format: slice entry points for the facade's hot *_in calls (local listing back to main's speed)

Since the M2 merge the facade handed in-memory files to the generic
`*_in` parsers as `&[u8]` (`with_bytes!`), which instantiates them in
the facade crate, where the format crate's private helpers do not
inline without LTO: listing a 400-group v1 file through `File::open`
was 7-10% slower than main. `ObjectHeader::parse_in`,
`group_v2::{resolve_child_in, resolve_group_children_in,
resolve_path_any_in}` and `attribute::{extract_attributes_tolerant_in,
find_attribute_in}` now pass a storage with `as_contiguous()` to their
non-generic slice entry point, compiled once in the format crate; other
storages reach the same generic core as before.

Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>

BENCHMARKS.md

@@ -484,6 +484,45 @@ explain the slower windows.
 
 ## Concurrent reads
 
+### Results after in-place chunk decoding (2026-09-26, tank, `c5334b1`)
+
+Same machine, files and commands, re-run after chunked reads started
+decoding into reusable per-thread buffers straight into the (typed) output,
+with the calling thread decoding alongside the pool. Load average 1.78 at
+the start; it rose to 6-9 during the runs (the clawhdf5 runs' own threads,
+and it stayed around 5-6 through the h5py runs, so something else was
+active). **This run was noisier than the previous one: h5py's own contiguous
+figures are about 40% lower than in the run below, and ours dropped
+similarly, so compare ratios within a run rather than MB/s across runs.**
+h5py was re-run in the same session.
+
+Each read decoding on its calling thread (`--decode-threads 1`, like h5py):
+
+| layout | mode | threads | clawhdf5 MB/s (eff) | h5py threads MB/s (eff) | h5py processes MB/s (eff) | vs h5py processes |
+|---|---|---:|---:|---:|---:|---:|
+| deflate | distinct | 1 | 670 (1.00) | 410 (1.00) | 397 (1.00) | 1.69x |
+| deflate | distinct | 4 | 2434 (0.91) | 406 (0.25) | 1470 (0.93) | 1.66x |
+| deflate | distinct | 8 | 3749 (0.70) | 406 (0.12) | 2398 (0.76) | 1.56x |
+| deflate | distinct | 16 | 4944 (0.46) | 390 (0.06) | 3135 (0.49) | 1.58x |
+| deflate | same | 1 | 211 (1.00) | 125 (1.00) | 124 (1.00) | 1.70x |
+| deflate | same | 16 | 1835 (0.54) | 122 (0.06) | 961 (0.48) | 1.91x |
+| contiguous | distinct | 1 | 6718 (1.00) | 5545 (1.00) | 5200 (1.00) | 1.29x |
+| contiguous | distinct | 16 | 11035 (0.10) | 4950 (0.06) | 10558 (0.13) | 1.05x |
+| contiguous | same | 1 | 14483 (1.00) | 2593 (1.00) | 2737 (1.00) | 5.29x |
+| contiguous | same | 16 | 132175 (0.57) | 2224 (0.05) | 14809 (0.34) | 8.93x |
+
+With the default rayon pool, deflate `distinct` reads 6143 MB/s from a single
+thread (15x h5py's 410 on one call) and 4556 MB/s at 16 threads (1.45x h5py
+processes); the other rows are within the noise of the table above.
+
+What changed: full reads of chunked datasets were 0.69x-0.76x of h5py
+processes at 16 threads in the run below, and are 1.58x here; with one
+thread they were 1.44x and are 1.69x. Minor page faults for the 16-thread
+run fell from about 4.6M to 0.2M (`/usr/bin/time -v`, provisional, loaded
+machine). clawhdf5 now reads faster than 16 h5py processes in every row of
+this benchmark except contiguous full reads at 16 threads, where both
+saturate memory bandwidth (1.05x).
+
 ### Results after the read fixes (2026-09-26, tank, `408f69e`)
 
 Same machine, files and commands as the first run below, re-run on an idle

CHANGELOG.md

@@ -2,6 +2,799 @@
 
 ## Unreleased
 
+### Range reads, milestone M3: remote files (2026-09-26)
+- **New crate `clawhdf5-remote`.** `open_url("http://host/file.h5")` gives
+  a `clawhdf5::File` (through `File::open_storage`) that reads the file by
+  HTTP `Range` requests; `storage_for_url` returns the cached storage to
+  read its statistics. Features: `http` (default; `ureq` without TLS, no
+  C), `https` (rustls with ring, which compiles C), `object-store` (any
+  `object_store` store, pure Rust), `s3`/`gcs`/`azure` (`s3://`, `gs://`,
+  `az://` URLs, configured from the environment; object_store's cloud
+  clients use aws-lc-rs, C).
+  - **`BlockCache`** wraps any `Storage`: aligned 1 MiB blocks (the size
+    `docs/design/range-reads.md` §2 measured), LRU with a byte budget
+    (64 MiB), the blocks one read misses fetched with one backend
+    `read_ranges` call as runs of consecutive blocks (a one-block gap is
+    fetched to merge two runs; at most 8 MiB per request). Readers on
+    several threads share it without holding its lock across a request, and
+    a block being fetched is waited for, not fetched again. A read that
+    misses more than half the budget is not kept (a large dataset does not
+    evict the metadata). A failed fetch is an error for every reader waiting
+    on it and is not cached.
+  - **`HttpStorage`**: opening is one ranged `GET` of the first block,
+    whose `Content-Range` gives the length. The file is pinned by its strong
+    `ETag` (`If-Match`) or else `Last-Modified` (`If-Unmodified-Since`), and
+    its length, checked on every response: a file changed while open is
+    `RemoteError::FileChanged`, never mixed data. A server that ignores
+    `Range` is refused without reading the body unless
+    `HttpOptions::allow_full_download`. Connection failures, timeouts,
+    `408`/`429`/`5xx` and cut-off bodies are retried with exponential
+    backoff; bodies are asked for with `Accept-Encoding: identity` and an
+    encoded one is refused. The ranges of one call are fetched in parallel.
+    A `200` answer to the first request whose body fits the range asked
+    for is taken as the whole file (a server may answer so for a small
+    file). Timeouts scale with the request: `HttpOptions::timeout` (30 s)
+    to connect and to get the headers, plus the body's size at
+    `HttpOptions::min_speed` (16 KiB/s), so a slow link is not cut off.
+    Redirects: at most `HttpOptions::max_redirects` (5), never from `https`
+    to `http`, and `HttpOptions::headers` are not sent to another origin.
+    No error or `Debug` output contains a URL's userinfo or query values
+    (`redact_url`; presigned URLs carry their signature there).
+  - **Hostile lengths**: the length a server claims is never used to
+    allocate. The cache's arithmetic is checked (a length near `u64::MAX`
+    used to overflow), a read spanning more than the budget is fetched
+    piece by piece with its output growing as data arrives, and
+    `download(storage, max_bytes)` reads a whole file only up to a limit
+    (`RemoteError::TooLarge` before any request otherwise).
+  - **`ObjectStoreStorage`** reads one object of any `object_store` store,
+    pinned by ETag (else version or modification time) and size. Each read
+    runs on a small tokio runtime the storage owns while the caller waits,
+    so it works from any thread, `spawn_blocking` and other runtimes
+    included.
+    `open_object(store, path, options)` opens a file through a block cache.
+  - Counted on the conformance corpus (tank, 2026-09-26,
+    `CLAWHDF5_REMOTE_CORPUS=conformance/.cache/corpus CLAWHDF5_REMOTE_REPORT=1
+    cargo test --release -p clawhdf5-remote --test http -- --nocapture corpus`):
+    the 621 files that open (254 MB) read over HTTP exactly as through
+    `File::open`; opening and listing them all (every group's entries, every
+    dataset's shape and type) took 640 requests and 55.5 MB, and then
+    reading each file's largest dataset under 64 MiB 96 more requests
+    (171 MB in all). Without the cache the same work is 141 936 requests.
+    Listing the 7.7 MB IMERG file (file A of the design's §2) takes 2
+    requests; the tests hold it to at most 3.
+- **`h5rs` takes URLs** with the new `remote` feature (`remote-https` for
+  `https://`): `ls`, `dump`, `stat` and `diff` read by range requests;
+  `check` downloads the file whole, since it validates every byte, up to
+  `--max-download N` (1 GiB by default). URLs are printed without their
+  credentials. The
+  tools now read through `File::storage` and the format crate's `*_in`
+  functions; local output is unchanged.
+- **`File::storage()`** (facade) returns the file's bytes from the
+  superblock on (the `as_bytes` view, cache image laid over) as a
+  `&(dyn Storage + Send + Sync)` for every backend, so code that parses a
+  file itself works on remote files too.
+- **`VlResolver::element_in` / `string_element_in`** (format): the
+  `element`/`string_element` lookups over any `Storage`.
+
+### Range reads, milestone M2: raw data and `File::open_storage` (2026-09-26)
+- **`clawhdf5::File::open_storage(Arc<dyn Storage + Send + Sync>)`** opens
+  a file served by any `clawhdf5_format::storage::Storage` and gives the
+  whole read API over it: groups and paths, datasets, attributes, the
+  `read_*` methods, selections, variable-length strings and sequences, and
+  virtual datasets. Every byte comes through `Storage::read_at` /
+  `read_ranges`; a user block is found and skipped, nothing past the
+  superblock's end of file is read, and a metadata cache image is laid over
+  the reads it covers (new `CacheImage::entries`). External virtual-dataset
+  sources are read through the new `File::set_vds_resolver` (any `File`;
+  without one a storage-backed file cannot follow them). `File::open` and
+  `File::from_bytes` keep their mmap and in-memory paths: the file's view
+  is now a `Storage` whose `as_contiguous()` is that buffer, and every hot
+  loop takes it. New exports: `clawhdf5::{Storage, SharedStorage,
+  VdsResolver}`, `File::contiguous_bytes()`.
+  - Over a storage without the whole file in memory the zero-copy methods
+    (`read_raw_ref`, `read_as_slice`, `read_*_zerocopy`) answer
+    `FormatError::ContiguousStorageRequired`, and `File::as_bytes` panics
+    (documented; use `contiguous_bytes`). The typed readers keep their fast
+    paths: a contiguous dataset is read in one piece and converted, and a
+    contiguous selection of a native type reads only its runs
+    (`data_read::read_selection_native_in`).
+- **Nothing in the format crate needs the whole file any more.** The
+  structures M1 left to `ContiguousStorageRequired` read through `Storage`:
+  v2 B-trees (`BTreeV2Header::parse_in`, `collect_btree_v2_records_in`,
+  `find_btree_v2_records_in`; one bounded read per node, whose size is known
+  before it is read), hence dense attributes, a SOHM B-tree index, huge
+  fractal-heap objects, and dense groups; v1 and v2 group listings, lookups
+  and paths (`group_v2::resolve_group_children_in`, `resolve_child_in`,
+  `resolve_path_any_in`, `group_v1::*_in`).
+- **Raw data reads through `Storage`**, each with a generic `*_in` core and
+  its `&[u8]` function as a thin wrapper (callers do not change):
+  `data_read` (`read_raw_data*_in`, `read_raw_data_selection_in`,
+  `read_chunked_native_in`), `chunked_read` (the v1 B-tree chunk index —
+  one read of each node's header, one of its entries — `list_chunks_in`,
+  and the full, cached, sweep and indexed reads), `parallel_read`,
+  `partial_read`, `fill_value` (`read_full_with_fill_in`,
+  `apply_to_unallocated_chunks_in`; `dataset_fill_value_from_storage` is
+  now generic, so a `&dyn Storage` still works), `vds`
+  (`read_virtual_dataset_in`, `virtual_dataset_extent_in`: the virtual
+  file through `Storage`, external source files still loaded whole through
+  the resolver), `vl_data` (`VlResolver<'a, S = [u8]>` with `new_in`;
+  `read_vl_strings_in`, `read_vl_bytes_in`),
+  `AttributeMessage::read_vl_strings_in`, `provenance::verify_dataset_in`.
+  - A chunked read first lists the chunks it needs, then fetches all their
+    stored bytes with **one `read_ranges` call** per batch of at most 64 MiB
+    (`storage::RAW_BATCH_BYTES`), so a remote backend can coalesce and
+    parallelise them, then decodes each batch as before (in parallel with
+    the `parallel` feature) before fetching the next. Every path that reads
+    chunks — full, cached, indexed, sweep, selection and the
+    `parallel_read` decoders — goes through the same batching, and no chunk
+    fetches more of its stored bytes than its decoded size can need (the
+    chunk size if unfiltered; else each applied filter's worst-case growth,
+    generously: `n + n/4 + 4096` per codec, unbounded only for a codec the
+    application registered). A crafted chunk index that points every chunk
+    at one huge extent therefore costs a bounded fetch, not
+    `chunks x extent` bytes (`tests/raw_fetch_bounds.rs`). Chunks the
+    file's chunk cache already holds are not fetched. A selection fetches
+    only the chunks its bounding box overlaps; a contiguous selection only
+    its runs, merged into reads of up to 8 MiB across gaps of up to 4 KiB
+    (a stride-2 selection of 32M `f32` is 32 reads and 0.3 s over a
+    `CountingStorage`, where one read per element was 16.8M reads, 2.0 s
+    and 2.1 GB peak). A global-heap collection is read once
+    per resolver and kept (within the resolver's 32 MiB budget).
+  - Each extent's bounds error is the one the slice readers gave, reported
+    when the read reaches that extent, so a damaged file fails with the
+    same error, in the same order, through either path.
+  - A backend that answers a read with more bytes than asked (breaking
+    `read_at`'s contract) never has the extra bytes used: every read is
+    cut to the range asked for (`storage::exact_len`), and a short answer
+    inside the file is an error.
+- **No behaviour change for in-memory and mapped files:** with
+  `as_contiguous()` every path slices the file as before (checked below).
+- **Speed on local files** (provisional: tank was shared with other jobs,
+  load 4–16 during the runs; Criterion `local_metadata_bench`, `main`
+  `8f59b2e` and this branch as separate binaries, 8 alternating rounds,
+  best round of each). Listing the 400-group version-1 file through the
+  facade (`File::open`, mmap) had become 7–10% slower than `main` (8.95–9.04
+  vs 8.14–8.36 ms), from the M2 merge on (bisected over the merges: `main`
+  8.17–8.27, `2893b6c` 8.69–8.76, `93e2d5f`/`7447dce`/`4313917` the same).
+  The parsers were unchanged: the facade now called the generic `*_in`
+  entry points with the slice (`with_bytes!`), which instantiates each
+  parser in the facade crate, where the format crate's private helpers do
+  not inline (no LTO); `main` called the `&[u8]` wrappers, compiled in the
+  format crate. Routing just `resolve_child`/`resolve_group_children` to
+  the wrappers took the listing from 8.66–8.70 to 8.33–8.46 ms. Fixed in
+  the format crate: `ObjectHeader::parse_in`, `group_v2::{resolve_child_in,
+  resolve_group_children_in, resolve_path_any_in}` and
+  `attribute::{extract_attributes_tolerant_in, find_attribute_in}` hand a
+  storage with `as_contiguous()` to their non-generic slice entry point,
+  compiled once in the format crate; everything else goes to the same
+  generic core as before, so `File::open_storage` is unchanged. Best
+  rounds, `main` / before / after: facade listing 8.49 / 9.07 / 8.68 ms
+  (+2.2% on `main`, was +6.9%); `ObjectHeader::parse` ×401 24.91 / 25.90 /
+  25.51 µs; symbol-table nodes 1.99 / 1.94 / 1.94 µs; group B-tree walk
+  360.6 / 365.0 / 361.1 ns. The B-tree walk's `btree_v1.rs` is identical to
+  `main` and it calls only format-crate code; its earlier +5–10% (323–355
+  vs 353–363 ns) was run-to-run layout noise (at `2893b6c` it measured
+  357–374 ns against `main`'s 355–371 in the same rounds).
+- Tests (2026-09-26, tank):
+  - `clawhdf5-format/tests/storage_equivalence.rs` now also reads every
+    dataset — whole, fill-aware, through a chunk cache (twice) and the
+    indexed path, three selections, virtual datasets with their sibling
+    sources, VL strings, sequences and bytes — through the read_at-only
+    `CountingStorage` and requires the slice results, and fails on any
+    `ContiguousStorageRequired`. With
+    `CLAWHDF5_STORAGE_CORPUS=conformance/.cache/corpus`, all 653 HDF5
+    files of the corpus agree (82 396 checks). The cached and indexed
+    paths are compared on values only when a read fails: they order
+    chunks by hash map, so which failing chunk a damaged dataset reports
+    varies between two caches even for the same slice (seen on
+    `cve-2025-2310.h5`; see `docs/known-issues.md`).
+  - A misbehaving storage (fails its N-th read; serves short reads) over
+    every fixture: each listing and dataset read is an error or exactly the
+    in-memory result, never other data (1 137 runs).
+  - A chunked read issues one `read_ranges` call with one range per chunk,
+    and a one-chunk selection one call with one range.
+  - `clawhdf5/tests/storage_equivalence.rs` reads every fixture (61 files)
+    and, with `CLAWHDF5_STORAGE_CORPUS`, every corpus file (701 files, 621
+    that open) through `File::open` and through `File::open_storage` over
+    `CountingStorage`: the tree, every attribute (all, and each by name),
+    every dataset's shape, types and values (all bytes, `f64`, `f32`,
+    `i64`, a box hyperslab, a strided one, out-of-order points, strings, VL
+    sequences) must be identical, and are — errors included, in full (open
+    errors and every read's). The one allowance is a line on which
+    `File::open` itself varies between two opens (the chunk cache lists a
+    damaged dataset's chunks in hash-map order, so which failing chunk a
+    full read reports varies: `cve-2025-2310.h5`), and then only if a fresh
+    `File::open` reproduces the storage's error. No storage read may answer
+    `ContiguousStorageRequired`. A storage that returns more bytes than
+    asked reads every fixture identically too.
+    It also counts what one pass — open, list, read every attribute and
+    every dataset once — asks of a storage with no cache: 176 092 `read_at`
+    calls and 208 MB for the 621 corpus files (254 MB of files); the most
+    are `h5stat_newgrat.h5` (35 001 groups: 92 489 calls) and
+    `ref_hdf5_compat1.nc` (16 062). A remote backend needs the block cache
+    of milestone M3. Command: `CLAWHDF5_STORAGE_CORPUS=… cargo test
+    --release -p clawhdf5 --test storage_equivalence -- --nocapture`.
+  - Conformance sweep (`conformance/run.sh --no-fetch`): 600 of 697 files
+    ok, `results.json` byte-identical to `8f59b2e`.
+
+### Correctness: Fletcher-32 (2026-09-26)
+- **Fletcher-32 checksums disagreed with libhdf5's on about one chunk in
+  32768** (fixed 2026-09-26). **Every release is affected, v2.1.0 through
+  v2.7.0**, both directions: `FileBuilder`/`FileWriter` (`with_fletcher32`)
+  and, before release, `FileEditor` wrote chunks that h5py and libhdf5
+  refuse ("filter returned failure during read"), and every reader
+  rejected valid libhdf5-written chunks with `Fletcher32Mismatch`. Our
+  checksum reduced its sums with `% 65535`; libhdf5's
+  `H5_checksum_fletcher32` uses the ones'-complement fold
+  `(s & 0xffff) + (s >> 16)`, which leaves 0xffff where the modulo leaves
+  0, so the two differ whenever a sum is a non-zero multiple of 65535.
+  `clawhdf5_format::checksum::fletcher32` (new, public) is a port of
+  `H5_checksum_fletcher32` and the only implementation; the filter writes
+  and verifies with it, and, as libhdf5 does, also accepts a stored
+  checksum with the bytes of each 16-bit half swapped (libhdf5 1.6.2 and
+  earlier) and the `% 65535` form v2.7.0 and earlier wrote, so their files
+  stay readable. Tests: `crates/clawhdf5/tests/fletcher32_interop.rs` compares
+  it with libhdf5's own function (through ctypes) on every 1- and 2-byte
+  input and 40 000 random and fold-heavy ones, and has h5py read
+  fold-case chunks written by `FileBuilder` and `FileEditor` and us read
+  h5py's. Files written by earlier releases read with a fixed build; to
+  make one readable by libhdf5, rewrite its Fletcher-32 datasets with a
+  fixed build (see `docs/known-issues.md`). `clawhdf5_accel::checksum_fletcher32` is a
+  different, textbook Fletcher-32 (sums start at 0xffff) and is not used
+  for HDF5.
+
+### In-place editing: version-2 B-tree indexes, shrinking, dense attributes (2026-09-26)
+- **`FileEditor` adds, moves and resizes chunks of datasets with two or
+  more unlimited dimensions** (version-2 B-tree chunk index, record types
+  10/11), as libhdf5's `H5B2` code does: `H5B2_update`'s insert-or-modify,
+  the preemptive split/redistribute loop, `split1`/`split_root` (depth
+  growth), `redistribute2/3`, and removal with `merge2/3`, root collapse
+  and the internal-record swap; node pointer widths and cumulative record
+  counts per depth; a missing index is created from the layout message's
+  parameters. After the same growth libhdf5's and the editor's trees are
+  node for node the same (tested through a depth increase).
+- **`FileEditor::resize` shrinks** along any dimension (h5py's
+  `Dataset.resize` to a smaller shape), as `H5D__chunk_prune_by_extent`
+  does, visiting the same chunks in the same order: chunks wholly outside
+  the new extent leave the index (version-1 B-tree removal with libhdf5's
+  sibling key and link fix-ups and empty-root case, version-2 B-tree
+  removal, Fixed/Extensible Array elements reset; an implicit index keeps
+  its chunks, as in libhdf5) and their space is freed; the part of a
+  partial edge chunk outside the extent is overwritten with the fill value,
+  so it reads as fill after a later growth. Growth under early allocation
+  now allocates and fills the new chunks (`H5D__chunk_allocate`), which an
+  implicit index needs. Shrinking was `Error::Unsupported`. Only the
+  chunks that exist are visited (placed in libhdf5's order), so shrinking
+  a sparse dataset costs memory and time in its chunks, not in the
+  coordinates cut off (a 2 x 10^12-coordinate shrink takes 0.6 s).
+- **`FileEditor::set_attr` handles dense attribute storage and creation
+  order**: objects that track (and index) attribute creation order; the
+  move to dense storage when an object reaches its compact limit (or an
+  attribute is too large for a header message), as `H5O__attr_create`
+  does it (new fractal heap, name index, creation-order index when
+  indexed, compact attributes moved over in header order); objects
+  already in dense storage (h5py- or clawhdf5-written): insertion,
+  same-size rewrites in place, other replacements by removal and
+  insertion. The heap is changed as `H5HF` changes it — best-fit free
+  sections from its free-space manager (kept as libhdf5 keeps `FSHD`/
+  `FSSE`), new direct blocks through the root indirect block (created,
+  doubled), blocks too small for an attribute skipped as libhdf5 skips
+  them (`H5HF__hdr_skip_blocks`: an indirect free section with its row
+  sections, serialized as libhdf5 serializes them, merged with the range
+  skipped just before it, and later attributes given skipped blocks from
+  either end or the middle of a range, which splits it), huge objects
+  through the huge-object B-tree (deleted with the last huge object),
+  removed objects' space merged back — with libhdf5's statistics: after
+  the same attribute workload the heap, its free space and both index
+  B-trees equal libhdf5's (`dense_skipped_blocks_match_libhdf5` covers
+  every way of skipping, with libhdf5 doing one edit per session as the
+  editor does). In a random attribute workload (1-4 KiB attributes among
+  small ones) 24% of `set_attr` calls were refused before skipping was
+  implemented; 2.2% are now, all replacements of the last attribute in a
+  heap block. Attributes are encoded as libhdf5
+  encodes them for a file h5py opens `r+` (message version 1, 3 for
+  non-ASCII names; simple dataspaces with their maximum dimensions).
+  Still refused: see `docs/known-issues.md`.
+- **Freed space is reused within an editing session.** A `FileEditor`
+  reuses (best fit, zeroed) what its earlier edits freed — moved filtered
+  chunks, pruned chunks, merged B-tree nodes, replaced heap blocks — never
+  what the current edit frees, and writes reused blocks with the new space
+  before any existing byte changes. `FileEditor::reusable_bytes`. The
+  append workload of `measure_append_waste` leaks less (sizes in
+  `docs/known-issues.md`).
+- **Reader: implicit chunk indexes below their maximum shape.** libhdf5
+  places an implicit index's chunks by their position in the *maximum*
+  chunk grid; the reader used the current grid and returned other chunks'
+  values from the second chunk row on (h5py early allocation with a fixed
+  `maxshape` larger than the shape).
+  `chunked_read::generate_implicit_chunks_in_grid` takes the maximum.
+- **Reader: object headers with long continuation chains.** A version-1
+  header whose continuation chunks chain more than 32 deep (a header that
+  gains a chunk per attribute added when full, as libhdf5 and the editor
+  grow it) was refused with `NestingDepthExceeded`; version-2 headers
+  stopped at 256 chunks. Chunks are now read one at a time from a queue,
+  in the order their continuation messages are found (libhdf5's
+  `H5O_protect` order, which the editor already used; a version-1
+  chunk's messages used to be inserted at its continuation message), each
+  buffer released before the next is read; a chunk address seen twice (a
+  cycle), chunks adding up to more than the file (a crafted chain of
+  chunks nested in each other made storage with owned buffers read and
+  hold the square of the file's size), or more than 65 536 chunks are
+  refused, so a header's chunks read at most the file's size.
+  The first version allocated a queue and a set of chunk starts for every
+  header and made `ObjectHeader::parse` of 401 small headers 1.8x slower
+  (`local_metadata_bench`: 45.7 vs 24.8 µs on `main` `8f59b2e`). The first
+  8 chunks now live in an inline array (cycle check by scan; only a longer
+  header allocates), and the per-chunk message loop is its own function
+  instead of being inlined into the generic parser. Provisional (tank load
+  3–4; Criterion, separate binaries, 2 alternating rounds): 24.96–25.12 vs
+  24.85–25.15 µs on `main`.
+- Tests: `crates/clawhdf5-tools/tests/edit_coverage_interop.rs` (h5py
+  `earliest`/`v110`/`latest` and clawhdf5-written files; structure
+  comparisons with libhdf5 for version-2 B-trees, shrink on every index,
+  and dense attribute heaps); the random-operation property test in
+  `edit_interop.rs` now shrinks, grows two unlimited dimensions and moves
+  attributes to dense storage (`CLAWHDF5_EDIT_SEED` for other seeds).
+
+### Name lookups through the name index (2026-09-26)
+- **Finding one link or attribute by name reads the name index, not every
+  entry.** In a dense group (links in a fractal heap) the v2 B-tree name
+  index (record type 5, lookup3 hash of the name) is descended to the
+  records with the name's hash and only their links are read — O(log n)
+  instead of all n. Path resolution (`File::dataset`, `resolve_path_any`,
+  soft-link targets) and `Group::dataset`/`Group::group` (on `File`,
+  `MmapFile` and `LazyFile`, which listed the whole group per call) use it;
+  names whose hashes collide are all compared, so the order libhdf5 gives
+  them does not matter. New `clawhdf5_format::group_v2::resolve_child`,
+  `btree_v2::find_btree_v2_records` (records in one key range), and a
+  `lookup-stats` feature counting heap objects read, for tests. Huge heap
+  objects are found through their index the same way.
+- **`attr(name)`** on the facade's groups and datasets (all three file
+  types): one attribute, found in dense storage through its name index
+  (record type 8) instead of reading every attribute
+  (`clawhdf5_format::attribute::find_attribute_in_file`).
+- **Two links of one name: the first wins everywhere.** A group cannot
+  validly hold two links of one name, but a damaged or hand-made one can.
+  The listing, `resolve_child` (`Group::dataset`/`group`) and path
+  resolution now all use only the first link of a name (header message
+  order in a compact group, name index order in a dense one) and ignore
+  the rest, even if the first dangles — libhdf5's rule for compact groups
+  (h5py fails to open a dangling first link although a later one
+  resolves). Before, the listing skipped a dangling first link and listed
+  the name via a later one that lookup did not follow, and path resolution
+  followed the last.
+- **B-tree v2 internal nodes are checksum-verified.** Lookups prune
+  children by internal-node keys, so a corrupted internal node could hide
+  a name with no error; a mismatch is now `ChecksumMismatch`, as in
+  libhdf5, for lookups and listings alike.
+- **`Group::entries()` and `File::group_at(address)`**: a listing's
+  `(name, address)` pairs, to open children without looking names up again.
+- Test: `crates/clawhdf5/tests/indexed_lookup_interop.rs` — every child of
+  an h5py-written 35 001-link group (with colliding hashes) opened by name
+  reads at most two links per lookup (before: 35 001), matches h5py, and
+  every link kind (soft, relative, dangling, external) resolves as h5py
+  resolves it in dense and compact groups.
+
+### Checked address conversion (2026-09-26)
+- **No 64-bit file value is truncated on a 32-bit target.** All 119
+  truncating `u64 as usize` casts in `clawhdf5-format` that clippy's
+  `cast_possible_truncation` reports, under every feature the crate is
+  built with in CI except `szip` (115 with default features and
+  `plugin-filters`, 4 more behind `parallel`), are gone: file addresses,
+  lengths and counts go through `addr::to_usize`, which fails with
+  `FormatError::Overflow` where the value does not fit (wasm32 and other
+  32-bit targets; it used to wrap onto another part of the file), and
+  in-memory counts through `addr::saturating_usize`. On 64-bit targets
+  nothing changes. `scripts/check-32bit-casts.sh` (run by `ci-test.sh`)
+  lints the crate with no default features, with default features, and
+  with every optional feature but `szip` (for wasm32; the set with `zstd`,
+  which does not build for wasm32, for the host), and fails on any new
+  truncating cast. The facade, `clawhdf5-io` and `clawhdf5-ann` are not
+  covered.
+
+### Range reads, milestone M1: the `Storage` trait (2026-09-26)
+- **Breaking: `clawhdf5_format::error::FormatError` and `clawhdf5::Error`
+  are now `#[non_exhaustive]`.** An exhaustive `match` on either needs a
+  wildcard arm. `FormatError` has two new variants: `Storage(String)` (a
+  storage backend failed to serve a read) and
+  `ContiguousStorageRequired(&'static str)` (an operation not yet converted
+  to range reads was asked of a backend without the whole file in memory).
+- New `clawhdf5_format::storage::Storage`, the synchronous, `no_std` read
+  interface of `docs/design/range-reads.md` option (a): `read_at(offset:
+  u64, len) -> Cow<[u8]>`, `read_ranges`, `len()` and an `as_contiguous()`
+  fast path; implemented for `[u8]`, `Vec<u8>`, and references, `Box`es
+  and (with `std`) `Arc`s of a `Storage`. Slices and `Vec`s serve borrowed
+  bytes, so parsing an in-memory file costs no copy.
+- **The metadata parsers read through `Storage`.** Each converted parser has
+  an `*_in<S: Storage + ?Sized>(&S, ..)` core (a `&dyn Storage` works too),
+  and its `&[u8]` function is now a thin wrapper over it, so no caller
+  changes. The wrappers compile to a `[u8]` instance of the same code, so a
+  structure read in memory is a bounds check and a borrowed slice, with no
+  indirect call and no copy. Converted: the superblock
+  (`Superblock::parse_in`), its extension and cache image
+  (`read_superblock_extension_in`, `cache_image_state_in`), object headers
+  with their continuation chunks (`ObjectHeader::parse_in`), local and
+  global heaps, symbol-table nodes and the group B-tree (v1), fractal heaps,
+  fixed and extensible array chunk indexes, shared messages and the SOHM
+  table (`message_data_in`, `message_data_with_sohm_in`,
+  `load_sohm_table_in`, …), attributes (`extract_attributes_full_in`,
+  `extract_attributes_tolerant_in`, `AttributeMessage::parse_in_storage`),
+  fill values (`dataset_fill_value_from_storage`) and virtual-dataset
+  mappings (`DataLayout::resolve_vds_mappings_in`); also
+  `signature::find_signature_in`. Each structure is read with bounded reads
+  (a prefix, then the structure) instead of slicing the whole file; the
+  open-ended `&file_data[addr..]` slices in these modules are gone. Bounds
+  errors keep their values (absolute position, file length).
+- Reads sized by untrusted fields are bounded by what the parser uses, so
+  a crafted size cannot turn one structure into a read of the rest of the
+  file on a range backend: local-heap names are read in growing pieces
+  (64 bytes first) rather than to the end of the data segment; a fractal
+  heap indirect block is read up to the entry covering the object (the
+  whole block only when that entry is unallocated); paged fixed and
+  extensible array data blocks over 1 MiB are read page by page, only the
+  pages in use; and a block under one checksum whose claimed size runs
+  past the end of the file fails its bounds check before any read (with
+  the `checksum` feature). An object header's prefix is one read (was
+  two).
+- Structures still indexed by a v2 B-tree — dense attribute storage, a SOHM
+  B-tree index and huge fractal-heap objects found through their B-tree —
+  are not converted yet (the
+  v2 B-tree and dense groups come with milestone M3); over a backend without
+  the whole file in memory they are the clean `ContiguousStorageRequired`
+  error, never a partial result. Raw data, chunk B-tree (v1) indexes and VL
+  data are milestone M2. (All of them read through `Storage` since M2,
+  above.)
+- **No behaviour change**, checked three ways (2026-09-26, tank): every
+  existing test passes unchanged; the conformance sweep
+  (`conformance/run.sh --no-fetch`) gives a byte-identical `results.json`
+  at `f2ff2c4` and on this branch, and identical per-file probe output for
+  696 of the 697 files — the exception, `cve-2025-2310.h5`, reports one of
+  two errors depending on which parallel chunk decode fails first, at
+  `f2ff2c4` as on this branch; and a transcript of every converted `&[u8]`
+  function's result over the fixtures, the conformance corpus and the
+  h5py-written files below (748 files, 7 603 object headers) is
+  byte-identical between the two builds.
+- **Speed on local files** (provisional: tank was shared with other jobs;
+  both builds linked into one binary and timed alternately, 200 rounds;
+  new Criterion bench `clawhdf5/benches/local_metadata_bench.rs` over a
+  400-group version-1 file, `clawhdf5-format/tests/fixtures/v1_groups_400.h5`):
+  against `f2ff2c4`, listing the file through the facade is 2.7% faster,
+  `ObjectHeader::parse` is within ±1%, symbol-table nodes and the group
+  B-tree walk are about 19% faster (their entry loops were tightened),
+  local-heap names and `resolve_group_children` 1.5–3% faster. The same
+  harness run on two copies of the old code differs by up to 2%. The
+  Criterion bench itself, old and new as separate binaries run alternately
+  (3 rounds), agrees except for `ObjectHeader::parse`, which it puts about
+  7% slower (25.9 vs 24.1 µs for 401 headers) while the listing that
+  parses those headers is 5–8% faster. **Rechecked on an idle tank
+  (2026-09-26, load 1.95 at the start, 3–4 during; Criterion, `main`
+  `479d8b4` vs this branch `b49ec39` as separate binaries, 2 alternating
+  rounds):** `ObjectHeader::parse` for 401 headers 24.0–24.1 µs → 25.5–25.8
+  µs (+6–7%, about 4 ns per header — real, not noise); symbol-table nodes
+  2.39 → 1.94–2.04 µs (−17%); group B-tree walk 405–411 → 322–351 ns
+  (−16%); listing the 400-group file through the facade 8.56–8.71 →
+  8.40–8.42 ms (−2.4%). Net, local metadata reads are slightly faster; the
+  per-header cost is a known, small regression. Data reads are unchanged:
+  `concurrent_read --decode-threads 1` (the `BENCHMARKS.md` "Concurrent
+  reads" workload), `main` and this branch alternating, 2 rounds each, on
+  the same idle start: every deflate and contiguous full-read row within
+  ±2.4% of `main` (the cache-bound contiguous hyperslab rows vary by up to
+  ±40% between `main`'s own rounds and are not comparable).
+- New equivalence harness `clawhdf5-format/tests/storage_equivalence.rs`:
+  every converted parser runs over the file as a slice and over
+  `storage::CountingStorage` — a `Storage` that serves an in-memory buffer
+  through `read_at` only (`as_contiguous()` is `None`), copying what it
+  serves and counting reads — and must give identical results. It walks
+  the fixtures, files h5py writes for it (extensible arrays with super
+  blocks and paged data blocks, paged fixed arrays, large v1 and dense
+  groups, a user block, SOHM list and B-tree indexes, dense, shared and
+  committed-type attributes, fixed and extensible array data blocks over
+  1 MiB, whole and truncated), and with `CLAWHDF5_STORAGE_CORPUS=<dir>` a
+  corpus (all 653 HDF5 files of the conformance corpus pass). Only the
+  structures listed above as not converted may answer
+  `ContiguousStorageRequired`, and only in the checks that reach them; a
+  converted parser falling back to the whole file fails it. Milestones M2
+  and M3 extend it.
+
+### ZFP (2026-09-26)
+- **ZFP (filter 32013, H5Z-ZFP) reads, in pure Rust.** It was the last
+  filter in the conformance corpus that failed with `UnsupportedFilter`.
+  New feature `zfp` (`clawhdf5-format` and `clawhdf5`, included in
+  `plugin-filters`, no dependencies) ports the zfp 1.0.1 decoder and the
+  decompression half of H5Z-ZFP 1.1.1: every mode (fixed rate, fixed
+  precision, fixed accuracy, reversible, expert), int32, int64, float and
+  double, 1-4-D fields with partial blocks, and headers written by
+  big-endian machines (values byte-swapped as H5Z-ZFP does). Read only:
+  there is no ZFP encoder. The decoder is deterministic, so lossy modes have
+  one right answer, and clawhdf5 returns exactly libzfp's values.
+- Tests: `crates/clawhdf5/tests/zfp_interop.rs` has h5py + hdf5plugin 7.1
+  (H5Z-ZFP 1.1.1, zfp 1.0.1) write 2205 datasets over 16 mode settings
+  (including expert parameters at their edges) x the four types x 1-4-D
+  shapes with partial edge chunks, partial blocks and unit chunk
+  dimensions x smooth, noisy, wide-range, zero and inf/NaN data; each must
+  read byte for byte as h5py reads it. `tests/zfp_alloc_bounds.rs` fuzzes
+  headers and streams under a counting allocator: no panics, and the output
+  is allocated only when it matches the chunk's size and the stream holds at
+  least a bit per block. A stream that ends before the decoder is done is an
+  error (libzfp reads past its buffer).
+- Conformance on tank (2026-09-26, `conformance/run.sh --no-fetch`): 600 of
+  697 files ok (599 before); `h5ex_d_zfp.h5` now reads.
+
+### In-place modification (2026-09-26)
+- **`clawhdf5::FileEditor` modifies an existing file where it lies.**
+  `FileBuilder` builds whole files in memory; the editor opens a file
+  written by libhdf5 (any `libver`, including HDF5 2.0's own format) or by
+  clawhdf5 and changes only what an edit touches, recomputing the checksum
+  of every structure it changes. It takes an exclusive `flock` on the file
+  (the lock libhdf5 takes), so a second editor gets `Error::Locked`.
+  - `write_selection` / `write_all` / `write_values`: overwrite values of a
+    compact, contiguous (also never-written, late-allocated) or chunked
+    dataset, in its own datatype, under any selection. Chunks are decoded,
+    updated and re-encoded through the dataset's filters; a chunk that no
+    longer fits moves to the end of the file. New chunks are added to
+    version-1 B-tree (every chunked dataset of h5py's default `libver`),
+    Extensible Array, Fixed Array and single-chunk indexes — creating the
+    index, its data blocks, super blocks and pages, and splitting B-tree
+    nodes, as libhdf5 does: after the same sequence of writes the B-tree has
+    the same number of nodes per level and the Extensible Array header the
+    same block statistics as libhdf5's (tested). Filters run as libhdf5's
+    `H5Z_pipeline` runs them (new
+    `clawhdf5_format::filters::compress_chunk_masked`): an optional filter
+    that fails — LZF or Blosc output no smaller than the chunk — is skipped
+    and its filter-mask bit set, so the chunk is stored exactly as h5py
+    stores it; a mandatory filter that fails fails the edit. (Storing such
+    a chunk LZF-encoded at the raw size with a clear mask let a later
+    libhdf5 rewrite of it keep the stale mask, and h5py could no longer
+    read the dataset.)
+  - `resize`: grow a chunked dataset up to its maximum dimensions (h5py's
+    `Dataset.resize`).
+  - `set_attr`: add or replace an attribute in an object header, in free
+    space or in a new continuation chunk at the end of the file. A
+    version-2 header (h5py `libver='v110'` and later) without an Attribute
+    Info message gets one, as libhdf5's `H5O__attr_create` adds it: libhdf5
+    counts such a header's attributes through that message, and without it
+    h5py reported `len(obj.attrs) == 0` while listing them.
+  - Each edit is planned in memory and refused as a whole
+    (`Error::Unsupported`, file untouched) when any part is not supported:
+    new chunks in a version-2 B-tree index (two or more unlimited
+    dimensions) or an implicit index, shrinking, variable-length and
+    reference data, chunks through a filter this build cannot encode
+    (scale-offset, N-Bit, SZIP), attributes in dense storage, past an
+    object's compact limit or with tracked creation order, files with a
+    metadata cache
+    image, paged or persistent free space, or marked open by another
+    writer. New error variants `Error::Unsupported`,
+    `Error::InvalidArgument`, `Error::Locked`, and `clawhdf5::Error` is now
+    `#[non_exhaustive]` — a breaking change for code that matches it
+    exhaustively (the Python bindings map the new variants to
+    `NotImplementedError`, `ValueError` and `OSError`).
+  - Durability: the new space (chunks, index blocks) is written and synced
+    before any existing byte changes, then the metadata that links it in,
+    then a second sync. There is no journal: a crash during the second
+    phase can leave the file inconsistent (as with libhdf5 without SWMR).
+    Freed space is not reused (see `docs/known-issues.md`; since reused
+    within an editing session, above).
+  - Tests: `crates/clawhdf5-tools/tests/edit_interop.rs` (h5py `earliest`,
+    `v114` and `latest` files and clawhdf5 files; after every round h5py
+    reads the expected values, h5dump and `h5rs check --data` accept the
+    file, and h5py `r+` modifies it further; random operations against a
+    model) and `crates/clawhdf5/tests/edit_tests.rs`.
+- `clawhdf5_format::type_builders::build_attr_message` is public.
+
+### Chunked full reads (2026-09-26)
+- **Chunks are decoded straight into the output, into reused buffers.** A
+  full read of a chunked dataset faulted in about three times its size in
+  fresh pages: every chunk was decoded into a new buffer per filter stage
+  (the cached reader behind `read_*` decoded 128 chunks at a time before
+  placing any; the uncached one behind `MmapFile`, `LazyFile` and
+  `verify_provenance` decoded the whole dataset first), then assembled into
+  a byte buffer, which the typed readers copied once more. Now each chunk
+  is decoded into buffers the thread keeps between chunks and reads
+  (`clawhdf5_format::filters::DecodeScratch`,
+  `decompress_chunk_exact_with`: deflate inflates into a kept buffer with a
+  reset inflater, shuffle into the other one, Fletcher32 is checked in
+  place; other filters go through the registry as before) and copied
+  directly to its place in the output. Chunks still go into the file's
+  chunk cache when the whole dataset fits. Selection reads decode the
+  chunks they touch the same way.
+- **Typed full reads of chunked data skip the byte buffer.** `read_f32`,
+  `read_f64`, `read_i32`, `read_i64` and `read_u64` (on `File`, `MmapFile`
+  and `LazyFile`) of a chunked dataset stored as that type in native byte
+  order decode every chunk into the returned `Vec` (huge-page backed when
+  large, like the byte readers' output); other types and byte orders
+  convert as before. New public
+  `clawhdf5_format::data_read::read_chunked_native`.
+- **Reading threads no longer wait for a busy rayon pool.** A full read
+  handed its chunks to rayon and the calling thread slept until the pool had
+  decoded them, so with a small pool (2-4 threads) readers outside it queued
+  behind its workers. The calling thread now decodes too, and pool workers
+  join in only when free; a helper the pool starts after the read has
+  finished returns at once. A single read still spreads over the default
+  pool. This replaces the one-thread-pool special case below for full
+  reads. Chunks are placed from several threads only when the chunk index
+  puts them on the chunk grid at distinct places; a corrupt index is read
+  one chunk at a time, and the error reported is still the first failing
+  chunk's. New test `crates/clawhdf5/tests/busy_decode_pool.rs`.
+- **Fixed:** in a filtered dataset, a chunk stored with every filter skipped
+  (filter mask) and shorter than a chunk read with zeros in place of its
+  missing part through `File`'s `read_*`; it is now an error naming the
+  chunk, as `MmapFile`/`LazyFile` already made it.
+- New h5py comparison `crates/clawhdf5/tests/chunked_read_paths_interop.rs`:
+  every chunked read path (cached and uncached full reads, `MmapFile`,
+  `LazyFile`, small, strided and point selections, with and without the
+  `parallel` feature) for 1-8-byte integers and 2-8-byte floats in both
+  byte orders, through deflate, shuffle, Fletcher32, LZF, SZIP and Blosc,
+  with partial edge chunks, sparse datasets and fill values, and datasets
+  larger than the chunk cache.
+
+### Writer: large dense indexes (2026-09-26)
+- **`track_order` orders attributes too, as h5py's `track_order=True`
+  does.** It tracked link creation order only, so h5py listed a tracked
+  object's attributes by name. A tracking object's header now has the
+  attribute creation order tracked and indexed flags, an Attribute Info
+  message with the next creation order (also for inline attributes), and a
+  creation order on each inline attribute message; dense attribute storage
+  gets a creation-order index (B-tree type 9). `FileWriter::track_order` /
+  `FileBuilder::track_order` now apply to datasets' attributes as well, and
+  `DatasetBuilder::track_order` sets it per dataset. Groups and datasets
+  that track order are written differently from before; others are
+  unchanged. More than 65 535 attributes on a tracking object is an error
+  (libhdf5's creation order counter is 2 bytes). The reader
+  (`attribute::extract_attributes*`) lists a tracking object's attributes
+  in creation order. Test `track_order_lists_attributes_in_creation_order`
+  (h5py lists, reads and extends them in "r+" mode, including libhdf5's
+  move from inline to dense storage).
+- **No more 65 535-record limit on the writer's v2 B-trees.** Dense link
+  storage (name index and creation-order index), dense attribute storage
+  and the chunk index of datasets with more than one unlimited dimension
+  were written as a single leaf, so a group with more than 65 535 links, an
+  object with more than 65 535 dense attributes, or such a dataset with more
+  than 65 535 chunks was an error. The writer now builds internal nodes to
+  any depth (`clawhdf5_format::btree_v2_write`), with node capacities and
+  child-pointer widths from the same arithmetic as libhdf5's
+  `H5B2__hdr_init` (shared with the reader, `btree_v2::node_info`) and
+  libhdf5's node sizes (512 bytes for dense storage, 2048 for chunks).
+  Indexes that fit the old one-leaf layout are written byte for byte as
+  before. New tests `crates/clawhdf5/tests/deep_btree_interop.rs` (100 000
+  links, 70 000 attributes, 200 000 chunks; h5py, h5dump, clawhdf5, and
+  h5py "r+" edits) and `check_files_with_deep_btrees` (`h5rs check`).
+- **Links and attributes whose name hashes collide are found by name.** The
+  dense name indexes (a group's links: B-tree v2 type 5; an object's
+  attributes: type 8) are ordered by the name's lookup3 hash and, when two
+  hashes are equal, by the name itself, as libhdf5 compares them. The writer
+  broke ties by insertion order, so libhdf5 could not open one of two
+  colliding names (`"k69209"` and `"k155448"` share hash `0x3a0b13e6`;
+  collisions are likely from about 77 000 names). Regression test
+  `names_whose_hashes_collide_are_found_by_name` in
+  `crates/clawhdf5/tests/writer_groups_interop.rs`.
+
+### Blosc2 (2026-09-26)
+- **Blosc2 (filter 32026) reads, in pure Rust.** Files written with
+  hdf5plugin's `Blosc2` failed with `UnsupportedFilter`. New feature
+  `blosc2` (`clawhdf5-format` and `clawhdf5`, included in `plugin-filters`)
+  decodes the Blosc2 contiguous frame hdf5-blosc2 stores per chunk, the
+  B2ND arrays it uses for chunks of 2 or more dimensions (blocks gathered
+  back into C order), Blosc2 chunks with their special values (zeros, NaN,
+  uninitialised, one repeated value), and the shuffle, bit-shuffle, delta
+  and truncate-precision filters, over the BloscLZ, LZ4/LZ4HC, Zlib and
+  Zstandard codecs shared with Blosc 1. Read only: there is no Blosc2
+  encoder. Dictionaries, lazy chunks, variable-length blocks, user-defined
+  codecs and registered Blosc2 filters (e.g. bytedelta) are errors;
+  uninitialised chunks read as zeros. Tested against h5py 3.16 +
+  hdf5plugin 7.1 (every codec, filter and level 0-9; 1- to 5-D chunks with
+  partial edge chunks; every integer width and f4/f8; datasets of zeros,
+  one value and NaN; Fletcher32 before Blosc2) and against frames from
+  python-blosc2 4.13.1 for what hdf5plugin never writes
+  (`crates/clawhdf5-format/tests/fixtures/blosc2/`); the decoder is
+  fuzzed. Conformance: 576 of 697 files ok (was 575) — h5ex_d_blosc2.
+- **A crafted Blosc2 chunk cannot allocate more than a few times its HDF5
+  chunk size.** Found in review before release: the sizes a frame declares
+  sized the decoder's buffers. A 173-byte frame whose offsets chunk claimed
+  2 GiB was decoded in full for a 1 MiB chunk; an empty chunk allocated its
+  declared block size twice (about 1 GiB); a B2ND chunk was decoded whole
+  with its padding (up to 16x the chunk); and a Zstandard stream's declared
+  window (up to 100 MiB) was reserved as the decoder was reused, which also
+  affected Blosc 1 and bitshuffle with Zstandard. Now the offsets chunk is
+  capped at the chunk size, block sizes are clamped to the chunk, B2ND
+  blocks are placed as they are decoded (padding is never held), and the
+  Zstandard window is capped at twice the stream's output (at least
+  128 KiB). A B2ND chunk may no longer be larger than its array, which
+  hdf5-blosc2 never writes. `tests/blosc2_alloc_bounds.rs` measures peak
+  allocation for these frames and for 45,000 fuzzed ones: at most 6x the
+  chunk size, twice the input and 2 MiB of Zstandard state.
+
+### Remaining conformance errors (2026-09-26)
+Conformance on tank, `conformance/run.sh --no-fetch`: 598 of 697 files
+ok (575 before). Of the 5 our-errors left, 3 are corrupt data HDF5 2.0
+reads only through a bug (listed in `CONFORMANCE.md`), 2 are the Blosc2 and
+ZFP filters; the 2 mismatches are the known h5py big-endian VL bug. The
+five files whose cache image libhdf5 cannot load (`cve-2025-6269-*`,
+`cve-2025-6516`) count as ok because the library, like libhdf5, opens them
+and fails their objects (see below).
+- **Metadata cache images are read.** A file written with a metadata cache
+  image keeps its metadata cache entries in an image block the superblock
+  extension points at, and libhdf5 reads them in place of the file's own
+  bytes; in `h5clear_mdc_image.h5` the root group exists only there, and
+  every reader failed with `InvalidObjectHeaderVersion(0)`. `File`,
+  `MmapFile` and `LazyFile` (and `h5rs`) now apply the image at open
+  (`clawhdf5_format::superblock_ext::CacheImage`), with libhdf5's checks.
+  The file is not copied to do it: a mapped file gets the image's entries
+  written into a private copy-on-write mapping
+  (`clawhdf5_io::HDF5Read::private_copy`, `MAP_PRIVATE`), so only the pages
+  they land on are copied, and a buffer the opener owns (`File::from_bytes`,
+  `open_buffered`) is patched in place; files without an image are read
+  from the mapping exactly as before. (An interim version copied the whole
+  file onto the heap: 2 GB of memory to open a 1 GiB sparse file with an
+  image, and an abort for an 8 GiB one; `tests/cache_image_memory.rs`
+  guards it.) An image entry that runs past the end of file is refused
+  (libhdf5 checks only its start; the images it writes never do this). A
+  file whose image libhdf5 cannot load (`cve-2025-6269-*`, `cve-2025-6516`)
+  opens, as in libhdf5, and every object lookup fails with the image's
+  error (`File`, `MmapFile`; `LazyFile` reads the root group at open, so
+  its open fails). libhdf5 fails only its first metadata read and then
+  reads the file's own, possibly stale, bytes; those are never read here.
+  An interim version refused such a file at `File::open` while the
+  conformance probe reported it as libhdf5 does, so the gate counted five
+  files as agreeing with h5py that the library did not open; probe and
+  library now take the decision from the same
+  `superblock_ext::cache_image_state`.
+- **Every other opener applies the superblock extension and the cache
+  image too** (`superblock_ext::apply_cache_image_in_place`, writing into
+  the buffer each already owns): `clawhdf5_io`'s `NativeVol` (at `open`,
+  and on read for `from_bytes`), `AsyncHDF5File`, `MpiVol` (a minimal edit
+  through the same `vol::load_hdf5`; the `mpi-io` feature cannot be built
+  without an MPI installation, so it was not compiled), and the external
+  source files of a virtual dataset. They read a file with an image from
+  its own bytes — stale metadata, or none (`h5clear_mdc_image.h5` failed
+  with `InvalidObjectHeaderVersion(0)`) — and skipped the extension checks
+  `File::open` makes. These readers cannot open a file and fail each
+  object, so an image libhdf5 cannot load is refused with the image's
+  error.
+- **The superblock extension is decoded at open, as libhdf5 does:** a File
+  Space Info or Metadata Cache Image message libhdf5 cannot decode makes the
+  open fail (`cve-2020-10810`, `cve-2020-10812` were opened).
+  `FormatError::InvalidSuperblockExtension`, `InvalidCacheImage`.
+- **Dataset storage libhdf5 refuses at open is refused at open**
+  (`FormatError::InvalidDatasetStorage`, `data_read::check_dataset_storage`):
+  an element count times element size that overflows (`cve-2024-32624`
+  `/Dset_OBJREF` opened and reported its shape), contiguous storage past the
+  end of the file, compact data of the wrong size. An empty contiguous
+  dataset at a defined address, which clawhdf5 up to v2.7.0 wrote, still
+  opens.
+- **Wrong or missing data fixed:**
+  - a simple dataspace of rank 0 holds one element (it held 0;
+    `cve-2020-18494`), and contiguous storage larger than the dataset reads
+    (`cve-2024-32623`, `cve-2025-2309`; libhdf5 ignores the excess);
+  - scale-offset returned wrong values for ordinary h5py files — see
+    *Correctness* below; E-scale is refused, as in libhdf5; codes past the
+    end of the chunk stay an error (`cve-2025-2308`, where HDF5 2.0 reads
+    past its buffer);
+  - shuffle uses its own parameter as the element size, as libhdf5 does
+    (`cve-2025-44905`);
+  - an unfiltered chunk the index records at other than the chunk's size is
+    refused (it read with zeros for the missing bytes; `cve-2025-44904`);
+  - a v1 B-tree chunk index is read as libhdf5 reads it: each chunk is
+    looked up the way `H5B_find` / `H5D__btree_cmp3` / `H5D__btree_found`
+    look it up, and a chunk that lookup does not find reads as fill values.
+    A key with a non-zero element-size coordinate is found in a 1-D dataset
+    and not in one of rank 2 or more (`cve-2025-44905`
+    `/Shuffle_float_data_le`, which read the chunk's data where h5py reads
+    fill values); an interim fix refused every such key, including 1-D
+    files libhdf5 reads correctly.
+- **Refused as libhdf5 refuses them:** a v1 group with an empty link name
+  fails its listing (`FormatError::InvalidLinkName`; lookups still work,
+  `cve-2021-46244`); dataspaces with more than 32 dimensions, a rank on a
+  scalar or null dataspace, or a dimension over its maximum
+  (`FormatError::InvalidDataspace`).
+- `ObjectHeader::object_class` classifies a header as libhdf5 does (a
+  dataset needs a datatype *and* a dataspace).
+- Conformance harness: user-defined links were listed as objects by the
+  reference, unopenable objects were not deduplicated, nested array types
+  were hashed wrong (`tarray3.h5`), and the attributes of objects h5py
+  cannot open were compared; all fixed. `CONFORMANCE.md` lists the corrupt
+  objects HDF5 2.0 reads through a bug (`bad_nbit_parms_walk.h5` among
+  them: libhdf5's own test now requires that read to fail).
+
 ### Concurrent reads (2026-09-26)
 - **Full reads of chunked datasets scale with threads again when rayon's
   pool has one thread.** Each full read handed its chunks to rayon to
@@ -372,6 +1165,7 @@
   missing feature ("unsupported filter: 32026 (Blosc2, not implemented by
   clawhdf5)").
 - **Not implemented:** Blosc2 (32026) and ZFP (32013) remain a clear error.
+  (Blosc2 reads since the `blosc2` feature, see above.)
 - **Wrong data: a chunk that decodes short read as zeros** (pre-existing, every
   filter). HDF5 stores every chunk at the full chunk size, so a filter
   pipeline that decodes to fewer bytes means a corrupt chunk; every chunk
@@ -787,6 +1581,59 @@
 - CI keeps zlib-ng building and tested; the arm64 job no longer needs cmake.
 
 ### Correctness
+- **`FileBuilder` stored LZF and Blosc chunks with filter mask 0 even when
+  the filter had not shrunk them** (fixed 2026-09-26). Latent in the
+  unreleased LZF/Blosc writer only (added 2026-09-26, "Plugin filters"):
+  no tagged release writes LZF or Blosc, so v2.7.0 and earlier are
+  unaffected. libhdf5 treats LZF and Blosc output no smaller than the chunk
+  as a filter failure and, both being optional filters, stores such a chunk
+  unfiltered with the filter's mask bit set. clawhdf5 stored the filter's
+  output with a clear mask. For an LZF chunk whose stream was exactly the
+  chunk's size (h5py stores `[182, 0, 0, 0, 0]` in a 5-byte chunk raw),
+  the first libhdf5 rewrite of that chunk stored the new data raw at the
+  same size and, the size being unchanged, kept the stale mask 0: h5py
+  then failed to read the dataset ("filter returned failure during read").
+  The whole-file writer now runs chunks through the pipeline as libhdf5
+  does (`clawhdf5_format::filters::compress_chunk_masked`, as `FileEditor`
+  already did) and records each chunk's real mask in every chunk index it
+  builds (single chunk, Fixed Array, Extensible Array, version-2 B-tree;
+  it builds no version-1 B-tree or implicit index), in the sequential and
+  `parallel` paths and `create_datasets_parallel`. Files whose chunks all
+  compress are byte-identical to before. `PrecompressedChunks::chunks` is
+  now `(raw size, stored bytes, filter mask)` (**breaking** for code that
+  reads it). Files written before the fix read correctly; rewrite them
+  (with this build or `h5repack`) before modifying them with libhdf5.
+  Tests: `plugin_filters_interop`
+  `skipped_optional_filters_are_masked_as_libhdf5_masks_them` (LZF,
+  shuffle+LZF+fletcher32 and Blosc, random, compressible and alternating
+  chunks, every index: masks equal an h5py-written twin's; after h5py r+
+  rewrites and extends the datasets, h5py, h5dump and our reader read every
+  value — before the fix 20 of 24 datasets had other masks than h5py's,
+  and h5py could not read the rewritten `[x, 0, 0, 0, 0]` datasets) and
+  `files_whose_chunks_all_compress_are_unchanged`; `chunked_write`
+  `skipped_lzf_chunks_are_masked_in_every_index`.
+- **Scale-offset data read wrong values in every release that decoded it
+  (v2.2.0 to v2.7.0), silently, on ordinary h5py files** (fixed
+  2026-09-26). Of 1480 scale-offset datasets h5py writes across every
+  integer type (`i1` .. `u8`), `f4` and `f8`, both byte orders, with and
+  without a fill value, and `scaleoffset` from 0 to the full width, 332 did
+  not read as h5py reads them: **151 returned wrong values with no error**
+  and 181 failed to read. The common cause was a chunk libhdf5 stores at
+  full width (`minbits` equal to the type's width), which it does for any
+  full-width `scaleoffset` and on its own whenever a chunk's values span
+  most of the type's range: `scaleoffset=0` integer data with a wide range
+  (82 datasets, all wrong values), full-width `u4`/`i4`/`u8`/`i8` (51 wrong
+  values; the narrower types and the rest failed with "truncated minval" or
+  "implausible minbits"), and `f4` D-scale data with a large range (18,
+  wrong values). Such a chunk holds the elements as they are; they were
+  decoded as offsets from `minval`. Also fixed, found on crafted files: the
+  packed codes start at byte 21 whatever size the chunk records for
+  `minval` (`cve-2025-44905` `/Scale_offset_short_data_be`), and a chunk
+  with `minbits` 0 and a fill value is all fill values (it read as
+  `minval`). The whole matrix is now an interop test
+  (`crates/clawhdf5/tests/scaleoffset_interop.rs`, generated by h5py at
+  test time, every dataset compared); on v2.7.0's decoder it reports the
+  332. See `docs/known-issues.md`.
 - **Corrupt files libhdf5 refuses are now refused instead of read.** On the
   HDF Group's CVE reproducers, 18 objects that libhdf5 (HDF5 2.0, through
   h5py) refuses to open were read by clawhdf5, some as wrong data (a chunk

CLAUDE.md

@@ -5,13 +5,13 @@ Pure-Rust HDF5 format implementation with HNSW vector search, WAL-backed persist
 
 ## Architecture
 
-Cargo workspace with 18 crates under `crates/` (plus `libaec-sys`, an internal FFI bindings crate for the optional `szip` feature):
+Cargo workspace with 19 crates under `crates/` (plus `libaec-sys`, an internal FFI bindings crate for the optional `szip` feature):
 
 | Crate | Role |
 |-------|------|
 | `clawhdf5-format` | HDF5 binary spec parser (superblock, B-tree, heap) — also holds shared type definitions and physical constants |
 | `clawhdf5-io` | Read/write implementation |
-| `clawhdf5-filters` | Deflate backends (zlib-rs, zlib-ng, Apple Compression); the HDF5 filter pipeline, the filter registry (`clawhdf5_format::filter_registry`) and the other codecs (LZ4, Zstd, SZIP, N-Bit, scale-offset, pcodec, and the pure-Rust plugin filters LZF, bitshuffle, bzip2, Blosc 1) live in `clawhdf5-format`. No Blosc2 or ZFP. |
+| `clawhdf5-filters` | Deflate backends (zlib-rs, zlib-ng, Apple Compression); the HDF5 filter pipeline, the filter registry (`clawhdf5_format::filter_registry`) and the other codecs (LZ4, Zstd, SZIP, N-Bit, scale-offset, pcodec, and the pure-Rust plugin filters LZF, bitshuffle, bzip2, Blosc 1, and Blosc2 and ZFP read-only) live in `clawhdf5-format`. |
 | `clawhdf5-derive` | Proc-macro derive for HDF5-serializable structs |
 | `clawhdf5` | Main facade crate |
 | `clawhdf5-netcdf4` | NetCDF-4 compatibility layer |
@@ -26,6 +26,7 @@ Cargo workspace with 18 crates under `crates/` (plus `libaec-sys`, an internal F
 | `clawhdf5-napi` | Node.js native addon bindings |
 | `clawhdf5-py` | PyO3 Python bindings |
 | `clawhdf5-wasm` | WebAssembly (wasm-bindgen) reader for the browser; demo in `examples/wasm-viewer/` |
+| `clawhdf5-remote` | Remote files: `open_url` over HTTP(S) range requests and object stores (`object_store`: S3, GCS, Azure) through a mandatory block cache (`BlockCache`) |
 | `clawhdf5-bench` | Benchmark suite |
 
 ## Key Features
@@ -150,6 +151,35 @@ Cargo workspace with 18 crates under `crates/` (plus `libaec-sys`, an internal F
   Alerts never block a save — drain them with `HDF5Memory::take_anomaly_alerts`.
   `MemorySource` for this bookkeeping is inferred from the caller-supplied
   `source_channel` string (a heuristic, not an authenticated trust boundary).
+- In-place modification: `clawhdf5::FileEditor` (`crates/clawhdf5/src/edit/`)
+  overwrites values, grows and shrinks chunked datasets (every chunk index,
+  version-2 B-trees included) and sets attributes (compact and dense
+  storage) in existing files (h5py- or clawhdf5-written) without rewriting
+  them, changing indexes and heaps as libhdf5 does (index shapes and heap
+  bookkeeping are compared with libhdf5's in the tests); space an edit
+  frees is reused by later edits of the same editor. Anything it cannot do
+  safely is `Error::Unsupported` before any write (limits in
+  `docs/known-issues.md`). Test changes with
+  `cargo test -p clawhdf5-tools --test edit_interop --test
+  edit_coverage_interop` (h5py, h5dump, `h5rs check`, structure comparisons
+  with libhdf5; libhdf5 sources for the algorithms are at
+  github.com/HDFGroup/hdf5, tag `hdf5_1_14_6`).
+- Remote files (`clawhdf5-remote`, range-read milestone M3 of
+  `docs/design/range-reads.md`): `open_url("http://…")` gives a
+  `clawhdf5::File` over `File::open_storage`, read through `BlockCache`
+  (1 MiB blocks, LRU byte budget, per-block in-flight dedup across threads,
+  runs coalesced into parallel requests). `HttpStorage` pins the file by
+  ETag/Last-Modified and length (a change is `RemoteError::FileChanged`),
+  refuses servers that ignore `Range` unless a full download is allowed,
+  and retries transient failures. `ObjectStoreStorage` (feature
+  `object-store`, pure Rust) runs each read on a small owned tokio
+  runtime and waits on a channel, so it works from any thread, including
+  inside `spawn_blocking` or another runtime. Default build is plain HTTP with
+  no C; `https` (rustls + ring) and `s3`/`gcs`/`azure` (aws-lc-rs) are
+  opt-in. Tests run a std-only HTTP server
+  (`tests/common/server.rs`, also the `range_server` example);
+  `CLAWHDF5_REMOTE_CORPUS=conformance/.cache/corpus` compares every corpus
+  file over HTTP with `File::open`.
 - GPU-accelerated vector distance computation (`clawhdf5-gpu`, wgpu); HDF5 I/O itself is CPU-only
 - Browser: `clawhdf5-wasm` (wasm-bindgen, read-only, file held in memory;
   no Zstd/SZIP since they link C) and the `examples/wasm-viewer/` page.

CONFORMANCE.md

@@ -13,15 +13,15 @@ fatal. This file is generated by `conformance/run.sh`; do not edit it by hand.
 
 | | |
 |---|---|
-| date | 2026-09-26 14:18 UTC |
+| date | 2026-09-27 00:34 UTC |
-| clawhdf5 commit | `73a01f1256fb9bf1b1e7601f755af9e8273cec4e` |
+| clawhdf5 commit | `f37e7ae3263277319dba4bc39be5397194eb00c3` |
 | machine | `tank`: AMD Ryzen 7 7800X3D 8-Core Processor, 16 CPUs, 61 GiB, Linux 7.0.0-34-generic x86_64 |
 | command | `conformance/run.sh --no-fetch --update-baseline` |
 | rustc | rustc 1.98.1 (48a229cea 2026-09-01) |
 | reference | h5py 3.16.0, HDF5 2.0.0, numpy 2.5.3, hdf5plugin 7.1.0, Python 3.14.4 |
 | h5dump | Version 1.14.6 (CVE corpus only) |
 | limits | 20 s timeout (SIGKILL), 4096 MiB address space, per process; 16 files in parallel |
-| runtime | 23 s probing + comparing (0 s fetch/build before it) |
+| runtime | 21 s probing + comparing (0 s fetch/build before it) |
 
 ## Results
 
@@ -36,16 +36,18 @@ A file's class is the first that applies:
 | corpus | files | ok | our-error | mismatch | h5py-cannot-read | panic | hang | crash | oom |
 |---|---|---|---|---|---|---|---|---|---|
 | NCAS-CMS_pyfive | 33 | 32 | 0 | 1 | 0 | 0 | 0 | 0 | 0 |
-| cve_hdf5 | 147 | 100 | 6 | 9 | 32 | 0 | 0 | 0 | 0 |
+| cve_hdf5 | 147 | 113 | 2 | 0 | 32 | 0 | 0 | 0 | 0 |
 | h5py_data | 4 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
-| hdf5 | 466 | 392 | 4 | 10 | 60 | 0 | 0 | 0 | 0 |
+| hdf5 | 466 | 404 | 1 | 1 | 60 | 0 | 0 | 0 | 0 |
 | netcdf-c | 20 | 20 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
 | netcdf4-python | 18 | 18 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
 | usnistgov_h5wasm | 5 | 5 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
 | xarray-data | 4 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
-| **all** | **697** | **575** | **10** | **20** | **92** | **0** | **0** | **0** | **0** |
+| **all** | **697** | **600** | **3** | **2** | **92** | **0** | **0** | **0** | **0** |
 
-2 of the 20 mismatches are a known h5py bug, not ours (see *Known not-our-bug*).
+2 of the 2 mismatches are a known h5py bug, not ours (see *Known not-our-bug*).
+
+3 of the 3 our-errors are corrupt data that HDF5 2.0 reads only through a bug and clawhdf5 refuses (see *Known not-our-bug*).
 
 Corpora (fetched by `conformance/fetch-corpus.sh` into the gitignored `conformance/.cache/`):
 
@@ -70,26 +72,13 @@ Grouped by normalised error message. *files* counts files whose class this cause
 
 | files | objects | error | examples |
 |---:|---:|---|---|
-| 3 | 3 | `DataSizeMismatch { expected: N, actual: N }` | `cve_hdf5/cvefiles/cve-2020-18494.h5`, `cve_hdf5/cvefiles/cve-2024-32623.h5`, `cve_hdf5/cvefiles/cve-2025-2309.h5` |
+| 3 | 3 | `ChunkedReadError("…")` | `cve_hdf5/cvefiles/cve-2025-2308.h5`, `cve_hdf5/cvefiles/cve-2025-44904.h5`, `hdf5/test/testfiles/bad_nbit_parms_walk.h5` |
-| 2 | 2 | `ChunkedReadError("…")` | `cve_hdf5/cvefiles/cve-2025-2308.h5`, `hdf5/test/testfiles/bad_nbit_parms_walk.h5` |
-| 2 | 2 | `UnsupportedFilter(N)` | `hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_blosc2.h5`, `hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_zfp.h5` |
-| 1 | 1 | `UnexpectedEof { expected: N, available: N }` | `cve_hdf5/cvefiles/cve-2019-9151.h5` |
-| 1 | 1 | `MissingMessage(Dataspace)` | `cve_hdf5/cvefiles/cve-2024-33874.h5` |
-| 1 | 1 | `InvalidObjectHeaderVersion(N)` | `hdf5/tools/test/testfiles/h5clear_mdc_image.h5` |
 
 ## Mismatch root causes
 
 | files | objects | cause | examples |
 |---:|---:|---|---|
-| 13 | 14 | `missing-object` | `cve_hdf5/cvefiles/cve-2019-8397.h5`, `cve_hdf5/cvefiles/cve-2019-8398.h5`, `cve_hdf5/cvefiles/cve-2021-46243.h5` (+10 more) |
-| 2 | 6 | `extra-attr` | `cve_hdf5/cvefiles/cve-2018-17438`, `cve_hdf5/cvefiles/cve-2018-17439` |
 | 1 | 1 | `attr-values: ours=vlen(>u8) h5py=object layout=- filters=-` | `NCAS-CMS_pyfive/tests/data/attr_datatypes.hdf5` |
-| 1 | 4 | `extra-object` | `cve_hdf5/cvefiles/cve-2021-46244.h5` |
-| 1 | 1 | `values: ours=<f4 h5py=float32 layout=chunked filters=-` | `cve_hdf5/cvefiles/cve-2025-44904.h5` |
-| 1 | 1 | `values: ours=>i2 h5py=>i2 layout=chunked filters=[6]` | `cve_hdf5/cvefiles/cve-2025-44905.h5` |
-| 1 | 1 | `values: ours=>f4 h5py=>f4 layout=chunked filters=[2]` | `cve_hdf5/cvefiles/cve-2025-44905.h5` |
-| 1 | 1 | `values: ours=<f4 h5py=float32 layout=chunked filters=[2]` | `cve_hdf5/cvefiles/cve-2025-44905.h5` |
-| 1 | 1 | `values: ours=((<i4)[6, 3])[4] h5py=(('<i4', (6, 3)), (4,)) layout=contiguous filters=-` | `hdf5/tools/test/testfiles/tarray3.h5` |
 | 1 | 1 | `values: ours=vlen({r:>f4,i:>f4}8) h5py=object layout=contiguous filters=-` | `hdf5/tools/test/testfiles/tcomplex_be.h5` |
 
 ## CVE corpus: clawhdf5 vs h5dump vs h5py
@@ -102,7 +91,7 @@ columns are.
 
 | tool | read | error | panic | crash | hang | oom |
 |---|---:|---:|---:|---:|---:|---:|
-| clawhdf5 | 140 | 7 | 0 | 0 | 0 | 0 |
+| clawhdf5 | 121 | 26 | 0 | 0 | 0 | 0 |
 | h5dump 1.14.6 | 16 | 129 | 0 | 2 | 0 | 0 |
 | h5py 3.16.0 / HDF5 2.0.0 | 115 | 31 | 0 | 1 | 0 | 0 |
 
@@ -156,27 +145,27 @@ columns are.
 | cvefiles/cve-2018-17435.h5 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
 | cvefiles/cve-2018-17436 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2018-17437.h5 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
-| cvefiles/cve-2018-17438 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | mismatch |
+| cvefiles/cve-2018-17438 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
-| cvefiles/cve-2018-17439 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | mismatch |
+| cvefiles/cve-2018-17439 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
 | cvefiles/cve-2019-8396.h5 | error exit | read 3 obj, 2 errors | read 3 obj, 2 errors | ok |
-| cvefiles/cve-2019-8397.h5 | error exit | read 3 obj, 2 errors | read 2 obj, 1 errors | mismatch |
+| cvefiles/cve-2019-8397.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
-| cvefiles/cve-2019-8398.h5 | error exit | read 3 obj, 2 errors | read 2 obj, 1 errors | mismatch |
+| cvefiles/cve-2019-8398.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
-| cvefiles/cve-2019-9151.h5 | error exit | read 3 obj, 1 errors | read 3 obj, 2 errors | our-error |
+| cvefiles/cve-2019-9151.h5 | error exit | read 3 obj, 1 errors | read 3 obj, 2 errors | ok |
 | cvefiles/cve-2019-9152.h5 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
 | cvefiles/cve-2020-10809 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2020-10810.h5 | error exit | open error | read 2 obj | h5py-cannot-read |
+| cvefiles/cve-2020-10810.h5 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2020-10811.h5 | error exit | read 25 obj, 1 errors | read 25 obj, 1 errors | ok |
-| cvefiles/cve-2020-10812.h5 | error exit | open error | read 2 obj | h5py-cannot-read |
+| cvefiles/cve-2020-10812.h5 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2020-18232.h5 | error exit | read 3 obj, 2 errors | read 3 obj, 2 errors | ok |
-| cvefiles/cve-2020-18494.h5 | ok | read 2 obj | read 2 obj, 1 errors | our-error |
+| cvefiles/cve-2020-18494.h5 | ok | read 2 obj | read 2 obj | ok |
 | cvefiles/cve-2021-36977.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2021-37501.h5 | error exit | read 18 obj, 1 errors | read 18 obj, 1 errors | ok |
 | cvefiles/cve-2021-45829.h5 | error exit | read 1 obj, 2 errors | read 1 obj, 1 errors | ok |
-| cvefiles/cve-2021-45830.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2021-45830.h5 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2021-45833.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
 | cvefiles/cve-2021-46242.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
-| cvefiles/cve-2021-46243.h5 | error exit | read 3 obj, 2 errors | read 2 obj, 1 errors | mismatch |
+| cvefiles/cve-2021-46243.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
-| cvefiles/cve-2021-46244.h5 | error exit | read 2 obj, 1 errors | read 6 obj, 4 errors | mismatch |
+| cvefiles/cve-2021-46244.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
 | cvefiles/cve-2024-29157.h5 | error exit | read 4 obj, 7 errors | read 4 obj, 7 errors | ok |
 | cvefiles/cve-2024-29158.h5 | ok | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
 | cvefiles/cve-2024-29159.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
@@ -201,50 +190,50 @@ columns are.
 | cvefiles/cve-2024-32615.h5 | error exit | read 4 obj, 1 errors | read 4 obj, 1 errors | ok |
 | cvefiles/cve-2024-32616.h5 | error exit | read 10 obj, 7 errors | read 10 obj, 6 errors | ok |
 | cvefiles/cve-2024-32617.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
-| cvefiles/cve-2024-32618.h5 | error exit | read 4 obj, 2 errors | read 3 obj, 1 errors | mismatch |
+| cvefiles/cve-2024-32618.h5 | error exit | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
 | cvefiles/cve-2024-32619.h5 | error exit | read 3 obj, 2 errors | read 3 obj, 2 errors | ok |
 | cvefiles/cve-2024-32620.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2024-32621.h5 | ok | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
 | cvefiles/cve-2024-32622.h5 | ok | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
-| cvefiles/cve-2024-32623.h5 | ok | read 6 obj | read 6 obj, 1 errors | our-error |
+| cvefiles/cve-2024-32623.h5 | ok | read 6 obj | read 6 obj | ok |
-| cvefiles/cve-2024-32624.h5 | error exit | read 6 obj, 1 errors | read 6 obj | ok |
+| cvefiles/cve-2024-32624.h5 | error exit | read 6 obj, 1 errors | read 6 obj, 1 errors | ok |
 | cvefiles/cve-2024-33873.h5 | error exit | read 4 obj, 1 errors | read 4 obj, 1 errors | ok |
-| cvefiles/cve-2024-33874.h5 | ok | read 6 obj, 1 errors | read 6 obj, 2 errors | our-error |
+| cvefiles/cve-2024-33874.h5 | ok | read 6 obj, 1 errors | read 6 obj, 1 errors | ok |
 | cvefiles/cve-2024-33875.h5 | ok | read 2 obj | read 2 obj | ok |
 | cvefiles/cve-2024-33876.h5 | ok | read 3 obj, 1 errors | read 3 obj, 1 errors | ok |
 | cvefiles/cve-2024-33877.h5 | error exit | read 8 obj, 1 errors | read 8 obj, 1 errors | ok |
 | cvefiles/cve-2025-2153.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
 | cvefiles/cve-2025-2308.h5 | error exit | read 25 obj, 1 errors | read 25 obj, 2 errors | our-error |
-| cvefiles/cve-2025-2309.h5 | ok | read 6 obj, 1 errors | read 6 obj, 1 errors | our-error |
+| cvefiles/cve-2025-2309.h5 | ok | read 6 obj, 1 errors | read 6 obj | ok |
 | cvefiles/cve-2025-2310.h5 | error exit | read 24 obj, 8 errors | read 24 obj, 8 errors | ok |
-| cvefiles/cve-2025-2912.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-2912.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-2913.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-2913.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-2914.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-2914.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-2915.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-2915.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-2923.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-2923.h5 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2025-2924.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2025-2925.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
-| cvefiles/cve-2025-2926.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-2926.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-44904.h5 | error exit | read 25 obj, 1 errors | read 25 obj, 1 errors | mismatch |
+| cvefiles/cve-2025-44904.h5 | error exit | read 25 obj, 1 errors | read 25 obj, 2 errors | our-error |
-| cvefiles/cve-2025-44905.h5 | error exit | read 25 obj, 3 errors | read 25 obj, 3 errors | mismatch |
+| cvefiles/cve-2025-44905.h5 | error exit | read 25 obj, 3 errors | read 25 obj, 3 errors | ok |
 | cvefiles/cve-2025-6269-1.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2025-6269-2.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2025-6269-3.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2025-6269-4.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
-| cvefiles/cve-2025-6270-1.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-6270-1.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-6270-2.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-6270-2.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-6270-3.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-6270-3.h5 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2025-6516.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2025-6750.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
-| cvefiles/cve-2025-6816.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-6816.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-6817.h5 | error exit | open error | read 1 obj | h5py-cannot-read |
+| cvefiles/cve-2025-6817.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-6818.h5 | error exit | open error | read 1 obj | h5py-cannot-read |
+| cvefiles/cve-2025-6818.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-6856.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-6856.h5 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2025-6857.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
-| cvefiles/cve-2025-6858.h5 | SIGSEGV | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-6858.h5 | SIGSEGV | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2025-7067.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
-| cvefiles/cve-2025-7068.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-7068.h5 | error exit | open error | open error | h5py-cannot-read |
-| cvefiles/cve-2025-7069.h5 | error exit | open error | read 1 obj, 1 errors | h5py-cannot-read |
+| cvefiles/cve-2025-7069.h5 | error exit | open error | open error | h5py-cannot-read |
 | cvefiles/cve-2026-26200.h5 | error exit | read 1 obj, 1 errors | read 1 obj, 1 errors | ok |
 | cvefiles/cve-2026-34734.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
 | cvefiles/cve-2026-92627.h5 | error exit | read 2 obj, 1 errors | read 2 obj, 1 errors | ok |
@@ -275,6 +264,11 @@ columns are.
 - **Types h5py widens.** Where h5py reads a type into a numpy type of a different size
   (FP8 -> float16, bfloat16 -> float32, x87 long double -> float128) the values are not
   compared (shape and presence still are): dataset file type size 1 -> numpy float16 (2) (15x), attr file type size 1 -> numpy float16 (2) (15x), dataset file type size 2 -> numpy float32 (4) (2x), dataset file type size 8 -> numpy float128 (16) (1x), dataset file type size 12 -> numpy float128 (16) (1x), attr file type size 2 -> numpy float32 (4) (1x), dataset file type size 2 -> numpy >f4 (4) (1x), attr file type size 2 -> numpy >f4 (4) (1x).
+- **Corrupt data HDF5 2.0 reads through a bug.** clawhdf5 refuses these objects; h5py 3.16 /
+  HDF5 2.0 returns values for them that the file does not hold:
+  - `cve_hdf5/cvefiles/cve-2025-2308.h5` `/Scale_offset_long_long_data_le`: scale-offset codes run past the end of the chunk: HDF5 2.0 reads past its buffer; libhdf5's develop branch refuses the chunk ("Buffer too short").
+  - `cve_hdf5/cvefiles/cve-2025-44904.h5` `/Scale_offset_float_data_le`: unfiltered chunks of 38 and 37 bytes for 48-byte chunks: HDF5 2.0 fills the rest with whatever its buffer held; libhdf5's develop branch refuses them ("incorrect chunk size returned from index for unfiltered chunk").
+  - `hdf5/test/testfiles/bad_nbit_parms_walk.h5` `/Nbit_int_data_le`: an N-Bit parameter list one value short: HDF5 2.0 reads past the list; libhdf5's own test (`test_filter_bad_params`, test/dsets.c) now requires the read to fail.
 - **References** are compared by presence only (`R`), not by target.
 
 ## Objects h5py fails on but clawhdf5 reads
@@ -282,7 +276,6 @@ columns are.
 - 19 x `OSError: Can't synchronously read data (no appropriate function for conversion path)`
 - 1 x `TypeError: unhandled dtype kind M (dtype('…'))`
 - 1 x `TypeError: No NumPy equivalent for TypeTimeID exists`
-- 1 x `KeyError: "…"`
 - 1 x `ValueError: Insufficient precision in available types to represent (N, N, N, N, N)`
 
 ## Reproduce

Cargo.toml

@@ -18,6 +18,7 @@ members = [
     "crates/clawhdf5-bench",
     "crates/clawhdf5-tools",
     "crates/clawhdf5-wasm",
+    "crates/clawhdf5-remote",
     "crates/libaec-sys",
 ]
 resolver = "2"

README.md

@@ -96,6 +96,12 @@ breaking change, are in [CHANGELOG.md](CHANGELOG.md).
   filtered top-k, never slower than unfiltered), and opt-in re-ranking and
   confidence rejection, which used to be reachable only through `ClawhdfBackend`.
 
+**Remote files (unreleased)**
+- New crate `clawhdf5-remote`: `open_url("http://…")` reads a file on an
+  HTTP server (or in S3/GCS/Azure, opt-in) by range requests through a
+  block cache, without downloading it; `h5rs` takes URLs with its `remote`
+  feature. See [Reading remote files](#reading-remote-files).
+
 **Tooling**
 - CI now runs the h5py/netCDF4 interop suites for real (they had been skipping
   silently) and runs an aarch64 job for the NEON kernels.
@@ -433,6 +439,71 @@ b.write("groups.h5")?;
 A group holds at most 65 535 links; more is an error, as is a link over
 65 515 bytes (a very long soft-link target) in a group of more than 8 links.
 
+### Modifying an existing file
+
+```rust
+use clawhdf5::{AttrValue, FileEditor, Selection};
+
+// A file from h5py or clawhdf5, dataset "x" chunked with maxshape=(None,).
+let mut ed = FileEditor::open("data.h5")?; // exclusive lock, like libhdf5
+ed.resize("x", &[1100])?;                   // h5py: ds.resize((1100,))
+let sel = Selection::Hyperslab { start: vec![1000], stride: vec![1], count: vec![100], block: vec![1] };
+ed.write_values("x", &sel, &[0.5f64; 100])?; // ds[1000:1100] = 0.5
+ed.set_attr("x", "units", &AttrValue::String("m/s".into()))?;
+ed.resize("x", &[900])?;                    // shrinking prunes chunks, like h5py
+```
+
+Each call changes the file in place (no rewrite) and syncs it. Any chunk
+index (version-2 B-trees for several unlimited dimensions included) and
+attributes in compact or dense storage are handled as libhdf5 handles
+them; space an edit frees is reused by later edits of the same editor. What it
+cannot change safely is refused before anything is written; see
+[known issues](docs/known-issues.md) for the limits.
+
+### Reading remote files
+
+[`clawhdf5-remote`](crates/clawhdf5-remote/README.md) opens a file on an
+HTTP server (or, with its `s3`/`gcs`/`azure` features, in an object store)
+without downloading it: the read API is the same `clawhdf5::File`, and
+only the bytes an operation needs are fetched, by `Range` requests through
+a block cache (1 MiB blocks; opening fetches the first one). A file that
+changes on the server while it is open is an error, never a mix of old and
+new bytes.
+
+```rust
+let file = clawhdf5_remote::open_url("http://127.0.0.1:8000/tall.h5")?;
+let values = file.dataset("/g2/dset2.1")?.read_f64()?;
+```
+
+To try it without a server of your own, the crate's test server serves a
+directory with range support:
+
+```bash
+cargo run -p clawhdf5-remote --example range_server -- crates/clawhdf5/tests/fixtures 127.0.0.1:8000
+# in another shell: list the file, read one dataset, print what it cost
+cargo run -p clawhdf5-remote --example read_url -- http://127.0.0.1:8000/tall.h5 /g2/dset2.1
+```
+
+```text
+/g1  group
+/g2  group
+/g2/dset2.1  dataset [10] F32
+/g2/dset2.2  dataset [3, 5] F32
+/g1/g1.1  group
+/g1/g1.2  group
+/g1/g1.2/g1.2.1  group
+/g1/g1.1/dset1.1.1  dataset [10, 10] I32
+/g1/g1.1/dset1.1.2  dataset [20] I32
+/g2/dset2.1: 10 values, first [1.0, 1.100000023841858, 1.2000000476837158, ...]
+1 range requests (the one at open included), 9968 bytes fetched, 9968 bytes cached
+```
+
+(`tall.h5` is 9 968 bytes, so the first block holds all of it.) `h5rs`
+built with `--features remote` takes the same URLs:
+`h5rs ls -r http://127.0.0.1:8000/tall.h5`. Plain HTTP builds no C;
+`https://` is the `https` feature (rustls with ring, which compiles C).
+Limits are in [known issues](docs/known-issues.md).
+
 ### Python
 
 `crates/clawhdf5-py` is a Python package (PyO3 + numpy) that reads HDF5 with
@@ -661,19 +732,20 @@ let exported = backend.export_markdown("MEMORY.md")?;
 ## Crate Map
 
 ```
-clawhdf5 workspace (17 crates, ~86K lines of Rust in src/, ~104K with tests
+clawhdf5 workspace (19 crates, ~86K lines of Rust in src/, ~104K with tests
                      and benches; plus libaec-sys, an internal FFI bindings
                      crate for the optional szip feature)
 │
 ├── Core HDF5
 │   ├── clawhdf5-format      — Binary parser/writer (no_std-capable), shared type definitions
 │   ├── clawhdf5-io          — I/O abstraction (file/memory readers; optional mmap, async, HSDS, MPI)
-│   ├── clawhdf5-filters     — Fast deflate path (zlib-ng); the filter registry and the lz4/zstd/pcodec/szip/LZF/bitshuffle/bzip2/Blosc filters live in clawhdf5-format
+│   ├── clawhdf5-filters     — Fast deflate path (zlib-ng); the filter registry and the lz4/zstd/pcodec/szip/LZF/bitshuffle/bzip2/Blosc/Blosc2 filters live in clawhdf5-format
 │   ├── clawhdf5-derive      — Proc macros
 │   ├── clawhdf5             — High-level API
 │   ├── clawhdf5-netcdf4     — NetCDF-4 support
 │   ├── clawhdf5-accel       — SIMD (AVX2, NEON incl. SDOT int8; AVX-512 behind `avx512`)
-│   └── clawhdf5-gpu         — GPU compute (wgpu, hand-written WGSL compute shaders)
+│   ├── clawhdf5-gpu         — GPU compute (wgpu, hand-written WGSL compute shaders)
+│   └── clawhdf5-remote      — Remote files: HTTP(S) range requests, object stores, block cache
 │
 ├── Agent Memory
 │   ├── clawhdf5-agent       — Memory engine (24.7K lines, 32 modules; chained-CRC WAL)
@@ -688,6 +760,7 @@ clawhdf5 workspace (17 crates, ~86K lines of Rust in src/, ~104K with tests
 │   └── clawhdf5-wasm        — Browser (WebAssembly, wasm-bindgen; read-only)
 │
 └── Tooling
+    ├── clawhdf5-tools       — h5rs: ls, dump, stat, diff, check
     └── clawhdf5-bench       — Benchmark suite
 ```
 
@@ -779,13 +852,15 @@ stores keep their setting. Opt out with `float16 = false` or
 | `bitshuffle` | no | Bitshuffle filter (id 32008) with its LZ4 and Zstandard modes: read and write. Pure Rust (lz4_flex, ruzstd) |
 | `bzip2` | no | bzip2 filter (id 307): read and write. Pure Rust (the `bzip2` crate's libbz2-rs-sys backend compiles no C) |
 | `blosc` | no | Blosc 1 filter (id 32001): reads BloscLZ, LZ4/LZ4HC, Snappy, Zlib and Zstandard frames with byte or bit shuffle; writes LZ4, Snappy, Zlib or Zstandard (not BloscLZ). Pure Rust |
-| `plugin-filters` | no | All four above |
+| `blosc2` | no | Blosc2 filter (id 32026), read only: hdf5plugin's frames and B2ND (n-D) chunks, BloscLZ, LZ4/LZ4HC, Zlib and Zstandard, with shuffle, bit shuffle, delta or truncated precision. Pure Rust |
+| `zfp` | no | ZFP filter (id 32013, H5Z-ZFP), read only: every mode (rate, precision, accuracy, reversible, expert) for int32, int64, float and double, 1-4-D, returning exactly libzfp's values. Pure Rust, no dependencies |
+| `plugin-filters` | no | All six above |
 
-Blosc2 (32026) and ZFP (32013) are not implemented: reading them fails with
+clawhdf5 cannot write Blosc2 or ZFP. Any other
-`UnsupportedFilter`, whose message names the filter. Any other filter can be
+filter can be supplied at run time with `filter_registry::register_filter` (a
-supplied at run time with `filter_registry::register_filter` (a decoder
+decoder closure, or a `FilterCodec` that also encodes). The facade
-closure, or a `FilterCodec` that also encodes). The facade (`clawhdf5`)
+(`clawhdf5`) forwards `lzf`, `bitshuffle`, `bzip2`, `blosc`, `blosc2`, `zfp`
-forwards `lzf`, `bitshuffle`, `bzip2`, `blosc` and `plugin-filters`. Write
+and `plugin-filters`. Write
 with `DatasetBuilder::with_lzf()`, `with_bitshuffle(..)`, `with_bzip2(..)`
 and `with_blosc(..)`; h5py + hdf5plugin read the result (tested both ways in
 `crates/clawhdf5/tests/plugin_filters_interop.rs`). The pure-Rust Zstandard

conformance/baseline.json

@@ -1,15 +1,15 @@
 {
  "comment": "conformance/run.sh fails if the ok count drops below `ok` or a file in `ok_files` stops being ok. Regenerate with `conformance/run.sh --update-baseline` after an intended change.",
- "commit": "73a01f1256fb9bf1b1e7601f755af9e8273cec4e",
+ "commit": "f37e7ae3263277319dba4bc39be5397194eb00c3",
- "date": "2026-09-26 14:18 UTC",
+ "date": "2026-09-27 00:34 UTC",
  "reference": "h5py 3.16.0 / HDF5 2.0.0",
  "files": 697,
- "ok": 575,
+ "ok": 600,
  "counts": {
   "h5py-cannot-read": 92,
-  "mismatch": 20,
+  "mismatch": 2,
-  "ok": 575,
+  "ok": 600,
-  "our-error": 10
+  "our-error": 3
  },
  "per_corpus": {
   "NCAS-CMS_pyfive": {
@@ -18,18 +18,17 @@
   },
   "cve_hdf5": {
    "h5py-cannot-read": 32,
-   "mismatch": 9,
+   "ok": 113,
-   "ok": 100,
+   "our-error": 2
-   "our-error": 6
   },
   "h5py_data": {
    "ok": 4
   },
   "hdf5": {
    "h5py-cannot-read": 60,
-   "mismatch": 10,
+   "mismatch": 1,
-   "ok": 392,
+   "ok": 404,
-   "our-error": 4
+   "our-error": 1
   },
   "netcdf-c": {
    "ok": 20
@@ -117,14 +116,22 @@
   "cve_hdf5/cvefiles/cve-2018-17434.h5",
   "cve_hdf5/cvefiles/cve-2018-17435.h5",
   "cve_hdf5/cvefiles/cve-2018-17437.h5",
+  "cve_hdf5/cvefiles/cve-2018-17438",
+  "cve_hdf5/cvefiles/cve-2018-17439",
   "cve_hdf5/cvefiles/cve-2019-8396.h5",
+  "cve_hdf5/cvefiles/cve-2019-8397.h5",
+  "cve_hdf5/cvefiles/cve-2019-8398.h5",
+  "cve_hdf5/cvefiles/cve-2019-9151.h5",
   "cve_hdf5/cvefiles/cve-2019-9152.h5",
   "cve_hdf5/cvefiles/cve-2020-10811.h5",
   "cve_hdf5/cvefiles/cve-2020-18232.h5",
+  "cve_hdf5/cvefiles/cve-2020-18494.h5",
   "cve_hdf5/cvefiles/cve-2021-36977.h5",
   "cve_hdf5/cvefiles/cve-2021-37501.h5",
   "cve_hdf5/cvefiles/cve-2021-45829.h5",
   "cve_hdf5/cvefiles/cve-2021-45833.h5",
+  "cve_hdf5/cvefiles/cve-2021-46243.h5",
+  "cve_hdf5/cvefiles/cve-2021-46244.h5",
   "cve_hdf5/cvefiles/cve-2024-29157.h5",
   "cve_hdf5/cvefiles/cve-2024-29158.h5",
   "cve_hdf5/cvefiles/cve-2024-29159.h5",
@@ -148,18 +155,23 @@
   "cve_hdf5/cvefiles/cve-2024-32615.h5",
   "cve_hdf5/cvefiles/cve-2024-32616.h5",
   "cve_hdf5/cvefiles/cve-2024-32617.h5",
+  "cve_hdf5/cvefiles/cve-2024-32618.h5",
   "cve_hdf5/cvefiles/cve-2024-32619.h5",
   "cve_hdf5/cvefiles/cve-2024-32620.h5",
   "cve_hdf5/cvefiles/cve-2024-32621.h5",
   "cve_hdf5/cvefiles/cve-2024-32622.h5",
+  "cve_hdf5/cvefiles/cve-2024-32623.h5",
   "cve_hdf5/cvefiles/cve-2024-32624.h5",
   "cve_hdf5/cvefiles/cve-2024-33873.h5",
+  "cve_hdf5/cvefiles/cve-2024-33874.h5",
   "cve_hdf5/cvefiles/cve-2024-33875.h5",
   "cve_hdf5/cvefiles/cve-2024-33876.h5",
   "cve_hdf5/cvefiles/cve-2024-33877.h5",
+  "cve_hdf5/cvefiles/cve-2025-2309.h5",
   "cve_hdf5/cvefiles/cve-2025-2310.h5",
   "cve_hdf5/cvefiles/cve-2025-2924.h5",
   "cve_hdf5/cvefiles/cve-2025-2925.h5",
+  "cve_hdf5/cvefiles/cve-2025-44905.h5",
   "cve_hdf5/cvefiles/cve-2025-6269-1.h5",
   "cve_hdf5/cvefiles/cve-2025-6269-2.h5",
   "cve_hdf5/cvefiles/cve-2025-6269-3.h5",
@@ -183,12 +195,14 @@
   "h5py_data/vlen_string_s390x.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_bitgroom.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_blosc.h5",
+  "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_blosc2.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_bshuf.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_bzip2.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_granularbr.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_jpeg.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_lz4.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_lzf.h5",
+  "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_zfp.h5",
   "hdf5/HDF5Examples/C/H5FLT/tfiles/h5ex_d_zstd.h5",
   "hdf5/HDF5Examples/C/H5G/16/h5ex_g_iterate.h5",
   "hdf5/HDF5Examples/C/H5G/16/h5ex_g_traverse.h5",
@@ -264,6 +278,7 @@
   "hdf5/test/testfiles/tmtimeo.h5",
   "hdf5/test/testfiles/tnullspace.h5",
   "hdf5/test/testfiles/tsizeslheap.h5",
+  "hdf5/tools/test/testfiles/bigendian/tall.h5",
   "hdf5/tools/test/testfiles/bigendian/tdset2.h5",
   "hdf5/tools/test/testfiles/binfp64.h5",
   "hdf5/tools/test/testfiles/binin16.h5",
@@ -284,6 +299,7 @@
   "hdf5/tools/test/testfiles/h5clear_fsm_persist_noclose.h5",
   "hdf5/tools/test/testfiles/h5clear_fsm_persist_user_equal.h5",
   "hdf5/tools/test/testfiles/h5clear_fsm_persist_user_less.h5",
+  "hdf5/tools/test/testfiles/h5clear_mdc_image.h5",
   "hdf5/tools/test/testfiles/h5clear_sec2_v0.h5",
   "hdf5/tools/test/testfiles/h5clear_sec2_v2.h5",
   "hdf5/tools/test/testfiles/h5copy_extlinks_src.h5",
@@ -337,6 +353,7 @@
   "hdf5/tools/test/testfiles/h5diff_softlinks.h5",
   "hdf5/tools/test/testfiles/h5diff_strings1.h5",
   "hdf5/tools/test/testfiles/h5diff_strings2.h5",
+  "hdf5/tools/test/testfiles/h5diff_types.h5",
   "hdf5/tools/test/testfiles/h5fc_edge_v3.h5",
   "hdf5/tools/test/testfiles/h5fc_err_level.h5",
   "hdf5/tools/test/testfiles/h5fc_ext1_f.h5",
@@ -414,9 +431,11 @@
   "hdf5/tools/test/testfiles/tCVE_2018_11206_fill_new.h5",
   "hdf5/tools/test/testfiles/tCVE_2018_11206_fill_old.h5",
   "hdf5/tools/test/testfiles/taindices.h5",
+  "hdf5/tools/test/testfiles/tall.h5",
   "hdf5/tools/test/testfiles/tarray1.h5",
   "hdf5/tools/test/testfiles/tarray1_big.h5",
   "hdf5/tools/test/testfiles/tarray2.h5",
+  "hdf5/tools/test/testfiles/tarray3.h5",
   "hdf5/tools/test/testfiles/tarray4.h5",
   "hdf5/tools/test/testfiles/tarray5.h5",
   "hdf5/tools/test/testfiles/tarray8.h5",
@@ -449,6 +468,7 @@
   "hdf5/tools/test/testfiles/textlinksrc.h5",
   "hdf5/tools/test/testfiles/textlinktar.h5",
   "hdf5/tools/test/testfiles/textpfe.h5",
+  "hdf5/tools/test/testfiles/tfcontents1.h5",
   "hdf5/tools/test/testfiles/tfcontents2.h5",
   "hdf5/tools/test/testfiles/tfilters.h5",
   "hdf5/tools/test/testfiles/tfloat16.h5",
@@ -505,6 +525,7 @@
   "hdf5/tools/test/testfiles/tstr3.h5",
   "hdf5/tools/test/testfiles/tudfilter.h5",
   "hdf5/tools/test/testfiles/tudfilter2.h5",
+  "hdf5/tools/test/testfiles/tudlink.h5",
   "hdf5/tools/test/testfiles/tvldtypes1.h5",
   "hdf5/tools/test/testfiles/tvldtypes2.h5",
   "hdf5/tools/test/testfiles/tvldtypes3.h5",
@@ -513,6 +534,8 @@
   "hdf5/tools/test/testfiles/tvlenstr_array.h5",
   "hdf5/tools/test/testfiles/tvlstr.h5",
   "hdf5/tools/test/testfiles/tvms.h5",
+  "hdf5/tools/test/testfiles/twithub.h5",
+  "hdf5/tools/test/testfiles/twithub513.h5",
   "hdf5/tools/test/testfiles/txtfp32.h5",
   "hdf5/tools/test/testfiles/txtfp64.h5",
   "hdf5/tools/test/testfiles/txtin16.h5",
@@ -558,6 +581,7 @@
   "hdf5/tools/test/testfiles/xml/tenum.h5",
   "hdf5/tools/test/testfiles/xml/test35.nc",
   "hdf5/tools/test/testfiles/xml/tloop2.h5",
+  "hdf5/tools/test/testfiles/xml/tmany.h5",
   "hdf5/tools/test/testfiles/xml/tname-amp.h5",
   "hdf5/tools/test/testfiles/xml/tname-apos.h5",
   "hdf5/tools/test/testfiles/xml/tname-gt.h5",

conformance/compare.py

@@ -146,7 +146,13 @@ for rel in files:
             if a.get("kind") != b.get("kind") and "error" not in b and "error" not in a:
                 issues.append(("mismatch", f"{p}: kind {a.get('kind')} vs ours {b.get('kind')}", "kind", b))
                 ok = False
+            # h5py could not open the object at all: it read none of its
+            # attributes or links, so there is nothing to compare ours with
+            # (the object's own error is compared above and below).
+            ref_unopened = a.get("kind") == "unknown" and "error" in a
             for k in ("error", "list_error", "attrs_error"):
+                if ref_unopened and k != "error":
+                    continue
                 if k in b and k not in a:
                     issues.append(("our-error", f"{p}: {k}: {b[k]}", b[k], b))
                     ok = False
@@ -164,7 +170,7 @@ for rel in files:
                     issues.append(("mismatch", f"{p}: values differ (h5py {a.get('dtype')} vs ours {b.get('dtype')})", "values", b | {"ref_head": a.get("head"), "ref_dtype": a.get("dtype")}))
                     ok = False
             ra, oa = a.get("attrs") or {}, b.get("attrs") or {}
-            if "attrs_error" not in b and "attrs_error" not in a:
+            if "attrs_error" not in b and "attrs_error" not in a and not ref_unopened:
                 for an in sorted(set(ra) | set(oa)):
                     x, y = ra.get(an), oa.get(an)
                     if x is None:

conformance/probe/src/main.rs

@@ -31,7 +31,7 @@ use clawhdf5_format::filter_pipeline::FilterPipeline;
 use clawhdf5_format::group_v1::{self, GroupEntry};
 use clawhdf5_format::group_v2;
 use clawhdf5_format::message_type::MessageType;
-use clawhdf5_format::object_header::ObjectHeader;
+use clawhdf5_format::object_header::{ObjectClass, ObjectHeader};
 use clawhdf5_format::signature;
 use clawhdf5_format::superblock::Superblock;
 use clawhdf5_format::symbol_table::SymbolTableMessage;
@@ -308,18 +308,20 @@ impl<'a> Ctx<'a> {
             )
             .map_err(e)?;
         }
-        let (shape, n) = Self::shape(&ds);
-        rec.insert("shape".into(), shape);
-        if n.saturating_mul(dt.type_size() as u64) > MAX_BYTES {
-            rec.insert("skipped".into(), Value::String("too large".into()));
-            return Ok(());
-        }
         let lm = h
             .messages
             .iter()
             .find(|m| m.msg_type == MessageType::DataLayout)
             .ok_or("MissingMessage(DataLayout)")?;
         let dl = DataLayout::parse(&lm.data, self.os, self.ls).map_err(e)?;
+        // What libhdf5 checks when it opens the dataset (as File::dataset).
+        data_read::check_dataset_storage(&dl, &ds, &dt, self.data.len() as u64).map_err(e)?;
+        let (shape, n) = Self::shape(&ds);
+        rec.insert("shape".into(), shape);
+        if n.saturating_mul(dt.type_size() as u64) > MAX_BYTES {
+            rec.insert("skipped".into(), Value::String("too large".into()));
+            return Ok(());
+        }
         rec.insert(
             "layout".into(),
             Value::String(
@@ -657,6 +659,20 @@ fn is_group(h: &ObjectHeader) -> bool {
     })
 }
 
+/// The probe's kind for an object header: libhdf5's object class
+/// ([`ObjectHeader::object_class`]: group, then dataset — a datatype *and* a
+/// dataspace — then named datatype), which is what h5py opens the object as.
+/// The root group, and a header with only link messages, count as groups.
+fn kind_of(h: &ObjectHeader, is_root: bool) -> &'static str {
+    match h.object_class() {
+        Some(ObjectClass::Group) => "group",
+        Some(ObjectClass::Dataset) => "dataset",
+        _ if is_root || is_group(h) => "group",
+        Some(ObjectClass::NamedDatatype) => "datatype",
+        None => "unknown",
+    }
+}
+
 fn main() {
     install_hook();
     let path = std::env::args().nth(1).expect("usage: probe <file>");
@@ -696,6 +712,42 @@ fn main() {
             return;
         }
     };
+    // libhdf5 decodes the superblock extension at open (an error refuses
+    // the file), and loads a metadata cache image over the file's own
+    // metadata. It loads the image only when it first reads metadata — the
+    // root group — so a file whose image it cannot load still opens and
+    // that read fails. The library decides all three cases with the same
+    // `cache_image_state`: `File` and `MmapFile` open such a file and fail
+    // every object lookup with the image's error, which is what the probe
+    // records here (on the root group, where libhdf5 reports it).
+    use clawhdf5_format::superblock_ext::{self, CacheImageState};
+    let state = match guarded(|| superblock_ext::cache_image_state(hdf5, &sb).map_err(e)) {
+        Ok(x) => x,
+        Err(msg) => {
+            top.insert("open_error".into(), Value::String(msg));
+            println!("{}", Value::Object(top));
+            return;
+        }
+    };
+    let mut image_error = None;
+    let view = match state {
+        CacheImageState::Absent => None,
+        CacheImageState::Unloadable(err) => {
+            image_error = Some(e(err));
+            None
+        }
+        CacheImageState::Loaded(image) => {
+            let mut v = hdf5.to_vec();
+            match image.block(hdf5).and_then(|b| image.apply(b, &mut v)) {
+                Ok(()) => Some(v),
+                Err(err) => {
+                    image_error = Some(e(err));
+                    None
+                }
+            }
+        }
+    };
+    let hdf5: &[u8] = view.as_deref().unwrap_or(hdf5);
     top.insert("superblock_version".into(), json!(sb.version));
     let ctx = Ctx {
         data: hdf5,
@@ -723,6 +775,9 @@ fn main() {
         let mut rec = Map::new();
         rec.insert("path".into(), Value::String(p.clone()));
         let r = guarded(|| {
+            if let Some(msg) = &image_error {
+                return Err(msg.clone());
+            }
             let h = ctx.header(addr)?;
             Ok(h)
         });
@@ -735,23 +790,7 @@ fn main() {
                 continue;
             }
         };
-        let is_ds = h
+        let kind = kind_of(&h, addr == sb.root_group_address);
-            .messages
-            .iter()
-            .any(|m| m.msg_type == MessageType::DataLayout);
-        let kind = if is_ds {
-            "dataset"
-        } else if is_group(&h) || addr == sb.root_group_address {
-            "group"
-        } else if h
-            .messages
-            .iter()
-            .any(|m| m.msg_type == MessageType::Datatype)
-        {
-            "datatype"
-        } else {
-            "unknown"
-        };
         rec.insert("kind".into(), Value::String(kind.into()));
         if kind == "dataset"
             && let Err(msg) = guarded(|| ctx.read_dataset(&h, &mut rec))
@@ -867,6 +906,42 @@ mod tests {
         assert!(ieee_layout(&f32le));
     }
 
+    #[test]
+    fn kind_follows_libhdf5_object_class() {
+        use clawhdf5_format::object_header::HeaderMessage;
+        let header = |types: &[MessageType]| ObjectHeader {
+            version: 2,
+            messages: types
+                .iter()
+                .map(|&msg_type| HeaderMessage {
+                    msg_type,
+                    size: 0,
+                    flags: 0,
+                    creation_order: None,
+                    data: Vec::new(),
+                })
+                .collect(),
+            reference_count: None,
+            flags: 0,
+            access_time: None,
+            modification_time: None,
+            change_time: None,
+            birth_time: None,
+        };
+        use MessageType::*;
+        // cve-2024-33874 `/Dset1`: a datatype and a layout but no dataspace
+        // is a named datatype to libhdf5 (h5py opens it as one).
+        assert_eq!(kind_of(&header(&[Datatype, DataLayout]), false), "datatype");
+        assert_eq!(
+            kind_of(&header(&[Datatype, Dataspace, DataLayout]), false),
+            "dataset"
+        );
+        assert_eq!(kind_of(&header(&[SymbolTable]), false), "group");
+        assert_eq!(kind_of(&header(&[Link]), false), "group");
+        assert_eq!(kind_of(&header(&[]), true), "group");
+        assert_eq!(kind_of(&header(&[]), false), "unknown");
+    }
+
     #[test]
     fn partial_precision_int_is_shifted_and_sign_extended() {
         let dt = Datatype::FixedPoint {

conformance/ref.py

@@ -111,8 +111,11 @@ def note_conversion(tid, dt, rec):
 
 
 def hash_values(arr, dt, rec):
-    if dt.subdtype is not None:
+    # h5py expands an HDF5 array element type into trailing array dims, a
-        # h5py expands an HDF5 array element type into trailing array dims
+    # nested array type (an array of arrays) into all of them. Converting the
+    # expanded array back to the inner subarray type would broadcast every
+    # element into a whole subarray, so strip every level.
+    while dt.subdtype is not None:
         dt = dt.subdtype[0]
     arr = np.asarray(arr, dtype=dt)
     if simple(dt):
@@ -173,9 +176,13 @@ def main(path):
         return
     objects = []
     seen = set()
-    stack = [("/", None)]
+    # Objects h5py cannot open have no ObjectID to deduplicate by; they are
+    # deduplicated by the address their hard link points at instead, as the
+    # probe deduplicates every object by header address.
+    seen_unopenable = set()
+    stack = [("/", None, None)]
     while stack:
-        p, obj = stack.pop()
+        p, obj, link_addr = stack.pop()
         if len(objects) >= MAX_OBJECTS:
             top["truncated"] = True
             break
@@ -185,6 +192,10 @@ def main(path):
                 obj = f[p]
             key = hash(obj.id)  # h5py ObjectID hash = (fileno, object address/token)
         except Exception as e:  # noqa: BLE001
+            if link_addr is not None:
+                if link_addr in seen_unopenable:
+                    continue
+                seen_unopenable.add(link_addr)
             rec["kind"] = "unknown"
             rec["error"] = err(e)
             objects.append(rec)
@@ -232,15 +243,18 @@ def main(path):
                 base = "" if p == "/" else p
                 kids = []
                 for n in names:
+                    # The link's own type: `obj.get(n, getlink=True)` reports
+                    # a user-defined link (type 64-255) as a HardLink.
                     try:
-                        link = obj.get(n, getlink=True)
+                        info = obj.id.links.get_info(n.encode("utf-8", "surrogateescape"))
                     except Exception:  # noqa: BLE001
-                        link = None
+                        info = None
-                    if link is not None and not isinstance(link, h5py.HardLink):
+                    if info is not None and info.type != h5py.h5l.TYPE_HARD:
                         continue
-                    kids.append(f"{base}/{n}")
+                    addr = info.u if info is not None else None
-                for k in reversed(kids):
+                    kids.append((f"{base}/{n}", addr))
-                    stack.append((k, None))
+                for k, addr in reversed(kids):
+                    stack.append((k, None, addr))
             except Exception as e:  # noqa: BLE001
                 rec["list_error"] = err(e)
         objects.append(rec)

conformance/report.py

@@ -98,13 +98,37 @@ def is_h5py_be_vlen(i):
             and ">" in (i.get("ours_dtype") or ""))
 
 
+# Objects the reference (h5py 3.16 / HDF5 2.0) reads only because of an
+# HDF5 2.0 bug, and that clawhdf5 refuses: each one reads past a buffer or
+# returns bytes the file does not hold, and libhdf5's develop branch refuses all
+# three. (file, object) -> why. Checked 2026-09-26 against HDF5 2.0.0
+# and HDFGroup/hdf5 develop sources; see docs/known-issues.md.
+LIBHDF5_BUGS = {
+    ("cve_hdf5/cvefiles/cve-2025-2308.h5", "/Scale_offset_long_long_data_le"):
+        "scale-offset codes run past the end of the chunk: HDF5 2.0 reads past its buffer; "
+        "libhdf5's develop branch refuses the chunk (\"Buffer too short\")",
+    ("cve_hdf5/cvefiles/cve-2025-44904.h5", "/Scale_offset_float_data_le"):
+        "unfiltered chunks of 38 and 37 bytes for 48-byte chunks: HDF5 2.0 fills the rest with "
+        "whatever its buffer held; libhdf5's develop branch refuses them (\"incorrect chunk size returned "
+        "from index for unfiltered chunk\")",
+    ("hdf5/test/testfiles/bad_nbit_parms_walk.h5", "/Nbit_int_data_le"):
+        "an N-Bit parameter list one value short: HDF5 2.0 reads past the list; libhdf5's own "
+        "test (`test_filter_bad_params`, test/dsets.c) now requires the read to fail",
+}
+
+
+def is_libhdf5_bug(rel, i):
+    return i["kind"] == "our-error" and any(
+        f == rel and i["detail"].startswith(obj + ":") for (f, obj) in LIBHDF5_BUGS)
+
+
 known = collections.defaultdict(list)
 for r in rows:
-    if r["class"] != "mismatch":
-        continue
     iss = issues.get(r["file"], [])
-    if iss and all(is_h5py_be_vlen(i) for i in iss):
+    if r["class"] == "mismatch" and iss and all(is_h5py_be_vlen(i) for i in iss):
         known["h5py-be-vlen"].append(r["file"])
+    if r["class"] == "our-error" and iss and all(is_libhdf5_bug(r["file"], i) for i in iss):
+        known["libhdf5-2.0"].append(r["file"])
 
 
 # --- the CVE corpus: clawhdf5 vs h5dump vs h5py ------------------------------
@@ -230,9 +254,13 @@ for c in sorted(by_corpus):
     w(f"| {c} | {sum(cnt.values())} | " + " | ".join(str(cnt.get(k, 0)) for k in CLASSES) + " |")
 w(f"| **all** | **{len(rows)}** | " + " | ".join(f"**{total.get(k, 0)}**" for k in CLASSES) + " |")
 w("")
-n_known = sum(len(v) for v in known.values())
+if known["h5py-be-vlen"]:
-if n_known:
+    w(f"{len(known['h5py-be-vlen'])} of the {total.get('mismatch', 0)} mismatches are a known h5py bug, "
-    w(f"{n_known} of the {total.get('mismatch', 0)} mismatches are a known h5py bug, not ours (see *Known not-our-bug*).")
+      "not ours (see *Known not-our-bug*).")
+    w("")
+if known["libhdf5-2.0"]:
+    w(f"{len(known['libhdf5-2.0'])} of the {total.get('our-error', 0)} our-errors are corrupt data that "
+      "HDF5 2.0 reads only through a bug and clawhdf5 refuses (see *Known not-our-bug*).")
     w("")
 w("Corpora (fetched by `conformance/fetch-corpus.sh` into the gitignored `conformance/.cache/`):")
 w("")
@@ -311,6 +339,11 @@ if res["incomparable"]:
     w("  (FP8 -> float16, bfloat16 -> float32, x87 long double -> float128) the values are not")
     w("  compared (shape and presence still are): "
       + ", ".join(f"{k} ({n}x)" for k, n in res["incomparable"]) + ".")
+w("- **Corrupt data HDF5 2.0 reads through a bug.** clawhdf5 refuses these objects; h5py 3.16 /")
+w("  HDF5 2.0 returns values for them that the file does not hold:")
+for (f, obj), why in sorted(LIBHDF5_BUGS.items()):
+    here = "" if f in known["libhdf5-2.0"] else " (not an our-error in this run)"
+    w(f"  - `{f}` `{obj}`: {why}{here}.")
 w("- **References** are compared by presence only (`R`), not by target.")
 w("")
 if res.get("ref_only_errors"):

crates/clawhdf5-accel/src/lib.rs

@@ -237,7 +237,10 @@ pub fn f16_to_f32_batch(input: &[u16], output: &mut [f32]) {
     convert::f16_to_f32_batch(input, output);
 }
 
-/// Compute Fletcher-32 checksum.
+/// Compute a textbook Fletcher-32 checksum (both sums start at 0xffff).
+///
+/// This is not HDF5's checksum; the Fletcher-32 I/O filter uses
+/// `clawhdf5_format::checksum::fletcher32`.
 pub fn checksum_fletcher32(data: &[u8]) -> u32 {
     checksum::checksum_fletcher32(data)
 }

crates/clawhdf5-format/Cargo.toml

@@ -76,8 +76,16 @@ bitshuffle = ["lz4_flex", "ruzstd"]
 bzip2 = ["dep:bzip2", "std"]
 # Blosc 1 (32001) with its BloscLZ, LZ4, Snappy, Zlib and Zstandard codecs.
 blosc = ["lz4_flex", "ruzstd", "snap", "deflate", "std"]
+# Blosc2 (32026), read-only: frames, B2ND arrays, and the Blosc codecs above.
+blosc2 = ["blosc"]
+# ZFP (32013, H5Z-ZFP), read-only: every mode, for int32, int64, float and
+# double fields of 1 to 4 dimensions.
+zfp = []
 # Every plugin filter above.
-plugin-filters = ["lzf", "bitshuffle", "bzip2", "blosc"]
+plugin-filters = ["lzf", "bitshuffle", "bzip2", "blosc", "blosc2", "zfp"]
+# Test instrumentation: per-thread counts of heap objects read (see
+# `lookup_stats`), so tests can bound the cost of a name lookup.
+lookup-stats = ["std"]
 
 [[bench]]
 name = "parallel_decompress_bench"

crates/clawhdf5-format/src/addr.rs

@@ -0,0 +1,121 @@
+//! File address and length → in-memory index conversion.
+//!
+//! HDF5 addresses and lengths are 64-bit; the file is parsed through a
+//! `&[u8]` indexed by `usize`. On a 64-bit target every `u64` fits, but on a
+//! 32-bit one (`wasm32`, `i686`, `thumbv7em`) an address past `usize::MAX`
+//! used to be truncated by an `as usize` cast — silently pointing at another
+//! part of the file — or to panic. [`to_usize`] is the one conversion the
+//! parsers use instead: such an address is a clean
+//! [`FormatError::Overflow`]. It cannot be inside the data anyway: no slice
+//! is longer than `isize::MAX` bytes.
+
+#[cfg(not(feature = "std"))]
+use alloc::format;
+
+use crate::error::FormatError;
+
+/// A file address, offset or length from the file as a `usize` index.
+///
+/// Fails with [`FormatError::Overflow`] when the value does not fit this
+/// platform's `usize` (only possible on targets narrower than 64 bits).
+#[inline]
+pub fn to_usize(value: u64) -> Result<usize, FormatError> {
+    to_index::<usize>(value)
+}
+
+/// A file address for a [`crate::storage::Storage`] read, checked as
+/// [`to_usize`] checks it: the parsers read through 64-bit offsets, but an
+/// address that could not index an in-memory file on this platform is the
+/// same [`FormatError::Overflow`] the slice parsers gave for it.
+#[inline]
+pub fn checked_addr(value: u64) -> Result<u64, FormatError> {
+    to_usize(value).map(|_| value)
+}
+
+/// [`to_usize`] for an index type of any width. `usize` is 64 bits wide on
+/// the hosts CI tests on, where the error path cannot be reached through
+/// `usize`; tests run the same code with `u32` in its place, as on a 32-bit
+/// target.
+#[inline]
+fn to_index<T: TryFrom<u64>>(value: u64) -> Result<T, FormatError> {
+    T::try_from(value).map_err(|_| too_large(value))
+}
+
+/// A count or offset into an in-memory buffer (a codec's progress counter,
+/// a size the writer computed from data it holds) as a `usize`, saturating
+/// at `usize::MAX` instead of truncating.
+///
+/// For values that are bounded by the length of something in memory, so
+/// always fit; if one ever did not, a saturated index fails its bounds check
+/// or allocation instead of silently addressing the wrong bytes. A value
+/// read from the file uses [`to_usize`].
+#[inline]
+pub fn saturating_usize(value: u64) -> usize {
+    saturating_index(value, usize::MAX)
+}
+
+/// [`saturating_usize`] for an index type of any width, whose largest
+/// value is `max` (see [`to_index`]).
+#[inline]
+fn saturating_index<T: TryFrom<u64>>(value: u64, max: T) -> T {
+    T::try_from(value).unwrap_or(max)
+}
+
+#[cold]
+#[inline(never)]
+fn too_large(value: u64) -> FormatError {
+    FormatError::Overflow(format!(
+        "file address or length {value:#x} exceeds this platform's address space"
+    ))
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn values_that_fit_convert_exactly() {
+        assert_eq!(to_usize(0), Ok(0));
+        assert_eq!(to_usize(0x1234), Ok(0x1234));
+        assert_eq!(to_usize(usize::MAX as u64), Ok(usize::MAX));
+    }
+
+    #[test]
+    fn saturating_conversion_never_wraps() {
+        assert_eq!(saturating_usize(0), 0);
+        assert_eq!(saturating_usize(0x1234), 0x1234);
+        assert_eq!(saturating_usize(usize::MAX as u64), usize::MAX);
+        // Past usize::MAX (32-bit targets) or at u64::MAX: saturates.
+        assert_eq!(saturating_usize(u64::MAX), usize::MAX);
+    }
+
+    #[test]
+    fn values_past_usize_max_are_an_error_not_truncated() {
+        // Reachable through `usize` only where it is narrower than u64 (no
+        // such target runs tests in CI), so the same conversion is run with
+        // u32 standing in for a 32-bit usize.
+        let max = u64::from(u32::MAX);
+        assert_eq!(to_index::<u32>(max), Ok(u32::MAX));
+        for past in [max + 1, max + 0x10, 0x1_0000_1234, u64::MAX] {
+            let err = to_index::<u32>(past).unwrap_err();
+            assert!(
+                matches!(err, FormatError::Overflow(_)),
+                "{past:#x}: {err:?}"
+            );
+        }
+        // Where an `as` cast would have wrapped to a small, valid-looking
+        // index, it is not returned.
+        assert_eq!(0x1_0000_1234_u64 as u32, 0x1234);
+        assert!(to_index::<u32>(0x1_0000_1234).is_err());
+
+        assert_eq!(saturating_index(max + 1, u32::MAX), u32::MAX);
+        assert_eq!(saturating_index(0x1_0000_1234, u32::MAX), u32::MAX);
+        assert_eq!(saturating_index(0x1234, u32::MAX), 0x1234);
+
+        // And through `usize` itself, whichever width it has here.
+        match (usize::MAX as u64).checked_add(1) {
+            Some(past) => assert!(matches!(to_usize(past), Err(FormatError::Overflow(_)))),
+            None => assert_eq!(to_usize(u64::MAX), Ok(usize::MAX)),
+        }
+    }
+}

crates/clawhdf5-format/src/attribute.rs

@@ -5,8 +5,10 @@ use alloc::{borrow::Cow, string::String, vec::Vec};
 #[cfg(feature = "std")]
 use std::borrow::Cow;
 
+use crate::addr::to_usize;
 use crate::attribute_info::AttributeInfoMessage;
-use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records};
+use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records_in, find_btree_v2_records_in};
+use crate::checksum::jenkins_lookup3;
 use crate::data_read;
 use crate::dataspace::Dataspace;
 use crate::datatype::Datatype;
@@ -15,6 +17,7 @@ use crate::fractal_heap::FractalHeapHeader;
 use crate::message_type::MessageType;
 use crate::object_header::ObjectHeader;
 use crate::shared_message;
+use crate::storage::Storage;
 use crate::vl_data;
 
 /// A parsed HDF5 attribute message.
@@ -50,7 +53,7 @@ impl AttributeMessage {
     ///
     /// `length_size` is needed for dataspace dimension parsing.
     pub fn parse(data: &[u8], length_size: u8) -> Result<AttributeMessage, FormatError> {
-        Self::parse_impl(data, length_size, None)
+        Self::parse_impl(data, length_size, None::<(&[u8], u8)>)
     }
 
     /// [`AttributeMessage::parse`] with access to the rest of the file, which
@@ -65,13 +68,24 @@ impl AttributeMessage {
         offset_size: u8,
         length_size: u8,
     ) -> Result<AttributeMessage, FormatError> {
-        Self::parse_impl(data, length_size, Some((file_data, offset_size)))
+        Self::parse_in_storage(data, file_data, offset_size, length_size)
     }
 
-    fn parse_impl(
+    /// [`AttributeMessage::parse_in_file`] with the file behind any
+    /// [`Storage`].
+    pub fn parse_in_storage<S: Storage + ?Sized>(
+        data: &[u8],
+        file: &S,
+        offset_size: u8,
+        length_size: u8,
+    ) -> Result<AttributeMessage, FormatError> {
+        Self::parse_impl(data, length_size, Some((file, offset_size)))
+    }
+
+    fn parse_impl<S: Storage + ?Sized>(
         data: &[u8],
         length_size: u8,
-        file: Option<(&[u8], u8)>,
+        file: Option<(&S, u8)>,
     ) -> Result<AttributeMessage, FormatError> {
         ensure_len(data, 0, 2)?;
         let version = data[0];
@@ -86,19 +100,19 @@ impl AttributeMessage {
 
     /// The bytes of an embedded datatype/dataspace message, following the
     /// shared-message reference when `shared` is set.
-    fn embedded_message<'a>(
+    fn embedded_message<'a, S: Storage + ?Sized>(
         bytes: &'a [u8],
         shared: bool,
         msg_type: MessageType,
         length_size: u8,
-        file: Option<(&[u8], u8)>,
+        file: Option<(&S, u8)>,
     ) -> Result<Cow<'a, [u8]>, FormatError> {
         if !shared {
             return Ok(Cow::Borrowed(bytes));
         }
         let (file_data, offset_size) = file.ok_or(FormatError::UnresolvedSharedMessage)?;
         let shared_ref = shared_message::parse_shared_ref_sized(bytes, offset_size, length_size)?;
-        shared_message::resolve_shared_message(
+        shared_message::resolve_shared_message_in(
             file_data,
             &shared_ref,
             msg_type,
@@ -143,10 +157,10 @@ impl AttributeMessage {
         })
     }
 
-    fn parse_v2(
+    fn parse_v2<S: Storage + ?Sized>(
         data: &[u8],
         length_size: u8,
-        file: Option<(&[u8], u8)>,
+        file: Option<(&S, u8)>,
     ) -> Result<AttributeMessage, FormatError> {
         // Flags: bit 0 = datatype is shared, bit 1 = dataspace is shared.
         let flags = data.get(1).copied().unwrap_or(0);
@@ -197,10 +211,10 @@ impl AttributeMessage {
         })
     }
 
-    fn parse_v3(
+    fn parse_v3<S: Storage + ?Sized>(
         data: &[u8],
         length_size: u8,
-        file: Option<(&[u8], u8)>,
+        file: Option<(&S, u8)>,
     ) -> Result<AttributeMessage, FormatError> {
         // Flags: bit 0 = datatype is shared, bit 1 = dataspace is shared.
         let flags = data.get(1).copied().unwrap_or(0);
@@ -322,9 +336,19 @@ impl AttributeMessage {
         file_data: &[u8],
         offset_size: u8,
         length_size: u8,
+    ) -> Result<Vec<String>, FormatError> {
+        self.read_vl_strings_in(file_data, offset_size, length_size)
+    }
+
+    /// [`Self::read_vl_strings`] over any [`Storage`].
+    pub fn read_vl_strings_in<S: Storage + ?Sized>(
+        &self,
+        file_data: &S,
+        offset_size: u8,
+        length_size: u8,
     ) -> Result<Vec<String>, FormatError> {
         let num_elements = self.dataspace.num_elements();
-        vl_data::read_vl_strings(
+        vl_data::read_vl_strings_in(
             file_data,
             &self.raw_data,
             num_elements,
@@ -341,7 +365,8 @@ fn compute_raw_data(
     dataspace: &Dataspace,
     datatype: &Datatype,
 ) -> Vec<u8> {
-    let num_elements = dataspace.num_elements() as usize;
+    // Saturating, like the product: the size is capped at what is there.
+    let num_elements = usize::try_from(dataspace.num_elements()).unwrap_or(usize::MAX);
     let elem_size = datatype.type_size() as usize;
     let expected_size = num_elements.saturating_mul(elem_size);
     let available = data.len().saturating_sub(pos);
@@ -415,7 +440,20 @@ pub fn extract_attributes_full(
     offset_size: u8,
     length_size: u8,
 ) -> Result<Vec<AttributeMessage>, FormatError> {
-    extract_attributes_with(file_data, header, offset_size, length_size, &mut Err)
+    extract_attributes_full_in(file_data, header, offset_size, length_size)
+}
+
+/// [`extract_attributes_full`] over any [`Storage`]. Dense attribute
+/// storage is indexed by a v2 B-tree, which is not read over [`Storage`]
+/// yet: on a backend without the whole file in memory an object with dense
+/// attributes is [`FormatError::ContiguousStorageRequired`].
+pub fn extract_attributes_full_in<S: Storage + ?Sized>(
+    file: &S,
+    header: &ObjectHeader,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<AttributeMessage>, FormatError> {
+    extract_attributes_with(file, header, offset_size, length_size, &mut Err)
 }
 
 /// Like [`extract_attributes_full`], but an attribute that cannot be read
@@ -431,6 +469,32 @@ pub fn extract_attributes_tolerant(
     header: &ObjectHeader,
     offset_size: u8,
     length_size: u8,
+) -> Result<(Vec<AttributeMessage>, Vec<FormatError>), FormatError> {
+    extract_attributes_tolerant_core(file_data, header, offset_size, length_size)
+}
+
+/// [`extract_attributes_tolerant`] over any [`Storage`] (see
+/// [`extract_attributes_full_in`] for dense storage). One with the whole
+/// file in memory is read as the slice, by code compiled in this crate (see
+/// [`crate::storage`], "Slice entry points").
+#[inline]
+pub fn extract_attributes_tolerant_in<S: Storage + ?Sized>(
+    file_data: &S,
+    header: &ObjectHeader,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<(Vec<AttributeMessage>, Vec<FormatError>), FormatError> {
+    match file_data.as_contiguous() {
+        Some(all) => extract_attributes_tolerant(all, header, offset_size, length_size),
+        None => extract_attributes_tolerant_core(file_data, header, offset_size, length_size),
+    }
+}
+
+fn extract_attributes_tolerant_core<S: Storage + ?Sized>(
+    file_data: &S,
+    header: &ObjectHeader,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<(Vec<AttributeMessage>, Vec<FormatError>), FormatError> {
     let mut errors = Vec::new();
     let attrs = extract_attributes_with(file_data, header, offset_size, length_size, &mut |e| {
@@ -442,23 +506,198 @@ pub fn extract_attributes_tolerant(
 
 /// Read every attribute; each one that fails goes to `on_error`, which
 /// either stops the read (returns the error) or skips that attribute.
-fn extract_attributes_with(
+fn extract_attributes_with<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     header: &ObjectHeader,
     offset_size: u8,
     length_size: u8,
     on_error: &mut dyn FnMut(FormatError) -> Result<(), FormatError>,
 ) -> Result<Vec<AttributeMessage>, FormatError> {
     let mut attrs = Vec::new();
+    // Each attribute's creation order, where the file records one.
+    let mut orders: Vec<u32> = Vec::new();
 
-    // Collect compact attributes (inline in OH)
+    extract_compact_attributes(
+        file_data,
+        header,
+        offset_size,
+        length_size,
+        &mut attrs,
+        &mut orders,
+        on_error,
+    )?;
+
+    // Check for dense attributes via AttributeInfo message
+    let attr_info = find_attribute_info(header, offset_size)?;
+    if let Some(info) = &attr_info
+        && let Some(fh_addr) = info.fractal_heap_address
+    {
+        extract_dense_attributes(
+            file_data,
+            info,
+            fh_addr,
+            offset_size,
+            length_size,
+            &mut attrs,
+            &mut orders,
+            on_error,
+        )?;
+    }
+
+    // An object that tracks attribute creation order lists its attributes
+    // in that order (h5py's `track_order=True`), as libhdf5 does; otherwise
+    // they come in storage order.
+    if attr_info.is_some_and(|i| i.max_creation_index.is_some()) {
+        let mut paired: Vec<(u32, AttributeMessage)> = orders.into_iter().zip(attrs).collect();
+        paired.sort_by_key(|(o, _)| *o);
+        attrs = paired.into_iter().map(|(_, a)| a).collect();
+    }
+
+    Ok(attrs)
+}
+
+/// B-tree v2 record type of dense attribute storage's name index.
+const ATTRIBUTE_NAME_INDEX: u8 = 8;
+
+/// The attribute called `name` on the object with header `header`: the
+/// first one [`extract_attributes_tolerant`] returns under that name, or
+/// `None` if it returns none (an attribute that cannot be read is not
+/// returned there either).
+///
+/// Compact attributes are in the header and are scanned. Dense attributes
+/// are found through the name index (a v2 B-tree of lookup3 name hashes,
+/// record type 8): only the attributes whose names hash like `name` are read
+/// from the heap, O(log n) instead of all of them. Errors in the structures
+/// that index the attributes fail the call, as they fail a listing.
+pub fn find_attribute_in_file(
+    file_data: &[u8],
+    header: &ObjectHeader,
+    name: &str,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Option<AttributeMessage>, FormatError> {
+    find_attribute_core(file_data, header, name, offset_size, length_size)
+}
+
+/// [`find_attribute_in_file`] over any [`Storage`] (see
+/// [`extract_attributes_full_in`] for dense storage, whose name index still
+/// needs the whole file in memory). One with the whole file in memory is
+/// read as the slice, by code compiled in this crate (see
+/// [`crate::storage`], "Slice entry points").
+#[inline]
+pub fn find_attribute_in<S: Storage + ?Sized>(
+    file_data: &S,
+    header: &ObjectHeader,
+    name: &str,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Option<AttributeMessage>, FormatError> {
+    match file_data.as_contiguous() {
+        Some(all) => find_attribute_in_file(all, header, name, offset_size, length_size),
+        None => find_attribute_core(file_data, header, name, offset_size, length_size),
+    }
+}
+
+fn find_attribute_core<S: Storage + ?Sized>(
+    file_data: &S,
+    header: &ObjectHeader,
+    name: &str,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Option<AttributeMessage>, FormatError> {
+    let attr_info = find_attribute_info(header, offset_size)?;
+    let dense = attr_info
+        .as_ref()
+        .and_then(|i| Some((i.fractal_heap_address?, i.btree_name_index_address?)));
+    let Some((fh_addr, btree_addr)) = dense else {
+        // Compact only (or dense storage without a name index, which a
+        // listing reports): as a listing finds it.
+        return Ok(
+            extract_attributes_tolerant_in(file_data, header, offset_size, length_size)?
+                .0
+                .into_iter()
+                .find(|a| a.name == name),
+        );
+    };
+    let btree_hdr = BTreeV2Header::parse_in(
+        file_data,
+        to_usize(btree_addr)? as u64,
+        offset_size,
+        length_size,
+    )?;
+    let fh = FractalHeapHeader::parse_in(file_data, fh_addr, offset_size, length_size)?;
+    if btree_hdr.tree_type != ATTRIBUTE_NAME_INDEX || btree_hdr.record_size < 4 {
+        return Ok(
+            extract_attributes_tolerant_in(file_data, header, offset_size, length_size)?
+                .0
+                .into_iter()
+                .find(|a| a.name == name),
+        );
+    }
+
+    // A listing has the compact attributes first.
+    let mut compact = Vec::new();
+    extract_compact_attributes(
+        file_data,
+        header,
+        offset_size,
+        length_size,
+        &mut compact,
+        &mut Vec::new(),
+        &mut |_| Ok(()),
+    )?;
+    if let Some(a) = compact.into_iter().find(|a| a.name == name) {
+        return Ok(Some(a));
+    }
+
+    // Record: heap ID + message flags(1) + creation order(4) + hash(4); the
+    // hash is the last field.
+    let hash = jenkins_lookup3(name.as_bytes());
+    let hash_at = usize::from(btree_hdr.record_size) - 4;
+    let records = find_btree_v2_records_in(file_data, &btree_hdr, offset_size, &mut |r| match r
+        .get(hash_at..hash_at + 4)
+    {
+        Some(h) => u32::from_le_bytes([h[0], h[1], h[2], h[3]]).cmp(&hash),
+        None => core::cmp::Ordering::Less,
+    })?;
+    let id_len = usize::from(fh.heap_id_length);
+    for record in &records {
+        let Some(id_bytes) = record.data.get(..id_len) else {
+            continue;
+        };
+        let attr = fh
+            .read_managed_object_in(file_data, id_bytes, offset_size)
+            .and_then(|d| {
+                AttributeMessage::parse_in_storage(&d, file_data, offset_size, length_size)
+            });
+        // One that cannot be read is left out, as from a listing.
+        if let Ok(attr) = attr
+            && attr.name == name
+        {
+            return Ok(Some(attr));
+        }
+    }
+    Ok(None)
+}
+
+/// The attributes stored in the object header itself (compact storage), and
+/// each one's creation order into `orders`.
+fn extract_compact_attributes<S: Storage + ?Sized>(
+    file_data: &S,
+    header: &ObjectHeader,
+    offset_size: u8,
+    length_size: u8,
+    attrs: &mut Vec<AttributeMessage>,
+    orders: &mut Vec<u32>,
+    on_error: &mut dyn FnMut(FormatError) -> Result<(), FormatError>,
+) -> Result<(), FormatError> {
     for msg in &header.messages {
         if msg.msg_type == MessageType::Attribute {
             let attr = if shared_message::is_shared(msg.flags) {
                 // Shared attribute: resolve the reference to get actual attribute data
                 shared_message::parse_shared_ref_sized(&msg.data, offset_size, length_size)
                     .and_then(|shared_ref| {
-                        shared_message::resolve_shared_message(
+                        shared_message::resolve_shared_message_in(
                             file_data,
                             &shared_ref,
                             MessageType::Attribute,
@@ -467,7 +706,7 @@ fn extract_attributes_with(
                         )
                     })
                     .and_then(|resolved| {
-                        AttributeMessage::parse_in_file(
+                        AttributeMessage::parse_in_storage(
                             &resolved,
                             file_data,
                             offset_size,
@@ -475,33 +714,19 @@ fn extract_attributes_with(
                         )
                     })
             } else {
-                AttributeMessage::parse_in_file(&msg.data, file_data, offset_size, length_size)
+                AttributeMessage::parse_in_storage(&msg.data, file_data, offset_size, length_size)
             };
             let attr = attr.and_then(|a| check_in_header(a, header));
             match attr {
-                Ok(attr) => attrs.push(attr),
+                Ok(attr) => {
+                    attrs.push(attr);
+                    orders.push(msg.creation_order.map_or(0, u32::from));
+                }
                 Err(e) => on_error(e)?,
             }
         }
     }
-
+    Ok(())
-    // Check for dense attributes via AttributeInfo message
-    let attr_info = find_attribute_info(header, offset_size)?;
-    if let Some(info) = attr_info
-        && let Some(fh_addr) = info.fractal_heap_address
-    {
-        extract_dense_attributes(
-            file_data,
-            &info,
-            fh_addr,
-            offset_size,
-            length_size,
-            &mut attrs,
-            on_error,
-        )?;
-    }
-
-    Ok(attrs)
 }
 
 /// Find and parse the Attribute Info message from an object header.
@@ -518,18 +743,21 @@ fn find_attribute_info(
     Ok(None)
 }
 
-/// Extract attributes from dense storage (fractal heap + B-tree v2).
+/// Extract attributes from dense storage (fractal heap + B-tree v2), and
-fn extract_dense_attributes(
+/// each one's creation order into `orders`.
-    file_data: &[u8],
+#[allow(clippy::too_many_arguments)]
+fn extract_dense_attributes<S: Storage + ?Sized>(
+    file_data: &S,
     attr_info: &AttributeInfoMessage,
     fh_addr: u64,
     offset_size: u8,
     length_size: u8,
     attrs: &mut Vec<AttributeMessage>,
+    orders: &mut Vec<u32>,
     on_error: &mut dyn FnMut(FormatError) -> Result<(), FormatError>,
 ) -> Result<(), FormatError> {
     // Parse fractal heap
-    let fh = FractalHeapHeader::parse(file_data, fh_addr as usize, offset_size, length_size)?;
+    let fh = FractalHeapHeader::parse_in(file_data, fh_addr, offset_size, length_size)?;
 
     // Parse B-tree v2 for name index (type 8)
     let btree_addr = attr_info
@@ -538,8 +766,13 @@ fn extract_dense_attributes(
             expected: 1,
             available: 0,
         })?;
-    let btree_hdr = BTreeV2Header::parse(file_data, btree_addr as usize, offset_size, length_size)?;
+    let btree_hdr = BTreeV2Header::parse_in(
-    let records = collect_btree_v2_records(file_data, &btree_hdr, offset_size, length_size)?;
+        file_data,
+        to_usize(btree_addr)? as u64,
+        offset_size,
+        length_size,
+    )?;
+    let records = collect_btree_v2_records_in(file_data, &btree_hdr, offset_size, length_size)?;
 
     for record in &records {
         // Per HDF5 spec, both type 8 and type 9 records start with heap_id:
@@ -556,12 +789,19 @@ fn extract_dense_attributes(
 
         // The data in the heap is a complete attribute message
         let attr = fh
-            .read_managed_object(file_data, id_bytes, offset_size)
+            .read_managed_object_in(file_data, id_bytes, offset_size)
             .and_then(|attr_data| {
-                AttributeMessage::parse_in_file(&attr_data, file_data, offset_size, length_size)
+                AttributeMessage::parse_in_storage(&attr_data, file_data, offset_size, length_size)
             });
         match attr {
-            Ok(attr) => attrs.push(attr),
+            Ok(attr) => {
+                attrs.push(attr);
+                let order = record
+                    .data
+                    .get(id_len + 1..id_len + 5)
+                    .map_or(0, |b| u32::from_le_bytes([b[0], b[1], b[2], b[3]]));
+                orders.push(order);
+            }
             Err(e) => on_error(e)?,
         }
     }
@@ -961,4 +1201,73 @@ mod tests {
         let strs = attr.read_as_strings().unwrap();
         assert_eq!(strs, vec!["abcd", "EFGH"]);
     }
+
+    /// Every object's attributes in h5py-written files read identically
+    /// through a read_at-only CountingStorage — compact ones, shared ones,
+    /// those behind an Attribute Info message and dense storage (its v2
+    /// B-tree name index included) — and through a slice as Storage.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::storage::CountingStorage;
+        let files: [(&str, &[u8]); 5] = [
+            ("attrs", include_bytes!("../tests/fixtures/attrs.h5")),
+            (
+                "mixed_attrs",
+                include_bytes!("../tests/fixtures/mixed_attrs.h5"),
+            ),
+            (
+                "dense_attrs",
+                include_bytes!("../tests/fixtures/dense_attrs.h5"),
+            ),
+            (
+                "dense_attrs_root",
+                include_bytes!("../tests/fixtures/dense_attrs_root.h5"),
+            ),
+            (
+                "shared_fill_value",
+                include_bytes!("../tests/fixtures/shared_fill_value.h5"),
+            ),
+        ];
+        let (mut same, mut dense, mut attrs) = (0, 0, 0);
+        for (name, file) in files {
+            let sb = crate::superblock::Superblock::parse(file, 0).unwrap();
+            let (os, ls) = (sb.offset_size, sb.length_size);
+            let mut addrs = vec![sb.root_group_address];
+            addrs.extend(
+                crate::group_v2::resolve_group_children(file, &sb, sb.root_group_address)
+                    .unwrap()
+                    .iter()
+                    .map(|e| e.object_header_address),
+            );
+            let storage = CountingStorage::new(file.to_vec());
+            for addr in addrs {
+                let header = ObjectHeader::parse(file, addr as usize, os, ls).unwrap();
+                let want = extract_attributes_full(file, &header, os, ls);
+                let slice_storage = extract_attributes_full_in(&file, &header, os, ls);
+                assert_eq!(format!("{slice_storage:?}"), format!("{want:?}"));
+                let got = extract_attributes_full_in(&storage, &header, os, ls);
+                let got_t = extract_attributes_tolerant_in(&storage, &header, os, ls);
+                let is_dense = find_attribute_info(&header, os)
+                    .unwrap()
+                    .is_some_and(|i| i.fractal_heap_address.is_some());
+                if is_dense {
+                    dense += 1;
+                }
+                attrs += want.as_ref().map_or(0, Vec::len);
+                assert_eq!(format!("{got:?}"), format!("{want:?}"), "{name}");
+                let want_t = extract_attributes_tolerant(file, &header, os, ls);
+                assert_eq!(format!("{got_t:?}"), format!("{want_t:?}"), "{name}");
+                same += 1;
+                for a in want.iter().flatten() {
+                    let one = find_attribute_in(&storage, &header, &a.name, os, ls);
+                    let want_one = find_attribute_in_file(file, &header, &a.name, os, ls);
+                    assert_eq!(format!("{one:?}"), format!("{want_one:?}"), "{name}");
+                }
+            }
+        }
+        assert!(
+            same >= 5 && dense >= 2 && attrs >= 5,
+            "{same} {dense} {attrs}"
+        );
+    }
 }

crates/clawhdf5-format/src/btree_v1.rs

@@ -4,6 +4,7 @@
 use alloc::vec::Vec;
 
 use crate::error::FormatError;
+use crate::storage::{Storage, read_exact_at};
 
 /// A parsed B-tree v1 node.
 #[derive(Debug, Clone)]
@@ -74,13 +75,28 @@ impl BTreeV1Node {
         file_data: &[u8],
         offset: usize,
         offset_size: u8,
+        length_size: u8,
+    ) -> Result<BTreeV1Node, FormatError> {
+        Self::parse_in(file_data, offset as u64, offset_size, length_size)
+    }
+
+    /// [`Self::parse`] over any [`Storage`]: one read of the node's header,
+    /// one of its keys and children.
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
         _length_size: u8,
     ) -> Result<BTreeV1Node, FormatError> {
         // signature(4) + node_type(1) + node_level(1) + entries_used(2) = 8
         // + left_sibling(offset_size) + right_sibling(offset_size)
         let os = offset_size as usize;
         let header_size = 8 + os * 2;
-        ensure_len(file_data, offset, header_size)?;
+        let header = read_exact_at(file, offset, header_size)?;
+        let file_data: &[u8] = &header;
+        // The header's read checked that `offset + header_size` fits.
+        let body_start = offset + header_size as u64;
+        let offset = 0usize;
 
         if &file_data[offset..offset + 4] != b"TREE" {
             return Err(FormatError::InvalidBTreeSignature);
@@ -102,31 +118,30 @@ impl BTreeV1Node {
         } else {
             Some(read_offset(file_data, pos, offset_size)?)
         };
-        pos += os;
 
         // For type 0: keys are offset_size bytes, children are offset_size bytes
         // Layout: key[0], child[0], key[1], child[1], ..., key[N-1], child[N-1], key[N]
         let eu = entries_used as usize;
         let key_size = os; // For type 0, key = offset_size
         let needed = eu * (key_size + os) + key_size; // eu children + (eu+1) keys
-        ensure_len(file_data, pos, needed)?;
+        let body = read_exact_at(file, body_start, needed)?;
+        let file_data: &[u8] = &body;
 
         let mut keys = Vec::with_capacity(eu + 1);
         let mut children = Vec::with_capacity(eu);
 
-        for _i in 0..eu {
+        if os == 0 {
-            // key[i]
+            // What reading the first key reports (and keeps `chunks_exact`
-            let key = read_offset(file_data, pos, offset_size)?;
+            // below from being given a zero size).
-            keys.push(key);
+            return Err(FormatError::InvalidOffsetSize(offset_size));
-            pos += key_size;
-            // child[i]
-            let child = read_offset(file_data, pos, offset_size)?;
-            children.push(child);
-            pos += os;
         }
-        // final key
+        // `needed` bytes: key[0], child[0], ..., child[eu - 1], key[eu].
-        let key = read_offset(file_data, pos, offset_size)?;
+        let (pairs, last) = file_data.split_at(eu * (key_size + os));
-        keys.push(key);
+        for pair in pairs.chunks_exact(key_size + os) {
+            keys.push(read_offset(pair, 0, offset_size)?);
+            children.push(read_offset(pair, key_size, offset_size)?);
+        }
+        keys.push(read_offset(last, 0, offset_size)?);
 
         Ok(BTreeV1Node {
             node_type,
@@ -150,11 +165,21 @@ pub fn collect_symbol_table_nodes(
     offset_size: u8,
     length_size: u8,
 ) -> Result<Vec<u64>, FormatError> {
-    collect_symbol_table_nodes_inner(file_data, btree_address, offset_size, length_size, 0)
+    collect_symbol_table_nodes_in(file_data, btree_address, offset_size, length_size)
 }
 
-fn collect_symbol_table_nodes_inner(
+/// [`collect_symbol_table_nodes`] over any [`Storage`]: two reads per node.
-    file_data: &[u8],
+pub fn collect_symbol_table_nodes_in<S: Storage + ?Sized>(
+    file: &S,
+    btree_address: u64,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<u64>, FormatError> {
+    collect_symbol_table_nodes_inner(file, btree_address, offset_size, length_size, 0)
+}
+
+fn collect_symbol_table_nodes_inner<S: Storage + ?Sized>(
+    file: &S,
     btree_address: u64,
     offset_size: u8,
     length_size: u8,
@@ -164,7 +189,7 @@ fn collect_symbol_table_nodes_inner(
         return Err(FormatError::NestingDepthExceeded);
     }
 
-    let node = BTreeV1Node::parse(file_data, btree_address as usize, offset_size, length_size)?;
+    let node = BTreeV1Node::parse_in(file, btree_address, offset_size, length_size)?;
 
     if node.node_type != 0 {
         return Err(FormatError::InvalidBTreeNodeType(node.node_type));
@@ -178,7 +203,7 @@ fn collect_symbol_table_nodes_inner(
         let mut result = Vec::new();
         for &child_addr in &node.children {
             let child_snods = collect_symbol_table_nodes_inner(
-                file_data,
+                file,
                 child_addr,
                 offset_size,
                 length_size,
@@ -317,4 +342,48 @@ mod tests {
         assert_eq!(node.entries_used, 1);
         assert_eq!(node.children, vec![0x50]);
     }
+
+    /// Nodes and trees, cut at every length, parse identically through a
+    /// `read_at`-only storage.
+    #[test]
+    fn storage_parse_matches_slice_parse() {
+        use crate::storage::CountingStorage;
+        let nodes = [
+            build_btree_node(0, 0, &[0, 5, 10], &[0x100, 0x200], None, None, 8),
+            build_btree_node(0, 0, &[0, 5], &[0x100], Some(0x40), Some(0x80), 4),
+            build_btree_node(1, 2, &[0, 5], &[0x100], None, Some(0x80), 8),
+        ];
+        for (n, node) in nodes.iter().enumerate() {
+            let os = if n == 1 { 4 } else { 8 };
+            for cut in 0..=node.len() {
+                let f = &node[..cut];
+                let storage = CountingStorage::new(f.to_vec());
+                let want = BTreeV1Node::parse(f, 0, os, 8);
+                let got = BTreeV1Node::parse_in(&storage, 0, os, 8);
+                assert_eq!(format!("{got:?}"), format!("{want:?}"));
+            }
+        }
+        let leaf1 = build_btree_node(0, 0, &[0, 5], &[0xA00], None, None, 8);
+        let leaf2 = build_btree_node(0, 0, &[5, 10], &[0xB00], None, None, 8);
+        let internal = build_btree_node(0, 1, &[0, 5, 10], &[0, 256], None, None, 8);
+        let mut file = vec![0u8; 512 + internal.len()];
+        file[..leaf1.len()].copy_from_slice(&leaf1);
+        file[256..256 + leaf2.len()].copy_from_slice(&leaf2);
+        file[512..].copy_from_slice(&internal);
+        for cut in [file.len(), 300, 260, 100, 10] {
+            let mut f = file.clone();
+            if cut < 512 {
+                // Truncate the leaves, keep the root.
+                f[cut..512].fill(0);
+            }
+            let storage = CountingStorage::new(f.clone());
+            assert_eq!(
+                collect_symbol_table_nodes_in(&storage, 512, 8, 8),
+                collect_symbol_table_nodes(&f, 512, 8, 8)
+            );
+        }
+        let storage = CountingStorage::new(file);
+        collect_symbol_table_nodes_in(&storage, 512, 8, 8).unwrap();
+        assert_eq!(storage.reads(), 6);
+    }
 }

crates/clawhdf5-format/src/btree_v2.rs

@@ -2,11 +2,14 @@
 
 #[cfg(not(feature = "std"))]
 use alloc::vec::Vec;
+use core::cmp::Ordering;
 
 #[cfg(feature = "checksum")]
 use byteorder::{ByteOrder, LittleEndian};
 
+use crate::addr::to_usize;
 use crate::error::FormatError;
+use crate::storage::{Storage, Window, len_usize};
 
 /// Parsed B-tree v2 header (signature "BTHD").
 #[derive(Debug, Clone)]
@@ -71,7 +74,7 @@ fn ensure_len(data: &[u8], pos: usize, needed: usize) -> Result<(), FormatError>
 
 /// Compute the number of bytes needed to represent a count, using variable-width encoding.
 /// B-tree v2 uses this for the number of records fields in internal nodes.
-fn bytes_for_max_records(max_nrec: u64) -> usize {
+pub(crate) fn bytes_for_max_records(max_nrec: u64) -> usize {
     if max_nrec == 0 {
         return 1;
     }
@@ -97,38 +100,52 @@ impl BTreeV2Header {
         offset_size: u8,
         length_size: u8,
     ) -> Result<BTreeV2Header, FormatError> {
-        ensure_len(file_data, offset, 4)?;
+        Self::parse_in(file_data, offset as u64, offset_size, length_size)
-        if &file_data[offset..offset + 4] != b"BTHD" {
+    }
+
+    /// [`Self::parse`] over any [`Storage`]: one bounded read of the
+    /// header.
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
+    ) -> Result<BTreeV2Header, FormatError> {
+        // Every field and the checksum; the window holds all of it or ends
+        // at the end of the file, so its bounds checks are the whole-file
+        // ones.
+        let full = 16 + usize::from(offset_size) + 2 + usize::from(length_size) + 4;
+        let w = Window::read(file, offset, full)?;
+        let d = &w.bytes;
+        w.ensure(0, 4)?;
+        if &d[..4] != b"BTHD" {
             return Err(FormatError::InvalidBTreeV2Signature);
         }
 
-        ensure_len(file_data, offset, 4 + 1 + 1 + 4 + 2 + 2 + 1 + 1)?;
+        w.ensure(0, 4 + 1 + 1 + 4 + 2 + 2 + 1 + 1)?;
-        let version = file_data[offset + 4];
+        let version = d[4];
         if version != 0 {
             return Err(FormatError::InvalidBTreeV2Version(version));
         }
 
-        let tree_type = file_data[offset + 5];
+        let tree_type = d[5];
-        let node_size = u32::from_le_bytes([
+        let node_size = u32::from_le_bytes([d[6], d[7], d[8], d[9]]);
-            file_data[offset + 6],
+        let record_size = u16::from_le_bytes([d[10], d[11]]);
-            file_data[offset + 7],
+        let depth = u16::from_le_bytes([d[12], d[13]]);
-            file_data[offset + 8],
+        let _split_percent = d[14];
-            file_data[offset + 9],
+        let _merge_percent = d[15];
-        ]);
-        let record_size = u16::from_le_bytes([file_data[offset + 10], file_data[offset + 11]]);
-        let depth = u16::from_le_bytes([file_data[offset + 12], file_data[offset + 13]]);
-        let _split_percent = file_data[offset + 14];
-        let _merge_percent = file_data[offset + 15];
 
-        let mut pos = offset + 16;
+        let mut pos = 16;
-        let root_node_address = read_offset(file_data, pos, offset_size)?;
+        w.ensure(pos, usize::from(offset_size))?;
+        let root_node_address = read_offset(d, pos, offset_size)?;
         pos += offset_size as usize;
 
-        ensure_len(file_data, pos, 2)?;
+        w.ensure(pos, 2)?;
-        let num_records_in_root = u16::from_le_bytes([file_data[pos], file_data[pos + 1]]);
+        let num_records_in_root = u16::from_le_bytes([d[pos], d[pos + 1]]);
         pos += 2;
 
-        let total_records = read_offset(file_data, pos, length_size)?;
+        w.ensure(pos, usize::from(length_size))?;
+        let total_records = read_offset(d, pos, length_size)?;
         #[allow(unused_assignments)]
         {
             pos += length_size as usize;
@@ -137,9 +154,9 @@ impl BTreeV2Header {
         // Validate header checksum
         #[cfg(feature = "checksum")]
         {
-            ensure_len(file_data, pos, 4)?;
+            w.ensure(pos, 4)?;
-            let stored = LittleEndian::read_u32(&file_data[pos..pos + 4]);
+            let stored = LittleEndian::read_u32(&d[pos..pos + 4]);
-            let computed = crate::checksum::jenkins_lookup3(&file_data[offset..pos]);
+            let computed = crate::checksum::jenkins_lookup3(&d[..pos]);
             if computed != stored {
                 return Err(FormatError::ChecksumMismatch {
                     expected: stored,
@@ -163,7 +180,7 @@ impl BTreeV2Header {
 /// Compute maximum records per node for a given depth level.
 /// leaf: (node_size - overhead) / record_size
 /// internal: depends on pointers
-fn max_records_leaf(node_size: u32, record_size: u16) -> u64 {
+pub(crate) fn max_records_leaf(node_size: u32, record_size: u16) -> u64 {
     // Leaf overhead: signature(4) + version(1) + type(1) + checksum(4) = 10
     let overhead = 10u32;
     if node_size <= overhead || record_size == 0 {
@@ -189,6 +206,17 @@ pub fn collect_btree_v2_records(
     header: &BTreeV2Header,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<BTreeV2Record>, FormatError> {
+    collect_btree_v2_records_in(file_data, header, offset_size, length_size)
+}
+
+/// [`collect_btree_v2_records`] over any [`Storage`]: one bounded read per
+/// node.
+pub fn collect_btree_v2_records_in<S: Storage + ?Sized>(
+    file: &S,
+    header: &BTreeV2Header,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<BTreeV2Record>, FormatError> {
     if header.total_records == 0 || header.num_records_in_root == 0 {
         return Ok(Vec::new());
@@ -208,24 +236,25 @@ pub fn collect_btree_v2_records(
     // millions of records from a few kilobytes. Counting against what the
     // file could physically contain bounds that without trusting the
     // header's own `total_records`.
-    let mut budget = file_data.len() / usize::from(header.record_size.max(1));
+    let mut budget = len_usize(file) / usize::from(header.record_size.max(1));
 
     let max_leaf_nrec = max_records_leaf(header.node_size, header.record_size);
 
     if header.depth == 0 {
         // Root is a leaf
         parse_leaf_records(
-            file_data,
+            file,
-            header.root_node_address as usize,
+            to_usize(header.root_node_address)?,
             header.num_records_in_root,
             header.record_size,
+            header.node_size,
         )
     } else {
         // Root is internal; traverse recursively
         let mut records = Vec::new();
         collect_internal_records(
-            file_data,
+            file,
-            header.root_node_address as usize,
+            to_usize(header.root_node_address)?,
             header.num_records_in_root,
             header.depth,
             header.record_size,
@@ -240,36 +269,72 @@ pub fn collect_btree_v2_records(
     }
 }
 
+/// A node's bytes: `want` bytes at `offset` (fewer only at the end of the
+/// file), after checking its 4-byte signature. A node is read in one piece
+/// when it fits in `node_size` (every valid node does); a larger claimed
+/// extent — record counts from a damaged parent — is first checked against
+/// the end of the file, so it costs a read only of bytes the file has.
+/// Bounds errors are the whole-file ones: the signature check needs the
+/// first 6 bytes, then `checks` — `(position, length)` pairs relative to
+/// the node, in the order the parser checks them — must lie in the file.
+fn read_node<'a, S: Storage + ?Sized>(
+    file: &'a S,
+    offset: usize,
+    want: usize,
+    node_size: u32,
+    signature: &[u8; 4],
+    checks: &[(usize, usize)],
+) -> Result<Window<'a>, FormatError> {
+    let one_read = usize::try_from(node_size).unwrap_or(usize::MAX).max(6);
+    let w = Window::read(file, offset as u64, want.min(one_read))?;
+    w.ensure(0, 6)?;
+    if &w.bytes[..4] != signature {
+        return Err(FormatError::InvalidBTreeV2Signature);
+    }
+    if want <= one_read {
+        return Ok(w);
+    }
+    for &(rel, len) in checks {
+        Window::check_extent(file, offset as u64, rel, len)?;
+    }
+    Window::read(file, offset as u64, want)
+}
+
 /// Parse records from a leaf node (signature "BTLF").
-fn parse_leaf_records(
+fn parse_leaf_records<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file: &S,
     offset: usize,
     num_records: u16,
     record_size: u16,
+    node_size: u32,
 ) -> Result<Vec<BTreeV2Record>, FormatError> {
     // signature(4) + version(1) + type(1) = 6 bytes header
-    ensure_len(file_data, offset, 6)?;
+    let pos = 6;
-    if &file_data[offset..offset + 4] != b"BTLF" {
-        return Err(FormatError::InvalidBTreeV2Signature);
-    }
-
-    let pos = offset + 6;
     let rs = record_size as usize;
     let total = (num_records as usize)
         .checked_mul(rs)
         .ok_or(FormatError::UnexpectedEof {
             expected: usize::MAX,
-            available: file_data.len(),
+            available: len_usize(file),
         })?;
-    ensure_len(file_data, pos, total)?;
+    let w = read_node(
+        file,
+        offset,
+        pos + total + 4,
+        node_size,
+        b"BTLF",
+        &[(pos, total)],
+    )?;
+    let d = &w.bytes;
+    w.ensure(pos, total)?;
 
     // Validate checksum: 4 bytes after records + padding
     #[cfg(feature = "checksum")]
     {
         let checksum_pos = pos + total;
-        if file_data.len() >= checksum_pos + 4 {
+        if d.len() >= checksum_pos + 4 {
-            let stored = LittleEndian::read_u32(&file_data[checksum_pos..checksum_pos + 4]);
+            let stored = LittleEndian::read_u32(&d[checksum_pos..checksum_pos + 4]);
-            let computed = crate::checksum::jenkins_lookup3(&file_data[offset..checksum_pos]);
+            let computed = crate::checksum::jenkins_lookup3(&d[..checksum_pos]);
             if computed != stored {
                 return Err(FormatError::ChecksumMismatch {
                     expected: stored,
@@ -283,45 +348,57 @@ fn parse_leaf_records(
     for i in 0..num_records as usize {
         let start = pos + i * rs;
         records.push(BTreeV2Record {
-            data: file_data[start..start + rs].to_vec(),
+            data: d[start..start + rs].to_vec(),
         });
     }
     Ok(records)
 }
 
-/// Recursively collect records from an internal node.
+/// An internal node read from the file: its bytes (from the signature on),
-#[allow(clippy::too_many_arguments, clippy::only_used_in_recursion)]
+/// where its records start, and its children as `(address, record count)`.
-fn collect_internal_records(
+struct InternalNode<'a> {
-    file_data: &[u8],
+    node: Window<'a>,
+    records_start: usize,
+    children: Vec<(u64, u16)>,
+}
+
+impl InternalNode<'_> {
+    /// Record `i`, `rs` bytes long.
+    fn record(&self, i: usize, rs: usize) -> Result<&[u8], FormatError> {
+        let overflow = || FormatError::UnexpectedEof {
+            expected: usize::MAX,
+            available: usize::MAX,
+        };
+        let rec_start = i
+            .checked_mul(rs)
+            .and_then(|o| self.records_start.checked_add(o))
+            .ok_or_else(overflow)?;
+        self.node.ensure(rec_start, rs)?;
+        Ok(&self.node.bytes[rec_start..rec_start + rs])
+    }
+}
+
+/// An internal node's layout: where its records start, and its children as
+/// `(address, record count)`.
+#[allow(clippy::too_many_arguments)]
+fn read_internal_node<S: Storage + ?Sized>(
+    file: &S,
     offset: usize,
     num_records: u16,
     depth: u16,
     record_size: u16,
     node_size: u32,
     offset_size: u8,
-    length_size: u8,
     max_leaf_nrec: u64,
-    budget: &mut usize,
+) -> Result<InternalNode<'_>, FormatError> {
-    out: &mut Vec<BTreeV2Record>,
-) -> Result<(), FormatError> {
-    // signature(4) + version(1) + type(1) = 6
-    ensure_len(file_data, offset, 6)?;
-    if &file_data[offset..offset + 4] != b"BTIN" {
-        return Err(FormatError::InvalidBTreeV2Signature);
-    }
-
     let nr = num_records as usize;
     let rs = record_size as usize;
-    let mut pos = offset + 6;
 
-    // Read all records first
+    // Records first
     let records_total = nr.checked_mul(rs).ok_or(FormatError::UnexpectedEof {
         expected: usize::MAX,
-        available: file_data.len(),
+        available: len_usize(file),
     })?;
-    ensure_len(file_data, pos, records_total)?;
-    let records_start = pos;
-    pos += records_total;
 
     // Child pointer layout, as libhdf5 computes it (H5B2__hdr_init): the
     // child's record count is always encoded in the width needed for a
@@ -344,32 +421,105 @@ fn collect_internal_records(
 
     let num_children = nr + 1;
     let child_ptr_size = offset_size as usize + nrec_width + total_nrec_width;
-    ensure_len(file_data, pos, num_children * child_ptr_size)?;
+    let pointers = num_children * child_ptr_size;
+
+    // signature(4) + version(1) + type(1) = 6, records, pointers, checksum.
+    let w = read_node(
+        file,
+        offset,
+        6 + records_total + pointers + 4,
+        node_size,
+        b"BTIN",
+        &[(6, records_total), (6 + records_total, pointers)],
+    )?;
+    let d = &w.bytes;
+    let mut pos = 6;
+    w.ensure(pos, records_total)?;
+    let records_start = pos;
+    pos += records_total;
+
+    w.ensure(pos, pointers)?;
 
-    // Read child pointers
     let mut children = Vec::with_capacity(num_children);
     for _ in 0..num_children {
-        let addr = read_offset(file_data, pos, offset_size)?;
+        let addr = read_offset(d, pos, offset_size)?;
         pos += offset_size as usize;
-        let child_nrec = read_var_uint(file_data, pos, nrec_width)? as u16;
+        let child_nrec = read_var_uint(d, pos, nrec_width)? as u16;
         pos += nrec_width;
         pos += total_nrec_width; // skip total records in subtree
         children.push((addr, child_nrec));
     }
 
+    // The checksum follows the child pointers and covers the node up to it.
+    // Lookups prune children by the keys in this node, so an unverified
+    // internal node could hide a record without any error: libhdf5 refuses
+    // a mismatch here, and so does this.
+    #[cfg(feature = "checksum")]
+    {
+        w.ensure(pos, 4)?;
+        let stored = LittleEndian::read_u32(&d[pos..pos + 4]);
+        let computed = crate::checksum::jenkins_lookup3(&d[..pos]);
+        if computed != stored {
+            return Err(FormatError::ChecksumMismatch {
+                expected: stored,
+                computed,
+            });
+        }
+    }
+    Ok(InternalNode {
+        node: w,
+        records_start,
+        children,
+    })
+}
+
+/// Recursively collect records from an internal node.
+#[allow(clippy::too_many_arguments, clippy::only_used_in_recursion)]
+fn collect_internal_records<S: Storage + ?Sized>(
+    file: &S,
+    offset: usize,
+    num_records: u16,
+    depth: u16,
+    record_size: u16,
+    node_size: u32,
+    offset_size: u8,
+    length_size: u8,
+    max_leaf_nrec: u64,
+    budget: &mut usize,
+    out: &mut Vec<BTreeV2Record>,
+) -> Result<(), FormatError> {
+    let nr = num_records as usize;
+    let rs = record_size as usize;
+    let node = read_internal_node(
+        file,
+        offset,
+        num_records,
+        depth,
+        record_size,
+        node_size,
+        offset_size,
+        max_leaf_nrec,
+    )?;
+    let child_depth = depth - 1;
+
     // Interleave: child[0], record[0], child[1], record[1], ..., child[nr]
     // We collect child[0] records, then record[0], then child[1], etc.
-    for (i, &(child_addr, child_nrec)) in children.iter().enumerate() {
+    for (i, &(child_addr, child_nrec)) in node.children.iter().enumerate() {
         if child_depth == 0 {
             // Before parsing, so a refused tree is not also a large allocation.
             spend(budget, usize::from(child_nrec))?;
-            let leaf_recs =
+            let leaf_recs = parse_leaf_records(
-                parse_leaf_records(file_data, child_addr as usize, child_nrec, record_size)?;
+                file,
+                to_usize(child_addr)?,
+                child_nrec,
+                record_size,
+                node_size,
+            )?;
             out.extend(leaf_recs);
         } else {
             collect_internal_records(
-                file_data,
+                file,
-                child_addr as usize,
+                to_usize(child_addr)?,
                 child_nrec,
                 child_depth,
                 record_size,
@@ -384,32 +534,10 @@ fn collect_internal_records(
 
         // Add record[i] (except after the last child)
         if i < nr {
-            let rec_offset = i.checked_mul(rs).ok_or(FormatError::UnexpectedEof {
+            let data = node.record(i, rs)?;
-                expected: usize::MAX,
-                available: file_data.len(),
-            })?;
-            let rec_start =
-                records_start
-                    .checked_add(rec_offset)
-                    .ok_or(FormatError::UnexpectedEof {
-                        expected: usize::MAX,
-                        available: file_data.len(),
-                    })?;
-            let rec_end = rec_start
-                .checked_add(rs)
-                .ok_or(FormatError::UnexpectedEof {
-                    expected: usize::MAX,
-                    available: file_data.len(),
-                })?;
-            if rec_end > file_data.len() {
-                return Err(FormatError::UnexpectedEof {
-                    expected: rec_end,
-                    available: file_data.len(),
-                });
-            }
             spend(budget, 1)?;
             out.push(BTreeV2Record {
-                data: file_data[rec_start..rec_end].to_vec(),
+                data: data.to_vec(),
             });
         }
     }
@@ -417,11 +545,135 @@ fn collect_internal_records(
     Ok(())
 }
 
+/// The records of a B-tree v2 that fall in one key range, found by
+/// descending the tree instead of reading all of it.
+///
+/// `cmp` places a record relative to the range: `Less` if the record sorts
+/// before it, `Greater` if after, `Equal` if the record is in it. The tree
+/// must be ordered consistently with `cmp`, as libhdf5 orders it (a link or
+/// attribute name index by name hash, so all records with one hash form a
+/// range whatever order their names are in). Only the nodes whose key
+/// interval overlaps the range are read: O(depth) nodes plus those holding
+/// the matches. Matches come in tree order.
+pub fn find_btree_v2_records(
+    file_data: &[u8],
+    header: &BTreeV2Header,
+    offset_size: u8,
+    cmp: &mut dyn FnMut(&[u8]) -> Ordering,
+) -> Result<Vec<BTreeV2Record>, FormatError> {
+    find_btree_v2_records_in(file_data, header, offset_size, cmp)
+}
+
+/// [`find_btree_v2_records`] over any [`Storage`]: one bounded read per
+/// node visited.
+pub fn find_btree_v2_records_in<S: Storage + ?Sized>(
+    file: &S,
+    header: &BTreeV2Header,
+    offset_size: u8,
+    cmp: &mut dyn FnMut(&[u8]) -> Ordering,
+) -> Result<Vec<BTreeV2Record>, FormatError> {
+    if header.total_records == 0 || header.num_records_in_root == 0 {
+        return Ok(Vec::new());
+    }
+    if header.depth > MAX_DEPTH {
+        return Err(FormatError::NestingDepthExceeded);
+    }
+    // As in `collect_btree_v2_records`: a valid tree cannot hold more
+    // records than the file has room for, however its children are shared.
+    let mut budget = len_usize(file) / usize::from(header.record_size.max(1));
+    let max_leaf_nrec = max_records_leaf(header.node_size, header.record_size);
+    let mut out = Vec::new();
+    find_in_node(
+        file,
+        header,
+        to_usize(header.root_node_address)?,
+        header.num_records_in_root,
+        header.depth,
+        offset_size,
+        max_leaf_nrec,
+        cmp,
+        &mut budget,
+        &mut out,
+    )?;
+    Ok(out)
+}
+
+#[allow(clippy::too_many_arguments)]
+fn find_in_node<S: Storage + ?Sized>(
+    file: &S,
+    header: &BTreeV2Header,
+    offset: usize,
+    num_records: u16,
+    depth: u16,
+    offset_size: u8,
+    max_leaf_nrec: u64,
+    cmp: &mut dyn FnMut(&[u8]) -> Ordering,
+    budget: &mut usize,
+    out: &mut Vec<BTreeV2Record>,
+) -> Result<(), FormatError> {
+    spend(budget, usize::from(num_records))?;
+    if depth == 0 {
+        let records = parse_leaf_records(
+            file,
+            offset,
+            num_records,
+            header.record_size,
+            header.node_size,
+        )?;
+        out.extend(
+            records
+                .into_iter()
+                .filter(|r| cmp(&r.data) == Ordering::Equal),
+        );
+        return Ok(());
+    }
+    let rs = usize::from(header.record_size);
+    let node = read_internal_node(
+        file,
+        offset,
+        num_records,
+        depth,
+        header.record_size,
+        header.node_size,
+        offset_size,
+        max_leaf_nrec,
+    )?;
+    let nr = usize::from(num_records);
+    let mut order = Vec::with_capacity(nr);
+    for i in 0..nr {
+        order.push(cmp(node.record(i, rs)?));
+    }
+    // Child `i` holds the keys between record `i - 1` and record `i`: it can
+    // hold a match unless the record before it is already past the range or
+    // the record after it is still before it.
+    for (i, &(child_addr, child_nrec)) in node.children.iter().enumerate() {
+        let after_left = i == 0 || order[i - 1] != Ordering::Greater;
+        let before_right = i == nr || order[i] != Ordering::Less;
+        if after_left && before_right {
+            find_in_node(
+                file,
+                header,
+                to_usize(child_addr)?,
+                child_nrec,
+                depth - 1,
+                offset_size,
+                max_leaf_nrec,
+                cmp,
+                budget,
+                out,
+            )?;
+        }
+        if i < nr && order[i] == Ordering::Equal {
+            out.push(BTreeV2Record {
+                data: node.record(i, rs)?.to_vec(),
+            });
+        }
+    }
+    Ok(())
+}
+
 /// Most records a subtree whose root is at `depth` can hold (libhdf5's
-/// `cum_max_nrec`): a leaf holds `max_leaf_nrec`; an internal node at depth
+/// `cum_max_nrec`). See [`node_info`].
-/// `d` holds `max_nrec(d)` records and `max_nrec(d) + 1` subtrees of depth
-/// `d - 1`, where `max_nrec(d)` is what fits in a node once each record is
-/// paired with a child pointer of the width depth `d` needs.
 fn cum_max_records(
     node_size: u32,
     record_size: u16,
@@ -429,24 +681,82 @@ fn cum_max_records(
     max_leaf_nrec: u64,
     depth: u16,
 ) -> u64 {
+    node_info_from_leaf(node_size, record_size, offset_size, max_leaf_nrec, depth)
+        .last()
+        .map_or(max_leaf_nrec, |n| n.cum_max_nrec)
+}
+
+/// Capacity of a B-tree v2 node at one depth, as libhdf5 computes it
+/// (`H5B2__hdr_init`'s `node_info`).
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub(crate) struct NodeInfo {
+    /// Most records one node at this depth holds.
+    pub(crate) max_nrec: u64,
+    /// Most records a subtree rooted at this depth holds.
+    pub(crate) cum_max_nrec: u64,
+    /// Bytes a subtree's total record count takes in a pointer to a node
+    /// at this depth (0 for a leaf, whose count is its own).
+    pub(crate) cum_max_nrec_size: usize,
+}
+
+/// Node capacities for depths `0..=depth` (entry `d` for depth `d`): a leaf
+/// holds `max_nrec(0)` records; an internal node at depth `d` holds
+/// `max_nrec(d)` records and `max_nrec(d) + 1` subtrees of depth `d - 1`,
+/// where `max_nrec(d)` is what fits in a node once each record is paired
+/// with a child pointer of the width depth `d` needs (address, the child's
+/// record count in the width a *leaf's* maximum needs, and below the first
+/// internal level the child subtree's total in the width its maximum
+/// needs), with one pointer more than records.
+pub(crate) fn node_info(
+    node_size: u32,
+    record_size: u16,
+    offset_size: u8,
+    depth: u16,
+) -> Vec<NodeInfo> {
+    let max_leaf = max_records_leaf(node_size, record_size);
+    node_info_from_leaf(node_size, record_size, offset_size, max_leaf, depth)
+}
+
+fn node_info_from_leaf(
+    node_size: u32,
+    record_size: u16,
+    offset_size: u8,
+    max_leaf_nrec: u64,
+    depth: u16,
+) -> Vec<NodeInfo> {
     // Internal node overhead: signature(4) + version(1) + type(1) + checksum(4).
     const PREFIX: u64 = 10;
     let nrec_width = bytes_for_max_records(max_leaf_nrec) as u64;
-    let mut cum = max_leaf_nrec;
+    let mut info = Vec::with_capacity(usize::from(depth) + 1);
-    let mut cum_width = 0u64;
+    info.push(NodeInfo {
+        max_nrec: max_leaf_nrec,
+        cum_max_nrec: max_leaf_nrec,
+        cum_max_nrec_size: 0,
+    });
     for d in 1..=depth {
-        let ptr = u64::from(offset_size) + nrec_width + if d > 1 { cum_width } else { 0 };
+        let below = info[usize::from(d) - 1];
+        let ptr = u64::from(offset_size)
+            + nrec_width
+            + if d > 1 {
+                below.cum_max_nrec_size as u64
+            } else {
+                0
+            };
         let max_nrec = u64::from(node_size)
             .saturating_sub(PREFIX)
             .saturating_sub(ptr)
             / (u64::from(record_size) + ptr).max(1);
-        cum = max_nrec
+        let cum = max_nrec
             .saturating_add(1)
-            .saturating_mul(cum)
+            .saturating_mul(below.cum_max_nrec)
             .saturating_add(max_nrec);
-        cum_width = bytes_for_max_records(cum) as u64;
+        info.push(NodeInfo {
+            max_nrec,
+            cum_max_nrec: cum,
+            cum_max_nrec_size: bytes_for_max_records(cum),
+        });
     }
-    cum
+    info
 }
 
 #[cfg(test)]
@@ -536,6 +846,8 @@ mod tests {
             buf.extend_from_slice(&child_nrec.to_le_bytes()[..nrec_width]);
             buf.resize(buf.len() + total_width, 0);
         }
+        let sum = crate::checksum::jenkins_lookup3(&buf);
+        buf.extend_from_slice(&sum.to_le_bytes());
         buf
     }
 

crates/clawhdf5-format/src/btree_v2_write.rs

@@ -0,0 +1,503 @@
+//! Writing version-2 B-trees: a header (`BTHD`) and its nodes, leaves
+//! (`BTLF`) and, for more records than one leaf holds, internal nodes
+//! (`BTIN`) to any depth.
+//!
+//! Node capacities come from [`crate::btree_v2::node_info`], the arithmetic
+//! libhdf5 uses (`H5B2__hdr_init`) and the reader decodes pointers with, so
+//! the pointer widths the writer encodes are the ones every reader expects.
+
+use crate::addr::saturating_usize;
+#[cfg(not(feature = "std"))]
+use alloc::{format, vec, vec::Vec};
+
+use crate::btree_v2::{NodeInfo, bytes_for_max_records, node_info};
+use crate::checksum::jenkins_lookup3;
+use crate::error::FormatError;
+
+/// How a B-tree is laid out: its record type and node geometry, as the
+/// header records them.
+#[derive(Debug, Clone, Copy)]
+pub(crate) struct BTreeV2Params {
+    /// Record type (5: link names, 6: link creation order, 8: attribute
+    /// names, 9: attribute creation order, 10/11: chunks).
+    pub(crate) tree_type: u8,
+    /// Bytes per node.
+    pub(crate) node_size: u32,
+    /// Bytes per record.
+    pub(crate) record_size: u16,
+    /// Split and merge percentages. The writer fills nodes itself; these
+    /// only tell libhdf5 when to split and merge as it modifies the tree.
+    pub(crate) split_percent: u8,
+    pub(crate) merge_percent: u8,
+}
+
+/// Size of a B-tree v2 header.
+pub(crate) fn header_size(offset_size: u8, length_size: u8) -> usize {
+    4 + 1 + 1 + 4 + 2 + 2 + 1 + 1 + offset_size as usize + 2 + length_size as usize + 4
+}
+
+/// Deepest tree the writer builds. Even at the smallest fan-out libhdf5's
+/// arithmetic allows, a few levels hold more records than any file could.
+const MAX_WRITE_DEPTH: u16 = 32;
+
+/// Write a B-tree v2 holding `records` (`record_size` bytes each,
+/// concatenated, already in the tree's key order) at `addr`: the header,
+/// then its nodes, each `node_size` bytes. No records gives a header with
+/// an undefined root.
+///
+/// The tree is as shallow as the node size allows: a single leaf when the
+/// records fit one, otherwise internal nodes above leaves. Records are
+/// spread evenly over each node's children, so every node but the root is
+/// at least about half full (above libhdf5's merge threshold, which is below
+/// half), and each node holds at most its depth's maximum.
+pub(crate) fn build_btree_v2(
+    p: BTreeV2Params,
+    records: &[u8],
+    addr: u64,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<u8>, FormatError> {
+    let rs = usize::from(p.record_size);
+    if rs == 0 || !records.len().is_multiple_of(rs) {
+        return Err(FormatError::SerializationError(format!(
+            "B-tree v2 records are {} bytes, not a multiple of the record size {rs}",
+            records.len()
+        )));
+    }
+    let n = (records.len() / rs) as u64;
+    let hdr_len = header_size(offset_size, length_size);
+
+    // The shallowest depth whose subtree can hold every record.
+    let mut info = node_info(p.node_size, p.record_size, offset_size, 0);
+    let max_leaf = info[0].max_nrec;
+    if max_leaf == 0 || max_leaf > u64::from(u16::MAX) {
+        return Err(FormatError::SerializationError(format!(
+            "a {}-byte B-tree v2 node holds {max_leaf} {}-byte records; \
+             a node holds 1 to 65535",
+            p.node_size, p.record_size
+        )));
+    }
+    let mut depth = 0u16;
+    while info[usize::from(depth)].cum_max_nrec < n {
+        depth += 1;
+        if depth > MAX_WRITE_DEPTH {
+            return Err(FormatError::SerializationError(format!(
+                "{n} records do not fit a B-tree v2 of {}-byte nodes",
+                p.node_size
+            )));
+        }
+        info = node_info(p.node_size, p.record_size, offset_size, depth);
+        let max = info[usize::from(depth)].max_nrec;
+        if max == 0 || max > u64::from(u16::MAX) {
+            return Err(FormatError::SerializationError(format!(
+                "a {}-byte B-tree v2 internal node holds {max} records; \
+                 a node holds 1 to 65535",
+                p.node_size
+            )));
+        }
+    }
+
+    let mut w = TreeWriter {
+        p,
+        records,
+        info: &info,
+        nrec_width: bytes_for_max_records(max_leaf),
+        offset_size,
+        first_node: addr + hdr_len as u64,
+        nodes: Vec::new(),
+    };
+    let root = (n > 0)
+        .then(|| w.node(depth, 0, saturating_usize(n)))
+        .transpose()?;
+
+    let mut out = Vec::with_capacity(hdr_len + w.nodes.len() * p.node_size as usize);
+    out.extend_from_slice(b"BTHD");
+    out.push(0); // version
+    out.push(p.tree_type);
+    out.extend_from_slice(&p.node_size.to_le_bytes());
+    out.extend_from_slice(&p.record_size.to_le_bytes());
+    out.extend_from_slice(&depth.to_le_bytes());
+    out.push(p.split_percent);
+    out.push(p.merge_percent);
+    match root {
+        Some(r) => push_uint(&mut out, r.addr, offset_size as usize),
+        None => out.extend(core::iter::repeat_n(0xFF, offset_size as usize)),
+    }
+    let root_nrec = root.map_or(0, |r| r.nrec);
+    out.extend_from_slice(&(root_nrec as u16).to_le_bytes());
+    push_uint(&mut out, n, length_size as usize);
+    let sum = jenkins_lookup3(&out);
+    out.extend_from_slice(&sum.to_le_bytes());
+    debug_assert_eq!(out.len(), hdr_len);
+    for node in &w.nodes {
+        out.extend_from_slice(node);
+    }
+    Ok(out)
+}
+
+/// A written node, as its parent points at it.
+#[derive(Debug, Clone, Copy)]
+struct NodeRef {
+    addr: u64,
+    /// Records in the node itself.
+    nrec: u64,
+    /// Records in the subtree it roots.
+    all_nrec: u64,
+}
+
+struct TreeWriter<'a> {
+    p: BTreeV2Params,
+    records: &'a [u8],
+    info: &'a [NodeInfo],
+    /// Width of a child's record count: what a leaf's maximum needs.
+    nrec_width: usize,
+    offset_size: u8,
+    /// Address of the first node (right after the header).
+    first_node: u64,
+    /// Nodes in file order (children before their parent).
+    nodes: Vec<Vec<u8>>,
+}
+
+impl TreeWriter<'_> {
+    fn record(&self, i: usize) -> &[u8] {
+        let rs = usize::from(self.p.record_size);
+        &self.records[i * rs..(i + 1) * rs]
+    }
+
+    fn push_node(&mut self, mut node: Vec<u8>) -> u64 {
+        // The checksum covers the node up to it, not the padding after.
+        let sum = jenkins_lookup3(&node);
+        node.extend_from_slice(&sum.to_le_bytes());
+        debug_assert!(node.len() <= self.p.node_size as usize);
+        node.resize(self.p.node_size as usize, 0);
+        let addr = self.first_node + self.nodes.len() as u64 * u64::from(self.p.node_size);
+        self.nodes.push(node);
+        addr
+    }
+
+    /// Write the subtree of `depth` holding records `first..first + n`.
+    fn node(&mut self, depth: u16, first: usize, n: usize) -> Result<NodeRef, FormatError> {
+        let rs = usize::from(self.p.record_size);
+        let mut node = Vec::with_capacity(self.p.node_size as usize);
+        if depth == 0 {
+            debug_assert!(n as u64 <= self.info[0].max_nrec);
+            node.extend_from_slice(b"BTLF");
+            node.push(0); // version
+            node.push(self.p.tree_type);
+            node.extend_from_slice(&self.records[first * rs..(first + n) * rs]);
+            let addr = self.push_node(node);
+            return Ok(NodeRef {
+                addr,
+                nrec: n as u64,
+                all_nrec: n as u64,
+            });
+        }
+
+        // As few children as hold the records, at least two, with the
+        // records spread evenly: `k` children and `k - 1` records between
+        // them.
+        let below = self.info[usize::from(depth) - 1].cum_max_nrec;
+        let k = (n as u64 + 1).div_ceil(below + 1).max(2);
+        let max = self.info[usize::from(depth)].max_nrec;
+        if k - 1 > max || (n as u64) < k - 1 + k {
+            return Err(FormatError::SerializationError(format!(
+                "cannot spread {n} B-tree v2 records over {k} children at depth {depth}"
+            )));
+        }
+        let k = saturating_usize(k);
+        let in_children = n - (k - 1);
+        let (base, extra) = (in_children / k, in_children % k);
+
+        let mut children = Vec::with_capacity(k);
+        let mut separators = Vec::with_capacity(k - 1);
+        let mut next = first;
+        for c in 0..k {
+            let m = base + usize::from(c < extra);
+            children.push(self.node(depth - 1, next, m)?);
+            next += m;
+            if c + 1 < k {
+                separators.push(next);
+                next += 1;
+            }
+        }
+        debug_assert_eq!(next, first + n);
+
+        node.extend_from_slice(b"BTIN");
+        node.push(0); // version
+        node.push(self.p.tree_type);
+        for &s in &separators {
+            node.extend_from_slice(self.record(s));
+        }
+        let total_width = if depth > 1 {
+            self.info[usize::from(depth) - 1].cum_max_nrec_size
+        } else {
+            0
+        };
+        for c in &children {
+            push_uint(&mut node, c.addr, self.offset_size as usize);
+            push_uint(&mut node, c.nrec, self.nrec_width);
+            if depth > 1 {
+                push_uint(&mut node, c.all_nrec, total_width);
+            }
+        }
+        let addr = self.push_node(node);
+        Ok(NodeRef {
+            addr,
+            nrec: (k - 1) as u64,
+            all_nrec: n as u64,
+        })
+    }
+}
+
+/// Append `v` as a `width`-byte little-endian integer.
+fn push_uint(buf: &mut Vec<u8>, v: u64, width: usize) {
+    let bytes = v.to_le_bytes();
+    buf.extend_from_slice(&bytes[..width.min(8)]);
+    buf.extend(vec![0u8; width.saturating_sub(8)]);
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records};
+
+    fn params(node_size: u32, record_size: u16) -> BTreeV2Params {
+        BTreeV2Params {
+            tree_type: 5,
+            node_size,
+            record_size,
+            split_percent: 100,
+            merge_percent: 40,
+        }
+    }
+
+    /// `n` 11-byte records: a big-endian counter, so byte order is key order.
+    fn records(n: usize, rs: usize) -> Vec<u8> {
+        let mut out = Vec::with_capacity(n * rs);
+        for i in 0..n {
+            let mut r = vec![0u8; rs];
+            r[..8].copy_from_slice(&(i as u64).to_be_bytes());
+            out.extend_from_slice(&r);
+        }
+        out
+    }
+
+    fn roundtrip(node_size: u32, rs: u16, n: usize, os: u8, ls: u8) -> BTreeV2Header {
+        let recs = records(n, usize::from(rs));
+        let base = 4096u64;
+        let tree = build_btree_v2(params(node_size, rs), &recs, base, os, ls).unwrap();
+        let mut file = vec![0u8; base as usize];
+        file.extend_from_slice(&tree);
+        let hdr = BTreeV2Header::parse(&file, base as usize, os, ls).unwrap();
+        assert_eq!(hdr.total_records, n as u64);
+        let got = collect_btree_v2_records(&file, &hdr, os, ls).unwrap();
+        assert_eq!(got.len(), n);
+        let flat: Vec<u8> = got.into_iter().flat_map(|r| r.data).collect();
+        assert_eq!(flat, recs, "node {node_size} rs {rs} n {n}");
+        hdr
+    }
+
+    #[test]
+    fn one_leaf_then_deeper_trees_read_back_in_order() {
+        // 512-byte nodes of 11-byte records: 45 per leaf, 1149 at depth 1,
+        // 26 449 at depth 2.
+        let info = node_info(512, 11, 8, 3);
+        assert_eq!(
+            info.iter().map(|i| i.cum_max_nrec).collect::<Vec<_>>(),
+            [45, 1149, 26_449, 608_349]
+        );
+        for (n, depth) in [
+            (0, 0),
+            (1, 0),
+            (45, 0),
+            (46, 1),
+            (1149, 1),
+            (1150, 2),
+            (26_449, 2),
+            (26_450, 3),
+            (100_000, 3),
+        ] {
+            let hdr = roundtrip(512, 11, n, 8, 8);
+            assert_eq!(hdr.depth, depth, "{n} records");
+        }
+    }
+
+    #[test]
+    fn pointer_widths_follow_the_offset_and_length_sizes() {
+        for (os, ls) in [(4, 4), (8, 4), (4, 8), (2, 2)] {
+            roundtrip(512, 11, 5000, os, ls);
+        }
+        // Wide counts: a leaf of 2048 bytes / 9-byte records (226, one byte)
+        // and deeper subtree totals of three bytes.
+        roundtrip(2048, 9, 300_000, 8, 8);
+    }
+
+    #[test]
+    fn every_node_is_within_its_capacity_and_above_the_merge_threshold() {
+        let rs = 17u16;
+        let n = 70_000usize;
+        let info = node_info(512, rs, 8, 3);
+        let recs = records(n, usize::from(rs));
+        let tree = build_btree_v2(params(512, rs), &recs, 0, 8, 8).unwrap();
+        let hdr_len = header_size(8, 8);
+        let nodes = (tree.len() - hdr_len) / 512;
+        for i in 0..nodes {
+            let node = &tree[hdr_len + i * 512..hdr_len + (i + 1) * 512];
+            let sig = &node[..4];
+            if sig == b"BTLF" {
+                continue; // counts checked through the parents below
+            }
+            assert_eq!(sig, b"BTIN");
+        }
+        // Walk from the header: each child's count within [40%, 100%].
+        let hdr = BTreeV2Header::parse(&tree, 0, 8, 8).unwrap();
+        assert_eq!(hdr.depth, 3);
+        assert!(u64::from(hdr.num_records_in_root) <= info[3].max_nrec);
+        fn walk(tree: &[u8], addr: usize, nrec: usize, depth: usize, info: &[NodeInfo], rs: usize) {
+            if depth == 0 {
+                return;
+            }
+            let nrec_w = bytes_for_max_records(info[0].max_nrec);
+            let tot_w = if depth > 1 {
+                info[depth - 1].cum_max_nrec_size
+            } else {
+                0
+            };
+            let mut pos = addr + 6 + nrec * rs;
+            for _ in 0..=nrec {
+                let a = u64::from_le_bytes(tree[pos..pos + 8].try_into().unwrap()) as usize;
+                pos += 8;
+                let mut c = 0usize;
+                for b in 0..nrec_w {
+                    c |= usize::from(tree[pos + b]) << (8 * b);
+                }
+                pos += nrec_w + tot_w;
+                let max = info[depth - 1].max_nrec as usize;
+                assert!(c <= max && c * 100 > max * 40, "{c} of {max}");
+                walk(tree, a, c, depth - 1, info, rs);
+            }
+        }
+        walk(
+            &tree,
+            hdr.root_node_address as usize,
+            usize::from(hdr.num_records_in_root),
+            3,
+            &info,
+            usize::from(rs),
+        );
+        assert!(nodes > 0);
+    }
+
+    /// Descending to a key range finds exactly the records a full read
+    /// holds in it — runs of equal keys that straddle node boundaries
+    /// included — at every depth, and nothing for keys not in the tree.
+    #[test]
+    fn a_key_range_search_matches_a_full_scan() {
+        use crate::btree_v2::find_btree_v2_records;
+        use core::cmp::Ordering;
+        let rs = 11usize;
+        // Keys 0, 0, 0, 2, 2, 2, 4, ...: runs of three, odd keys missing.
+        for n in [1usize, 45, 46, 1150, 30_000] {
+            let mut recs = Vec::with_capacity(n * rs);
+            for i in 0..n {
+                let mut r = vec![0u8; rs];
+                r[..8].copy_from_slice(&((i / 3 * 2) as u64).to_be_bytes());
+                r[8..].copy_from_slice(&[(i % 3) as u8, 0, 0]);
+                recs.extend_from_slice(&r);
+            }
+            let base = 4096u64;
+            let tree = build_btree_v2(params(512, 11), &recs, base, 8, 8).unwrap();
+            let mut file = vec![0u8; base as usize];
+            file.extend_from_slice(&tree);
+            let hdr = BTreeV2Header::parse(&file, base as usize, 8, 8).unwrap();
+            let all = collect_btree_v2_records(&file, &hdr, 8, 8).unwrap();
+            let key = |r: &[u8]| u64::from_be_bytes(r[..8].try_into().unwrap());
+            let last = key(&all[n - 1].data);
+            let probes = (0..=last + 1).step_by(if n > 1000 { 37 } else { 1 });
+            for k in probes.chain([last, last + 1, u64::MAX]) {
+                let found =
+                    find_btree_v2_records(&file, &hdr, 8, &mut |r: &[u8]| key(r).cmp(&k)).unwrap();
+                let want: Vec<&[u8]> = all
+                    .iter()
+                    .map(|r| r.data.as_slice())
+                    .filter(|r| key(r) == k)
+                    .collect();
+                let got: Vec<&[u8]> = found.iter().map(|r| r.data.as_slice()).collect();
+                assert_eq!(got, want, "n {n} key {k}");
+                assert_eq!(
+                    got.len(),
+                    if k % 2 == 0 && k <= last {
+                        want.len()
+                    } else {
+                        0
+                    }
+                );
+            }
+            // Every record, or none, when the whole tree is in or out of range.
+            let every = find_btree_v2_records(&file, &hdr, 8, &mut |_| Ordering::Equal).unwrap();
+            assert_eq!(every.len(), n);
+            let none = find_btree_v2_records(&file, &hdr, 8, &mut |_| Ordering::Less).unwrap();
+            assert!(none.is_empty());
+        }
+    }
+
+    /// A two-level tree read through a `read_at`-only storage gives what
+    /// the slice gives — records, descents and errors — whole, truncated
+    /// at every length, and with each byte of its nodes flipped, and each
+    /// node costs one read.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::btree_v2::{
+            collect_btree_v2_records_in, find_btree_v2_records, find_btree_v2_records_in,
+        };
+        use crate::storage::CountingStorage;
+        let (rs, n, base) = (11usize, 120usize, 64usize);
+        let recs = records(n, rs);
+        let tree = build_btree_v2(params(128, 11), &recs, base as u64, 8, 8).unwrap();
+        let mut whole = vec![0u8; base];
+        whole.extend_from_slice(&tree);
+        let hdr = BTreeV2Header::parse(&whole, base, 8, 8).unwrap();
+        assert!(hdr.depth >= 1, "{hdr:?}");
+        let key = |r: &[u8]| u64::from_be_bytes(r[..8].try_into().unwrap());
+        let mut files = Vec::new();
+        for cut in base..=whole.len() {
+            files.push(whole[..cut].to_vec());
+        }
+        for at in base..whole.len() {
+            let mut bad = whole.clone();
+            bad[at] ^= 0x5a;
+            files.push(bad);
+        }
+        let mut ok = 0;
+        for f in &files {
+            let st = CountingStorage::new(f.clone());
+            let want_h = BTreeV2Header::parse(f, base, 8, 8);
+            let got_h = BTreeV2Header::parse_in(&st, base as u64, 8, 8);
+            assert_eq!(format!("{got_h:?}"), format!("{want_h:?}"));
+            // The nodes of the intact header, over each damaged file.
+            let want = collect_btree_v2_records(f, &hdr, 8, 8);
+            st.reset();
+            let got = collect_btree_v2_records_in(&st, &hdr, 8, 8);
+            assert_eq!(format!("{got:?}"), format!("{want:?}"));
+            if want.is_ok() {
+                ok += 1;
+                assert!(st.reads() <= 1 + n as u64 / 3, "{} reads", st.reads());
+            }
+            for k in [0u64, 7, 60, 119, 500] {
+                let want = find_btree_v2_records(f, &hdr, 8, &mut |r: &[u8]| key(r).cmp(&k));
+                let got = find_btree_v2_records_in(&st, &hdr, 8, &mut |r: &[u8]| key(r).cmp(&k));
+                assert_eq!(format!("{got:?}"), format!("{want:?}"));
+            }
+        }
+        assert!(ok > 1);
+    }
+
+    #[test]
+    fn a_node_too_small_or_too_big_is_an_error() {
+        assert!(build_btree_v2(params(16, 11), &records(1, 11), 0, 8, 8).is_err());
+        // A leaf with room for more than 65 535 records.
+        assert!(build_btree_v2(params(1 << 20, 11), &records(1, 11), 0, 8, 8).is_err());
+        // Records that are not whole.
+        assert!(build_btree_v2(params(512, 11), &[0u8; 12], 0, 8, 8).is_err());
+    }
+}

crates/clawhdf5-format/src/checksum.rs

@@ -14,6 +14,45 @@ pub fn jenkins_lookup3(data: &[u8]) -> u32 {
     hashlittle(data, 0)
 }
 
+/// HDF5's Fletcher-32 checksum, as the Fletcher-32 I/O filter (filter id 3)
+/// stores it after each chunk.
+///
+/// A line-for-line port of `H5_checksum_fletcher32` (H5checksum.c, libhdf5
+/// 1.8 through 1.14): big-endian 16-bit words summed in blocks of 360, each
+/// sum reduced after a block by the ones'-complement fold
+/// `(s & 0xffff) + (s >> 16)` rather than `% 65535`, an odd trailing byte
+/// taken as the high byte of a last word, and a final fold of both sums.
+/// The fold and `% 65535` differ whenever a sum is a non-zero multiple of
+/// 65535: the fold leaves 0xffff where the modulo gives 0, so the two
+/// disagree on about one chunk in 32768 and libhdf5 rejects the other's
+/// checksum. This must stay the only implementation.
+pub fn fletcher32(data: &[u8]) -> u32 {
+    let mut sum1: u32 = 0;
+    let mut sum2: u32 = 0;
+    // 360 words keep both sums inside 32 bits between folds (the bound
+    // libhdf5 uses: after a fold sum1 < 0x10200, so sum2 stays below
+    // 360 * 361 / 2 * 0xffff + 360 * 0x10200 + 0x1fffe < 2^32). The adds wrap
+    // like the C unsigned arithmetic all the same.
+    let (words, odd) = data.as_chunks::<2>();
+    for block in words.chunks(360) {
+        for w in block {
+            sum1 = sum1.wrapping_add((u32::from(w[0]) << 8) | u32::from(w[1]));
+            sum2 = sum2.wrapping_add(sum1);
+        }
+        sum1 = (sum1 & 0xffff) + (sum1 >> 16);
+        sum2 = (sum2 & 0xffff) + (sum2 >> 16);
+    }
+    if let [last] = odd {
+        sum1 = sum1.wrapping_add(u32::from(*last) << 8);
+        sum2 = sum2.wrapping_add(sum1);
+        sum1 = (sum1 & 0xffff) + (sum1 >> 16);
+        sum2 = (sum2 & 0xffff) + (sum2 >> 16);
+    }
+    sum1 = (sum1 & 0xffff) + (sum1 >> 16);
+    sum2 = (sum2 & 0xffff) + (sum2 >> 16);
+    (sum2 << 16) | sum1
+}
+
 /// Compute CRC32 (IEEE / ISO 3309) over data.
 ///
 /// When the `fast-checksum` feature is enabled, this uses hardware CRC32
@@ -207,6 +246,19 @@ fn hashlittle(data: &[u8], initval: u32) -> u32 {
 mod tests {
     use super::*;
 
+    /// Values of libhdf5's `H5_checksum_fletcher32` (h5py 3.x's bundled
+    /// libhdf5, called through ctypes). The first three are sums that are
+    /// multiples of 65535, where `% 65535` gave 0 instead of 0xffff.
+    #[test]
+    fn fletcher32_matches_libhdf5() {
+        assert_eq!(fletcher32(&[0x00, 0x01, 0xff, 0xfe]), 0x0001_ffff);
+        assert_eq!(fletcher32(&[0xff; 720]), 0xffff_ffff);
+        assert_eq!(fletcher32(&[0xff; 721]), 0xff00_ff00);
+        assert_eq!(fletcher32(&[0xff; 1441]), 0xff00_ff00);
+        assert_eq!(fletcher32(&[]), 0);
+        assert_eq!(fletcher32(&[7]), 0x0700_0700);
+    }
+
     #[test]
     fn empty_input() {
         // Empty input should return the initial state after no mixing

crates/clawhdf5-format/src/chunk_index.rs

@@ -18,6 +18,7 @@ use alloc::collections::BTreeMap;
 #[cfg(feature = "std")]
 use std::collections::HashMap;
 
+use crate::addr::to_usize;
 use crate::chunk_cache::ChunkCoord;
 use crate::chunked_read::ChunkInfo;
 
@@ -167,7 +168,15 @@ impl ChunkLayout {
 
         for (_coord, ci) in index.iter() {
             let coord: ChunkCoord = ci.offsets.iter().take(rank).copied().collect();
-            let chunk_offsets: Vec<usize> = coord.iter().map(|&o| o as usize).collect();
+            // `ds_dims` are `usize`: a chunk at an offset past `usize::MAX`
+            // (only on a 32-bit target) lies outside the dataset.
+            let Ok(chunk_offsets) = coord
+                .iter()
+                .map(|&o| to_usize(o))
+                .collect::<Result<Vec<usize>, _>>()
+            else {
+                continue;
+            };
 
             let copies = if rank == 0 {
                 // Scalar dataset — single copy

crates/clawhdf5-format/src/chunked_write.rs

@@ -3,9 +3,11 @@
 #[cfg(not(feature = "std"))]
 extern crate alloc;
 
+use crate::addr::saturating_usize;
 #[cfg(not(feature = "std"))]
 use alloc::{format, vec, vec::Vec};
 
+use crate::btree_v2_write::{BTreeV2Params, build_btree_v2};
 use crate::checksum::jenkins_lookup3;
 use crate::chunk_cache::{CACHE_LINE_SIZE, align_to_cache_line};
 use crate::chunk_grid::ChunkGrid;
@@ -16,7 +18,7 @@ use crate::filter_pipeline::{
     FILTER_LZF, FILTER_PCODEC, FILTER_PCODEC_NAME, FILTER_SHUFFLE, FILTER_ZSTD, FilterDescription,
     FilterPipeline,
 };
-use crate::filters::compress_chunk;
+use crate::filters::compress_chunk_masked;
 /// Round a file offset up to the next cache-line boundary.
 ///
 /// This ensures chunk data starts at an address that is a multiple of the
@@ -413,18 +415,18 @@ pub fn split_into_chunks(
     // Dataset strides (row-major)
     let mut ds_strides = vec![1usize; rank];
     for i in (0..rank.saturating_sub(1)).rev() {
-        ds_strides[i] = ds_strides[i + 1] * shape[i + 1] as usize;
+        ds_strides[i] = ds_strides[i + 1] * saturating_usize(shape[i + 1]);
     }
 
     // Chunk strides
     let mut chunk_strides = vec![1usize; rank];
     for i in (0..rank.saturating_sub(1)).rev() {
-        chunk_strides[i] = chunk_strides[i + 1] * chunk_dims[i + 1] as usize;
+        chunk_strides[i] = chunk_strides[i + 1] * saturating_usize(chunk_dims[i + 1]);
     }
 
-    let chunk_total_elements: usize = chunk_dims.iter().map(|&d| d as usize).product();
+    let chunk_total_elements: usize = chunk_dims.iter().map(|&d| saturating_usize(d)).product();
 
-    let mut result = Vec::with_capacity(total_chunks as usize);
+    let mut result = Vec::with_capacity(saturating_usize(total_chunks));
 
     for linear_idx in 0..total_chunks {
         // Convert linear index to chunk grid coordinates
@@ -452,8 +454,8 @@ pub fn split_into_chunks(
                 let coord_in_chunk = remaining_idx / chunk_strides[d];
                 remaining_idx %= chunk_strides[d];
 
-                let global_coord = offsets[d] as usize + coord_in_chunk;
+                let global_coord = saturating_usize(offsets[d]) + coord_in_chunk;
-                if global_coord >= shape[d] as usize {
+                if global_coord >= saturating_usize(shape[d]) {
                     out_of_bounds = true;
                     break;
                 }
@@ -488,7 +490,12 @@ pub fn split_into_chunks(
 #[cfg(feature = "parallel")]
 const PARALLEL_COMPRESS_THRESHOLD: usize = 2;
 
-/// Compress all chunks, using parallel compression when beneficial.
+/// Compress all chunks, using parallel compression when beneficial, and
+/// return each chunk's stored bytes with its filter mask.
+///
+/// Chunks run through the pipeline as libhdf5 runs them
+/// ([`compress_chunk_masked`]): an optional filter that fails — LZF or Blosc
+/// output no smaller than its input — is skipped and its mask bit set.
 ///
 /// With the `parallel` feature and more than [`PARALLEL_COMPRESS_THRESHOLD`]
 /// filtered chunks, compression runs across rayon threads; otherwise it is
@@ -498,7 +505,7 @@ fn compress_all_chunks(
     chunks: &[(Vec<u64>, Vec<u8>)],
     pipeline: &Option<FilterPipeline>,
     element_size: u32,
-) -> Result<Vec<Vec<u8>>, FormatError> {
+) -> Result<Vec<(Vec<u8>, u32)>, FormatError> {
     #[cfg(feature = "parallel")]
     {
         if let Some(pl) = pipeline
@@ -507,7 +514,7 @@ fn compress_all_chunks(
             use rayon::prelude::*;
             return chunks
                 .par_iter()
-                .map(|(_offsets, chunk_bytes)| compress_chunk(chunk_bytes, pl, element_size))
+                .map(|(_offsets, chunk_bytes)| compress_chunk_masked(chunk_bytes, pl, element_size))
                 .collect();
         }
     }
@@ -517,9 +524,9 @@ fn compress_all_chunks(
         .iter()
         .map(|(_offsets, chunk_bytes)| {
             if let Some(pl) = pipeline {
-                compress_chunk(chunk_bytes, pl, element_size)
+                compress_chunk_masked(chunk_bytes, pl, element_size)
             } else {
-                Ok(chunk_bytes.clone())
+                Ok((chunk_bytes.clone(), 0))
             }
         })
         .collect()
@@ -797,8 +804,10 @@ pub fn build_fixed_array_at(
 /// writer passes eliminates the double-compression that the two-pass layout
 /// algorithm previously performed.
 pub struct PrecompressedChunks {
-    /// Per-chunk: (raw_size_bytes, compressed_bytes).
+    /// Per-chunk: (raw_size_bytes, stored_bytes, filter_mask). Bit `i` of
-    pub chunks: Vec<(u64, Vec<u8>)>,
+    /// the mask is set when filter `i` was skipped (an optional filter that
+    /// failed); 0 for every chunk of an unfiltered dataset.
+    pub chunks: Vec<(u64, Vec<u8>, u32)>,
     pub has_filters: bool,
     pub element_size: usize,
     pub shape: Vec<u64>,
@@ -833,7 +842,7 @@ pub fn precompress_chunks(
     let chunks = raw_chunks
         .into_iter()
         .zip(compressed)
-        .map(|((_offsets, raw_bytes), c)| (raw_bytes.len() as u64, c))
+        .map(|((_offsets, raw_bytes), (c, mask))| (raw_bytes.len() as u64, c, mask))
         .collect();
 
     Ok(PrecompressedChunks {
@@ -866,7 +875,7 @@ pub fn build_chunked_data_from_precompressed(
     let mut data_buf = Vec::new();
     let mut written_chunks = Vec::with_capacity(num_chunks);
 
-    for (raw_size, compressed) in &pre.chunks {
+    for (raw_size, compressed, filter_mask) in &pre.chunks {
         let aligned_offset = align_to_cache_line(data_buf.len());
         if aligned_offset > data_buf.len() {
             data_buf.resize(aligned_offset, 0u8);
@@ -878,7 +887,7 @@ pub fn build_chunked_data_from_precompressed(
             address,
             compressed_size,
             raw_size: *raw_size,
-            filter_mask: 0,
+            filter_mask: *filter_mask,
         });
     }
 
@@ -915,7 +924,7 @@ pub fn build_chunked_data_from_precompressed(
             } else {
                 None
             };
-            let filter_mask = if pre.has_filters { Some(0u32) } else { None };
+            let filter_mask = pre.has_filters.then_some(written_chunks[0].filter_mask);
             serialize_v4_single_chunk(
                 &chunk_dims_u32,
                 chunk_addr,
@@ -1035,7 +1044,7 @@ impl ChunkIndexPlan {
                     Ok(Self::SingleChunk)
                 } else {
                     let grid = ChunkGrid::fixed_array(shape, Some(max), chunk_dims)?;
-                    Ok(Self::FixedArray(grid, nslots as usize))
+                    Ok(Self::FixedArray(grid, saturating_usize(nslots)))
                 }
             }
             1 => Ok(Self::ExtensibleArray(ChunkGrid::extensible_array(
@@ -1105,11 +1114,12 @@ const BT2_CHUNK_FILTERED: u8 = 11;
 ///
 /// `records` are `(scaled coordinates, chunk)` in lexicographic order of the
 /// coordinates, which is the order the library's comparator
-/// (`H5VM_vector_cmp_u`) keeps them in. The tree is a single leaf: the
+/// (`H5VM_vector_cmp_u`) keeps them in. Up to 65 535 chunks go in a single
-/// library's 2048-byte node when the records fit, otherwise a leaf node
+/// leaf: the library's 2048-byte node when the records fit, otherwise a leaf
-/// sized to hold them all (the root's record count is 16-bit, so at most
+/// node sized to hold them all (the layout the writer has always used, kept
-/// 65535 chunks). Returns the bytes and the node size the layout message
+/// so those files do not change). More chunks get the library's 2048-byte
-/// must record.
+/// nodes with internal nodes above the leaves. Returns the bytes and the
+/// node size the layout message must record.
 fn build_btree_v2_chunk_index_at(
     rank: usize,
     records: &[(Vec<u64>, &WrittenChunk)],
@@ -1119,73 +1129,49 @@ fn build_btree_v2_chunk_index_at(
     base_address: u64,
 ) -> Result<(Vec<u8>, u32), FormatError> {
     let os = offset_size as usize;
-    let nrec = u16::try_from(records.len()).map_err(|_| {
-        FormatError::ChunkedReadError(
-            "more than 65535 chunks with more than one unlimited dimension: \
-             use larger chunks"
-                .into(),
-        )
-    })?;
     let chunk_size_bytes = has_filters.then(|| {
         let slots: Vec<Option<WrittenChunk>> =
             records.iter().map(|(_, c)| Some((*c).clone())).collect();
         filtered_chunk_size_len(&slots)
     });
     let record_size = os + chunk_size_bytes.map_or(0, |n| n + 4) + 8 * rank;
+    let record_size_u16 = u16::try_from(record_size)
+        .map_err(|_| FormatError::Overflow("B-tree v2 record size".into()))?;
+    let node_size = if records.len() <= usize::from(u16::MAX) {
         // Leaf: signature, version, type, records, checksum.
         let leaf_len = 4 + 1 + 1 + records.len() * record_size + 4;
-    let node_size = u32::try_from(leaf_len)
+        u32::try_from(leaf_len)
             .map_err(|_| FormatError::Overflow("B-tree v2 leaf size".into()))?
-        .max(BT2_NODE_SIZE);
+            .max(BT2_NODE_SIZE)
+    } else {
+        BT2_NODE_SIZE
+    };
     let tree_type = if has_filters {
         BT2_CHUNK_FILTERED
     } else {
         BT2_CHUNK_UNFILTERED
     };
 
-    let hdr_len = 4 + 1 + 1 + 4 + 2 + 2 + 1 + 1 + os + 2 + length_size as usize + 4;
+    let mut flat = Vec::with_capacity(records.len() * record_size);
-    let leaf_address = base_address + hdr_len as u64;
-
-    let mut out = Vec::with_capacity(hdr_len + node_size as usize);
-    out.extend_from_slice(b"BTHD");
-    out.push(0); // version
-    out.push(tree_type);
-    out.extend_from_slice(&node_size.to_le_bytes());
-    out.extend_from_slice(&(record_size as u16).to_le_bytes());
-    out.extend_from_slice(&0u16.to_le_bytes()); // depth
-    out.push(BT2_SPLIT_PERCENT);
-    out.push(BT2_MERGE_PERCENT);
-    if records.is_empty() {
-        out.extend(core::iter::repeat_n(0xFF, os));
-    } else {
-        push_addr(&mut out, leaf_address, offset_size);
-    }
-    out.extend_from_slice(&nrec.to_le_bytes());
-    match length_size {
-        4 => out.extend_from_slice(&(records.len() as u32).to_le_bytes()),
-        _ => out.extend_from_slice(&(records.len() as u64).to_le_bytes()),
-    }
-    let sum = jenkins_lookup3(&out);
-    out.extend_from_slice(&sum.to_le_bytes());
-    debug_assert_eq!(out.len(), hdr_len);
-    if records.is_empty() {
-        return Ok((out, node_size));
-    }
-
-    let leaf_start = out.len();
-    out.extend_from_slice(b"BTLF");
-    out.push(0); // version
-    out.push(tree_type);
     for (scaled, chunk) in records {
-        push_index_element(&mut out, Some(chunk), offset_size, chunk_size_bytes);
+        push_index_element(&mut flat, Some(chunk), offset_size, chunk_size_bytes);
         for &c in scaled {
-            out.extend_from_slice(&c.to_le_bytes());
+            flat.extend_from_slice(&c.to_le_bytes());
         }
     }
-    let sum = jenkins_lookup3(&out[leaf_start..]);
+    let out = build_btree_v2(
-    out.extend_from_slice(&sum.to_le_bytes());
+        BTreeV2Params {
-    // The library reads whole nodes; pad the leaf out to the node size.
+            tree_type,
-    out.resize(leaf_start + node_size as usize, 0);
+            node_size,
+            record_size: record_size_u16,
+            split_percent: BT2_SPLIT_PERCENT,
+            merge_percent: BT2_MERGE_PERCENT,
+        },
+        &flat,
+        base_address,
+        offset_size,
+        length_size,
+    )?;
     Ok((out, node_size))
 }
 
@@ -1273,7 +1259,7 @@ pub fn write_selection_to_buffer(
             let rank = dims.len();
             let mut ds_strides = vec![1usize; rank];
             for i in (0..rank.saturating_sub(1)).rev() {
-                ds_strides[i] = ds_strides[i + 1] * dims[i + 1] as usize;
+                ds_strides[i] = ds_strides[i + 1] * saturating_usize(dims[i + 1]);
             }
 
             let mut src_offset = 0usize;
@@ -1323,7 +1309,7 @@ pub fn write_selection_to_buffer(
                                 buffer,
                                 new_data,
                                 src_offset,
-                                current_ds_offset + coord as usize * ds_strides[d],
+                                current_ds_offset + saturating_usize(coord) * ds_strides[d],
                             );
                         }
                     }
@@ -1350,14 +1336,14 @@ pub fn write_selection_to_buffer(
             let rank = dims.len();
             let mut ds_strides = vec![1usize; rank];
             for i in (0..rank.saturating_sub(1)).rev() {
-                ds_strides[i] = ds_strides[i + 1] * dims[i + 1] as usize;
+                ds_strides[i] = ds_strides[i + 1] * saturating_usize(dims[i + 1]);
             }
 
             for (pi, pt) in pts.iter().enumerate() {
                 let flat: usize = pt
                     .iter()
                     .zip(ds_strides.iter())
-                    .map(|(&p, &s)| p as usize * s)
+                    .map(|(&p, &s)| saturating_usize(p) * s)
                     .sum();
                 let dst = flat * elem_size;
                 let src = pi * elem_size;
@@ -1965,6 +1951,98 @@ mod tests {
         bytes_to_f64(&output)
     }
 
+    /// Every chunk index the writer builds records each chunk's real filter
+    /// mask: LZF output no smaller than the chunk is skipped (bit 1, behind
+    /// shuffle) and the chunk stored shuffled only; compressible chunks keep
+    /// mask 0. The data reads back through both kinds of chunk.
+    #[cfg(feature = "lzf")]
+    #[test]
+    fn skipped_lzf_chunks_are_masked_in_every_index() {
+        let c = 64usize;
+        // Chunks alternate: random bytes (LZF cannot shrink them), then 7s.
+        let mut state = 0x1234_5678_u64;
+        let data: Vec<f64> = (0..4 * c)
+            .map(|i| {
+                if (i / c).is_multiple_of(2) {
+                    state ^= state << 13;
+                    state ^= state >> 7;
+                    state ^= state << 17;
+                    f64::from_bits(state)
+                } else {
+                    7.0
+                }
+            })
+            .collect();
+        let raw = f64_to_bytes(&data);
+        let options = ChunkOptions {
+            plugin: Some(PluginFilter::Lzf),
+            ..Default::default()
+        };
+        let c64 = c as u64;
+        #[allow(clippy::type_complexity)]
+        let cases: [(&[u64], &[u64], Option<&[u64]>, u8, &[u32]); 4] = [
+            (&[c64], &[c64], None, 1, &[2]),
+            (&[4 * c64], &[c64], None, 3, &[2, 0, 2, 0]),
+            (&[4 * c64], &[c64], Some(&[u64::MAX]), 4, &[2, 0, 2, 0]),
+            (
+                &[2, 2 * c64],
+                &[1, c64],
+                Some(&[u64::MAX, u64::MAX]),
+                5,
+                &[2, 0, 2, 0],
+            ),
+        ];
+        let base = 0x1000u64;
+        for (shape, chunks, maxshape, index_type, want_masks) in cases {
+            let n: u64 = shape.iter().product();
+            let raw = &raw[..n as usize * 8];
+            let result =
+                build_chunked_data_at_ext(raw, shape, chunks, 8, &options, base, maxshape).unwrap();
+            let mut file = vec![0u8; base as usize];
+            file.extend_from_slice(&result.data_bytes);
+            let layout = DataLayout::parse(&result.layout_message, 8, 8).unwrap();
+            assert!(
+                matches!(&layout, DataLayout::Chunked { chunk_index_type, .. }
+                    if *chunk_index_type == Some(index_type)),
+                "{layout:?}"
+            );
+            let dataspace = Dataspace {
+                space_type: DataspaceType::Simple,
+                rank: shape.len() as u8,
+                dimensions: shape.to_vec(),
+                max_dimensions: maxshape.map(<[u64]>::to_vec),
+            };
+            let (mut infos, _) =
+                crate::chunked_read::list_chunks(&file, &layout, &dataspace, 8, 8, 8).unwrap();
+            infos.sort_by(|a, b| a.offsets.cmp(&b.offsets));
+            let masks: Vec<u32> = infos.iter().map(|i| i.filter_mask).collect();
+            assert_eq!(masks, want_masks, "index type {index_type}");
+            for info in &infos {
+                // Skipped chunks are stored at the chunk's size (shuffled).
+                assert_eq!(
+                    info.chunk_size == (c * 8) as u32,
+                    info.filter_mask != 0,
+                    "{info:?}"
+                );
+            }
+            let pipeline = crate::filter_pipeline::FilterPipeline::parse(
+                result.pipeline_message.as_ref().unwrap(),
+            )
+            .unwrap();
+            let out = read_chunked_data(
+                &file,
+                &layout,
+                &dataspace,
+                &make_f64_type(),
+                Some(&pipeline),
+                8,
+                8,
+            )
+            .unwrap();
+            assert_eq!(out, raw, "index type {index_type}");
+        }
+    }
+
     #[test]
     fn ea_roundtrip_1d_inline_only() {
         let values: Vec<f64> = (0..10).map(|i| i as f64).collect();

crates/clawhdf5-format/src/data_layout.rs

@@ -6,7 +6,9 @@ use alloc::{format, string::String, vec::Vec};
 #[cfg(feature = "std")]
 use std::string::String;
 
+use crate::addr::to_usize;
 use crate::error::FormatError;
+use crate::storage::Storage;
 
 /// A single VDS (Virtual Dataset) source mapping.
 ///
@@ -207,7 +209,7 @@ pub fn parse_vds_mappings(
                     "VDS mapping shares a name with a later entry".into(),
                 ));
             }
-            Ok(idx as usize)
+            to_usize(idx)
         };
 
         let source_file = if flags & VDS_SOURCE_SAME_FILE != 0 {
@@ -309,6 +311,16 @@ impl DataLayout {
         &mut self,
         file_data: &[u8],
         length_size: u8,
+    ) -> Result<(), FormatError> {
+        self.resolve_vds_mappings_in(file_data, length_size)
+    }
+
+    /// [`Self::resolve_vds_mappings`] over any [`Storage`]: one read of the
+    /// global heap collection holding the mappings.
+    pub fn resolve_vds_mappings_in<S: Storage + ?Sized>(
+        &mut self,
+        file_data: &S,
+        length_size: u8,
     ) -> Result<(), FormatError> {
         if let DataLayout::Virtual {
             global_heap_address,
@@ -318,11 +330,8 @@ impl DataLayout {
         } = self
             && let Some(addr) = *global_heap_address
         {
-            let coll = crate::global_heap::GlobalHeapCollection::parse(
+            let coll =
-                file_data,
+                crate::global_heap::GlobalHeapCollection::parse_in(file_data, addr, length_size)?;
-                addr as usize,
-                length_size,
-            )?;
             let obj = coll.get_object(*global_heap_index as u16).ok_or(
                 FormatError::GlobalHeapObjectNotFound {
                     collection_address: addr,
@@ -1305,4 +1314,44 @@ mod tests {
         let blob = [0x01u8, 0, 0, 0, 0, 0, 0, 0, 0];
         assert!(parse_vds_mappings(&blob, 8).unwrap().is_empty());
     }
+
+    /// A virtual dataset's mappings resolve identically through a
+    /// read_at-only CountingStorage, in two reads of the global heap.
+    #[test]
+    fn vds_mappings_through_storage_match_slice() {
+        use crate::message_type::MessageType;
+        use crate::object_header::ObjectHeader;
+        use crate::storage::CountingStorage;
+        let file: &[u8] = include_bytes!("../tests/fixtures/vds_same_file.h5");
+        let sb = crate::superblock::Superblock::parse(file, 0).unwrap();
+        let (os, ls) = (sb.offset_size, sb.length_size);
+        let storage = CountingStorage::new(file.to_vec());
+        let mut virtuals = 0;
+        for child in
+            crate::group_v2::resolve_group_children(file, &sb, sb.root_group_address).unwrap()
+        {
+            let h =
+                ObjectHeader::parse(file, child.object_header_address as usize, os, ls).unwrap();
+            let Some(msg) = h
+                .messages
+                .iter()
+                .find(|m| m.msg_type == MessageType::DataLayout)
+            else {
+                continue;
+            };
+            let mut want = DataLayout::parse(&msg.data, os, ls).unwrap();
+            if !matches!(want, DataLayout::Virtual { .. }) {
+                continue;
+            }
+            let mut got = want.clone();
+            want.resolve_vds_mappings(file, ls).unwrap();
+            storage.reset();
+            got.resolve_vds_mappings_in(&storage, ls).unwrap();
+            assert_eq!(format!("{got:?}"), format!("{want:?}"));
+            assert!(matches!(&got, DataLayout::Virtual { mappings, .. } if !mappings.is_empty()));
+            assert_eq!(storage.reads(), 2);
+            virtuals += 1;
+        }
+        assert!(virtuals >= 1);
+    }
 }

crates/clawhdf5-format/src/data_read.rs

@@ -1,21 +1,25 @@
 //! Raw data reading and typed conversion for HDF5 datasets.
 
 #[cfg(not(feature = "std"))]
-use alloc::{collections::BTreeMap, format, string::String, vec, vec::Vec};
+use alloc::{borrow::Cow, collections::BTreeMap, format, string::String, vec, vec::Vec};
+#[cfg(feature = "std")]
+use std::borrow::Cow;
 
 #[cfg(feature = "std")]
 use std::collections::BTreeMap;
 
+use crate::addr::to_usize;
 #[cfg(feature = "std")]
 use crate::chunk_cache::ChunkCache;
-use crate::chunked_read::read_chunked_data;
+use crate::chunked_read::read_chunked_data_in;
 #[cfg(feature = "std")]
-use crate::chunked_read::{read_chunked_data_cached, read_chunked_data_indexed};
+use crate::chunked_read::{read_chunked_data_cached_in, read_chunked_data_indexed_in};
 use crate::data_layout::DataLayout;
 use crate::dataspace::Dataspace;
 use crate::datatype::{Datatype, DatatypeByteOrder};
 use crate::error::FormatError;
 use crate::filter_pipeline::FilterPipeline;
+use crate::storage::{Storage, read_exact_at};
 
 /// Checks that `[offset, offset + needed)` fits within `data`, guarding the
 /// addition against `usize` overflow from a crafted near-`usize::MAX` offset.
@@ -32,6 +36,80 @@ fn ensure_len(data: &[u8], offset: usize, needed: usize) -> Result<(), FormatErr
     Ok(())
 }
 
+/// The storage checks libhdf5 makes when it opens a dataset, before any
+/// data is read (`H5D__contig_check`, `H5D__compact_init`), so a dataset
+/// they refuse fails to open, as in libhdf5, instead of opening and
+/// reporting a shape nothing can be read from:
+///
+/// - the element count times the element size must not overflow 64 bits
+///   ("size of dataset's storage overflowed" — `cve-2024-32624`
+///   `/Dset_OBJREF`, 2^62 references of 8 bytes);
+/// - contiguous storage at a defined address must end within the file's
+///   `file_len` bytes (the HDF5 data up to the end of file the superblock
+///   records);
+/// - compact data must be exactly the dataset's size.
+///
+/// Deliberately not refused, unlike libhdf5: an empty contiguous dataset at
+/// a defined address (libhdf5's overflow test `addr + 0 <= addr` refuses
+/// it), which clawhdf5 up to v2.7.0 wrote. Chunked and virtual layouts are
+/// checked when their data is read.
+pub fn check_dataset_storage(
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    file_len: u64,
+) -> Result<(), FormatError> {
+    if !matches!(
+        layout,
+        DataLayout::Contiguous { .. } | DataLayout::Compact { .. }
+    ) {
+        return Ok(());
+    }
+    const OVERFLOWED: &str = "size of dataset's storage overflowed";
+    let n = dataspace
+        .checked_num_elements()
+        .map_err(|_| FormatError::InvalidDatasetStorage(OVERFLOWED))?;
+    let data_size = n
+        .checked_mul(u64::from(datatype.type_size()))
+        .ok_or(FormatError::InvalidDatasetStorage(OVERFLOWED))?;
+    match layout {
+        DataLayout::Contiguous {
+            address: Some(address),
+            ..
+        } if address
+            .checked_add(data_size)
+            .is_none_or(|end| end > file_len) =>
+        {
+            Err(FormatError::InvalidDatasetStorage(
+                "invalid dataset size, likely file corruption",
+            ))
+        }
+        DataLayout::Compact { data } if data.len() as u64 != data_size => {
+            Err(FormatError::InvalidDatasetStorage(
+                "bad value from dataset header - size of compact dataset's data buffer \
+                 doesn't match size of dataset data",
+            ))
+        }
+        _ => Ok(()),
+    }
+}
+
+/// How many bytes to read from a contiguous dataset's storage of
+/// `storage_size` bytes (the layout message's size) holding `needed` bytes
+/// of elements. libhdf5 reads the elements' bytes from the start of the
+/// storage and ignores storage past them (`H5D__contig_check` checks only
+/// that the elements fit in the file), so a larger storage reads; one too
+/// small to hold the elements is an error.
+pub fn contiguous_read_len(storage_size: u64, needed: usize) -> Result<usize, FormatError> {
+    if storage_size < needed as u64 {
+        return Err(FormatError::DataSizeMismatch {
+            expected: needed,
+            actual: usize::try_from(storage_size).unwrap_or(usize::MAX),
+        });
+    }
+    Ok(needed)
+}
+
 /// Zero-copy read of contiguous raw data, returning a borrowed slice.
 ///
 /// For contiguous layouts, returns a direct `&[u8]` slice into `file_data`.
@@ -43,7 +121,7 @@ pub fn read_raw_data_zerocopy<'a>(
     dataspace: &Dataspace,
     datatype: &Datatype,
 ) -> Result<Option<&'a [u8]>, FormatError> {
-    let num_elements = dataspace.num_elements() as usize;
+    let num_elements = to_usize(dataspace.num_elements())?;
     let elem_size = datatype.type_size() as usize;
     let expected_size = num_elements.checked_mul(elem_size).ok_or_else(|| {
         FormatError::Overflow(format!(
@@ -54,14 +132,8 @@ pub fn read_raw_data_zerocopy<'a>(
     match layout {
         DataLayout::Contiguous { address, size } => {
             let addr = address.ok_or(FormatError::NoDataAllocated)?;
-            let addr = addr as usize;
+            let addr = to_usize(addr)?;
-            let sz = *size as usize;
+            let sz = contiguous_read_len(*size, expected_size)?;
-            if sz != expected_size {
-                return Err(FormatError::DataSizeMismatch {
-                    expected: expected_size,
-                    actual: sz,
-                });
-            }
             ensure_len(file_data, addr, sz)?;
             Ok(Some(&file_data[addr..addr + sz]))
         }
@@ -80,7 +152,17 @@ pub fn read_raw_data(
     dataspace: &Dataspace,
     datatype: &Datatype,
 ) -> Result<Vec<u8>, FormatError> {
-    read_raw_data_full(file_data, layout, dataspace, datatype, None, 8, 8)
+    read_raw_data_in(file_data, layout, dataspace, datatype)
+}
+
+/// [`read_raw_data`] over any [`Storage`].
+pub fn read_raw_data_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+) -> Result<Vec<u8>, FormatError> {
+    read_raw_data_full_in(file_data, layout, dataspace, datatype, None, 8, 8)
 }
 
 /// Resolves a Virtual Dataset source **file name** (as stored in the mapping,
@@ -102,6 +184,27 @@ pub fn read_raw_data_full(
     pipeline: Option<&FilterPipeline>,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<u8>, FormatError> {
+    read_raw_data_full_in(
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        pipeline,
+        offset_size,
+        length_size,
+    )
+}
+
+/// [`read_raw_data_full`] over any [`Storage`].
+pub fn read_raw_data_full_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<u8>, FormatError> {
     read_raw_data_full_impl(
         file_data,
@@ -127,6 +230,30 @@ pub fn read_raw_data_full_with_resolver(
     offset_size: u8,
     length_size: u8,
     resolver: Option<&VdsSourceResolver>,
+) -> Result<Vec<u8>, FormatError> {
+    read_raw_data_full_with_resolver_in(
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        pipeline,
+        offset_size,
+        length_size,
+        resolver,
+    )
+}
+
+/// [`read_raw_data_full_with_resolver`] over any [`Storage`].
+#[allow(clippy::too_many_arguments)]
+pub fn read_raw_data_full_with_resolver_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
+    resolver: Option<&VdsSourceResolver>,
 ) -> Result<Vec<u8>, FormatError> {
     read_raw_data_full_impl(
         file_data,
@@ -141,8 +268,8 @@ pub fn read_raw_data_full_with_resolver(
 }
 
 #[allow(clippy::too_many_arguments)]
-fn read_raw_data_full_impl(
+fn read_raw_data_full_impl<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     layout: &DataLayout,
     dataspace: &Dataspace,
     datatype: &Datatype,
@@ -151,7 +278,7 @@ fn read_raw_data_full_impl(
     length_size: u8,
     resolver: Option<&VdsSourceResolver>,
 ) -> Result<Vec<u8>, FormatError> {
-    let num_elements = dataspace.num_elements() as usize;
+    let num_elements = to_usize(dataspace.num_elements())?;
     let elem_size = datatype.type_size() as usize;
     let expected_size = num_elements.checked_mul(elem_size).ok_or_else(|| {
         FormatError::Overflow(format!(
@@ -171,20 +298,19 @@ fn read_raw_data_full_impl(
         }
         DataLayout::Contiguous { address, size } => {
             let addr = address.ok_or(FormatError::NoDataAllocated)?;
-            let addr = addr as usize;
+            let addr = to_usize(addr)?;
-            let sz = *size as usize;
+            let sz = contiguous_read_len(*size, expected_size)?;
-            if sz != expected_size {
+            match read_exact_at(file_data, addr as u64, sz)? {
-                return Err(FormatError::DataSizeMismatch {
+                Cow::Borrowed(bytes) => {
-                    expected: expected_size,
-                    actual: sz,
-                });
-            }
-            ensure_len(file_data, addr, sz)?;
                     let mut out = crate::bulk_alloc::vec_for_bulk(sz);
-            out.extend_from_slice(&file_data[addr..addr + sz]);
+                    out.extend_from_slice(bytes);
                     Ok(out)
                 }
-        DataLayout::Chunked { .. } => read_chunked_data(
+                // Fetched for this read: already the caller's copy.
+                Cow::Owned(out) => Ok(out),
+            }
+        }
+        DataLayout::Chunked { .. } => read_chunked_data_in(
             file_data,
             layout,
             dataspace,
@@ -221,9 +347,34 @@ pub fn read_raw_data_cached(
     offset_size: u8,
     length_size: u8,
     cache: &ChunkCache,
+) -> Result<Vec<u8>, FormatError> {
+    read_raw_data_cached_in(
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        pipeline,
+        offset_size,
+        length_size,
+        cache,
+    )
+}
+
+/// [`read_raw_data_cached`] over any [`Storage`].
+#[cfg(feature = "std")]
+#[allow(clippy::too_many_arguments)]
+pub fn read_raw_data_cached_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
+    cache: &ChunkCache,
 ) -> Result<Vec<u8>, FormatError> {
     match layout {
-        DataLayout::Chunked { .. } => read_chunked_data_cached(
+        DataLayout::Chunked { .. } => read_chunked_data_cached_in(
             file_data,
             layout,
             dataspace,
@@ -233,7 +384,7 @@ pub fn read_raw_data_cached(
             length_size,
             cache,
         ),
-        _ => read_raw_data_full(
+        _ => read_raw_data_full_in(
             file_data,
             layout,
             dataspace,
@@ -262,9 +413,34 @@ pub fn read_raw_data_indexed(
     offset_size: u8,
     length_size: u8,
     cache: &ChunkCache,
+) -> Result<Vec<u8>, FormatError> {
+    read_raw_data_indexed_in(
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        pipeline,
+        offset_size,
+        length_size,
+        cache,
+    )
+}
+
+/// [`read_raw_data_indexed`] over any [`Storage`].
+#[cfg(feature = "std")]
+#[allow(clippy::too_many_arguments)]
+pub fn read_raw_data_indexed_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
+    cache: &ChunkCache,
 ) -> Result<Vec<u8>, FormatError> {
     match layout {
-        DataLayout::Chunked { .. } => read_chunked_data_indexed(
+        DataLayout::Chunked { .. } => read_chunked_data_indexed_in(
             file_data,
             layout,
             dataspace,
@@ -274,7 +450,7 @@ pub fn read_raw_data_indexed(
             length_size,
             cache,
         ),
-        _ => read_raw_data_full(
+        _ => read_raw_data_full_in(
             file_data,
             layout,
             dataspace,
@@ -303,6 +479,30 @@ pub fn read_raw_data_selection(
     offset_size: u8,
     length_size: u8,
     selection: &crate::selection::Selection,
+) -> Result<Vec<u8>, FormatError> {
+    read_raw_data_selection_in(
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        pipeline,
+        offset_size,
+        length_size,
+        selection,
+    )
+}
+
+/// [`read_raw_data_selection`] over any [`Storage`].
+#[allow(clippy::too_many_arguments)]
+pub fn read_raw_data_selection_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
+    selection: &crate::selection::Selection,
 ) -> Result<Vec<u8>, FormatError> {
     use crate::selection::Selection;
 
@@ -312,7 +512,7 @@ pub fn read_raw_data_selection(
     // Read only what the selection's bounding box touches when that is
     // possible; everything below is the decode-everything-then-pick path,
     // kept for the cases `partial_read` declines.
-    if let Some(selected) = crate::partial_read::read_selection(
+    if let Some(selected) = crate::partial_read::read_selection_in(
         file_data,
         layout,
         dataspace,
@@ -327,7 +527,7 @@ pub fn read_raw_data_selection(
 
     match selection {
         Selection::All => {
-            return read_raw_data_full(
+            return read_raw_data_full_in(
                 file_data,
                 layout,
                 dataspace,
@@ -347,7 +547,7 @@ pub fn read_raw_data_selection(
     match layout {
         DataLayout::Compact { .. } | DataLayout::Contiguous { .. } => {
             // Read all data, then extract the selection
-            let full_data = read_raw_data_full(
+            let full_data = read_raw_data_full_in(
                 file_data,
                 layout,
                 dataspace,
@@ -371,7 +571,7 @@ pub fn read_raw_data_selection(
             // implicit-index generator, which then indexed past the rank and
             // panicked — only to decode the full dataset anyway.
             crate::chunked_read::chunk_geometry(chunk_dimensions, *version, dataspace, elem_size)?;
-            let full_data = read_raw_data_full(
+            let full_data = read_raw_data_full_in(
                 file_data,
                 layout,
                 dataspace,
@@ -384,7 +584,7 @@ pub fn read_raw_data_selection(
         }
         DataLayout::Virtual { .. } => {
             // Assemble the full virtual dataset, then apply the read selection.
-            let full_data = read_raw_data_full(
+            let full_data = read_raw_data_full_in(
                 file_data,
                 layout,
                 dataspace,
@@ -408,8 +608,8 @@ pub fn read_raw_data_selection(
 /// would report differently from the stored dataspace (unlimited mappings).
 /// Use [`crate::vds::read_virtual_dataset`] to read those.
 #[allow(clippy::too_many_arguments)]
-fn read_virtual_data(
+fn read_virtual_data<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     layout: &DataLayout,
     dataspace: &Dataspace,
     datatype: &Datatype,
@@ -420,7 +620,7 @@ fn read_virtual_data(
     let wrapped =
         resolver.map(|r| move |name: &str| -> Result<Option<Vec<u8>>, FormatError> { Ok(r(name)) });
     let wrapped_ref = wrapped.as_ref().map(|w| w as &crate::vds::VdsFileResolver);
-    let v = crate::vds::read_virtual_dataset(
+    let v = crate::vds::read_virtual_dataset_in(
         file_data,
         layout,
         dataspace,
@@ -520,7 +720,7 @@ pub fn extract_selection_from_buffer(
             let rank = dims.len();
             let mut ds_strides = vec![1usize; rank];
             for i in (0..rank.saturating_sub(1)).rev() {
-                ds_strides[i] = ds_strides[i + 1] * dims[i + 1] as usize;
+                ds_strides[i] = ds_strides[i + 1] * to_usize(dims[i + 1])?;
             }
 
             let mut output = Vec::with_capacity(pts.len() * elem_size);
@@ -528,8 +728,8 @@ pub fn extract_selection_from_buffer(
                 let flat: usize = pt
                     .iter()
                     .zip(ds_strides.iter())
-                    .map(|(&p, &s)| p as usize * s)
+                    .map(|(&p, &s)| Ok(to_usize(p)? * s))
-                    .sum();
+                    .sum::<Result<usize, FormatError>>()?;
                 let src = flat * elem_size;
                 if src + elem_size <= full_data.len() {
                     output.extend_from_slice(&full_data[src..src + elem_size]);
@@ -756,6 +956,220 @@ pub fn read_selection_native<T: NativeElement>(
     crate::gather::gather::<T>(raw, dims, elem_size, selection).map(Some)
 }
 
+/// [`read_selection_native`] of a contiguous dataset in any [`Storage`],
+/// reading only the selected elements' runs (adjacent ones merged, one
+/// [`Storage::read_ranges`] call) instead of the whole dataset.
+///
+/// `Ok(None)` wherever the in-memory fast path does not apply and the
+/// caller converts through the byte readers instead: `datatype` is not
+/// `T`'s native representation, the layout is not contiguous, or the
+/// dataset's bytes cannot be located in the file (no address, storage too
+/// small, past the end of file: the cases [`read_raw_data_zerocopy`]
+/// fails). Otherwise the result and errors are [`read_selection_native`]'s
+/// over those bytes.
+pub fn read_selection_native_in<T: NativeElement, S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    selection: &crate::selection::Selection,
+) -> Result<Option<Vec<T>>, FormatError> {
+    if !T::is_native(datatype) {
+        return Ok(None);
+    }
+    let DataLayout::Contiguous {
+        address: Some(address),
+        size,
+    } = layout
+    else {
+        return Ok(None);
+    };
+    // Where the dataset's bytes are, as `read_raw_data_zerocopy` finds them.
+    let located = to_usize(dataspace.num_elements())
+        .ok()
+        .and_then(|n| n.checked_mul(datatype.type_size() as usize))
+        .filter(|&len| contiguous_read_len(*size, len).is_ok())
+        .filter(|&len| {
+            address
+                .checked_add(len as u64)
+                .is_some_and(|end| end <= file_data.len())
+        });
+    let Some(len) = located else {
+        return Ok(None);
+    };
+    if let Some(all) = file_data.as_contiguous() {
+        let start = to_usize(*address)?;
+        return read_selection_native(
+            &all[start..start + len],
+            &dataspace.dimensions,
+            datatype,
+            selection,
+        );
+    }
+    let dims = &dataspace.dimensions;
+    let elem_size = core::mem::size_of::<T>();
+    let total = dims
+        .iter()
+        .try_fold(1u64, |acc, &d| acc.checked_mul(d))
+        .ok_or_else(|| FormatError::Overflow("dataset shape overflows".into()))?;
+    let expected = crate::chunked_read::checked_byte_len(total, elem_size)?;
+    if len != expected {
+        return Err(FormatError::DataSizeMismatch {
+            expected,
+            actual: len,
+        });
+    }
+    let bytes = if let crate::selection::Selection::All = selection {
+        read_exact_at(file_data, *address, len)?.into_owned()
+    } else {
+        crate::partial_read::validate(selection, dims)?;
+        crate::gather::gather_storage(file_data, *address, len, dims, elem_size, selection)?
+    };
+    Ok(Some(native_to_vec(&bytes, bytes.len() / elem_size)))
+}
+
+/// The bytes of a slice of [`NativeElement`]s.
+#[cfg(feature = "std")]
+fn bytes_of_mut<T: NativeElement>(values: &mut [T]) -> &mut [u8] {
+    // SAFETY: `T: NativeElement` has no padding and every bit pattern is a
+    // valid value, so its storage may be viewed, and written, as bytes; the
+    // byte slice covers exactly the values' storage and borrows it
+    // exclusively for its lifetime.
+    unsafe {
+        core::slice::from_raw_parts_mut(
+            values.as_mut_ptr().cast::<u8>(),
+            core::mem::size_of_val(values),
+        )
+    }
+}
+
+/// `count` zeroed values of `T`, from zeroed pages where the allocator can
+/// (see [`crate::chunked_read::alloc_output`]) and backed by huge pages when
+/// large. A size taken from the file surfaces as an error, not an abort.
+#[cfg(feature = "std")]
+fn alloc_zeroed_values<T: NativeElement>(count: usize) -> Result<Vec<T>, FormatError> {
+    if count == 0 || core::mem::size_of::<T>() == 0 {
+        return Ok(Vec::new());
+    }
+    let failed = || {
+        FormatError::Overflow(format!(
+            "cannot allocate {count} values of {} bytes for dataset output",
+            core::mem::size_of::<T>()
+        ))
+    };
+    let layout = core::alloc::Layout::array::<T>(count).map_err(|_| failed())?;
+    // SAFETY: `layout` has non-zero size (count > 0, T not zero-sized).
+    let ptr = unsafe { std::alloc::alloc_zeroed(layout) };
+    if ptr.is_null() {
+        return Err(failed());
+    }
+    crate::bulk_alloc::advise_huge_pages(ptr, layout.size());
+    // SAFETY: allocated by the global allocator with the layout of
+    // `[T; count]`, which is what `Vec<T>` with capacity `count` frees; all
+    // bytes are zero, a valid `T` (`NativeElement`: any bit pattern is).
+    Ok(unsafe { Vec::from_raw_parts(ptr.cast::<T>(), count, count) })
+}
+
+/// Read a whole chunked dataset that stores `T` natively
+/// ([`NativeElement::is_native`]) straight into a `Vec<T>`: each chunk is
+/// decoded and copied to its place in the typed output, with no byte buffer
+/// to convert from afterwards. Unallocated chunks read as the dataset's fill
+/// value, as [`crate::fill_value::read_full_with_fill`] makes them.
+///
+/// `Ok(None)` when this does not apply — the datatype is not `T`'s native
+/// representation (another type, another byte order: the caller converts
+/// through the byte readers and the `read_as_*` functions), the layout is
+/// not chunked, no storage is allocated, or the data lives in external
+/// files. `cache` is the file's chunk cache, used as
+/// [`crate::chunked_read::read_chunked_data_cached`] uses it.
+#[cfg(feature = "std")]
+#[allow(clippy::too_many_arguments)]
+pub fn read_chunked_native<T: NativeElement>(
+    messages: &[crate::object_header::HeaderMessage],
+    file_data: &[u8],
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
+    cache: Option<&ChunkCache>,
+) -> Result<Option<Vec<T>>, FormatError> {
+    read_chunked_native_in(
+        messages,
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        pipeline,
+        offset_size,
+        length_size,
+        cache,
+    )
+}
+
+/// [`read_chunked_native`] over any [`Storage`].
+#[cfg(feature = "std")]
+#[allow(clippy::too_many_arguments)]
+pub fn read_chunked_native_in<T: NativeElement, S: Storage + ?Sized>(
+    messages: &[crate::object_header::HeaderMessage],
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
+    cache: Option<&ChunkCache>,
+) -> Result<Option<Vec<T>>, FormatError> {
+    use crate::fill_value;
+    use crate::message_type::MessageType;
+
+    if !T::is_native(datatype)
+        || !matches!(layout, DataLayout::Chunked { .. })
+        || !fill_value::has_storage(layout)
+        || messages
+            .iter()
+            .any(|m| m.msg_type == MessageType::ExternalDataFiles)
+    {
+        return Ok(None);
+    }
+    let size = core::mem::size_of::<T>();
+    let mut values = crate::chunked_read::read_chunked_full(
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        pipeline,
+        offset_size,
+        length_size,
+        cache,
+        |total_bytes| {
+            if !total_bytes.is_multiple_of(size) {
+                return Err(FormatError::DataSizeMismatch {
+                    expected: total_bytes.next_multiple_of(size),
+                    actual: total_bytes,
+                });
+            }
+            alloc_zeroed_values::<T>(total_bytes / size)
+        },
+        |values| bytes_of_mut(values),
+    )?;
+    let fill =
+        fill_value::dataset_fill_value_from_storage(file_data, messages, offset_size, length_size)?;
+    fill_value::apply_to_unallocated_chunks_in(
+        bytes_of_mut(&mut values),
+        file_data,
+        layout,
+        dataspace,
+        size,
+        fill.as_deref(),
+        offset_size,
+        length_size,
+    )?;
+    Ok(Some(values))
+}
+
 /// Convert raw bytes to `f64` values.
 pub fn read_as_f64(raw: &[u8], datatype: &Datatype) -> Result<Vec<f64>, FormatError> {
     // Array datatypes read as a flat sequence of their base elements, and
@@ -1165,7 +1579,7 @@ pub fn read_compound_fields(
             let mut fields = Vec::with_capacity(members.len());
             for m in members {
                 let field_size = m.datatype.type_size() as usize;
-                let offset = m.byte_offset as usize;
+                let offset = to_usize(m.byte_offset)?;
                 if offset
                     .checked_add(field_size)
                     .is_none_or(|end| end > elem_size)
@@ -2632,6 +3046,71 @@ mod tests {
         assert_eq!(result.unwrap(), &[1.5f32, 2.5, 3.5]);
     }
 
+    /// libhdf5 reads a contiguous dataset's elements from the start of its
+    /// storage and ignores storage past them (cve-2024-32623's scalar
+    /// `/Dset1` has 240 bytes of storage for one 4-byte element). Storage too
+    /// small for the elements is still an error.
+    #[test]
+    fn contiguous_storage_larger_than_the_elements_reads() {
+        let dt = make_f64_le_type();
+        let ds = make_simple_dataspace(&[2]);
+        let mut file_data = vec![0u8; 64];
+        file_data[..8].copy_from_slice(&1.5f64.to_le_bytes());
+        file_data[8..16].copy_from_slice(&2.5f64.to_le_bytes());
+        file_data[16..24].copy_from_slice(&9.0f64.to_le_bytes());
+        let layout = DataLayout::Contiguous {
+            address: Some(0),
+            size: 40,
+        };
+        let raw = read_raw_data(&file_data, &layout, &ds, &dt).unwrap();
+        assert_eq!(raw, file_data[..16]);
+        let zc = read_raw_data_zerocopy(&file_data, &layout, &ds, &dt).unwrap();
+        assert_eq!(zc, Some(&file_data[..16]));
+        let small = DataLayout::Contiguous {
+            address: Some(0),
+            size: 8,
+        };
+        assert!(matches!(
+            read_raw_data(&file_data, &small, &ds, &dt),
+            Err(FormatError::DataSizeMismatch { .. })
+        ));
+    }
+
+    /// `H5D__contig_check` / `H5D__compact_init`, run when a dataset opens.
+    #[test]
+    fn dataset_storage_checks_at_open() {
+        let dt = make_f64_le_type();
+        let contiguous = |address| DataLayout::Contiguous { address, size: 0 };
+        // cve-2024-32624 `/Dset_OBJREF`: 2^62 + 2 elements of 8 bytes.
+        let huge = make_simple_dataspace(&[(1 << 62) + 2]);
+        assert_eq!(
+            check_dataset_storage(&contiguous(None), &huge, &dt, 1 << 20),
+            Err(FormatError::InvalidDatasetStorage(
+                "size of dataset's storage overflowed"
+            ))
+        );
+        let ds = make_simple_dataspace(&[4]);
+        assert!(check_dataset_storage(&contiguous(Some(100)), &ds, &dt, 132).is_ok());
+        assert!(matches!(
+            check_dataset_storage(&contiguous(Some(100)), &ds, &dt, 131),
+            Err(FormatError::InvalidDatasetStorage(_))
+        ));
+        assert!(matches!(
+            check_dataset_storage(&contiguous(Some(u64::MAX - 8)), &ds, &dt, u64::MAX),
+            Err(FormatError::InvalidDatasetStorage(_))
+        ));
+        // Not allocated, and (unlike libhdf5) empty at a defined address.
+        assert!(check_dataset_storage(&contiguous(None), &ds, &dt, 0).is_ok());
+        let empty = make_simple_dataspace(&[0]);
+        assert!(check_dataset_storage(&contiguous(Some(64)), &empty, &dt, 64).is_ok());
+        let compact = |n: usize| DataLayout::Compact { data: vec![0; n] };
+        assert!(check_dataset_storage(&compact(32), &ds, &dt, 0).is_ok());
+        assert!(matches!(
+            check_dataset_storage(&compact(24), &ds, &dt, 0),
+            Err(FormatError::InvalidDatasetStorage(_))
+        ));
+    }
+
     #[test]
     fn zerocopy_size_mismatch() {
         let dt = make_f64_le_type();

crates/clawhdf5-format/src/dataspace.rs

@@ -7,6 +7,9 @@ use alloc::vec::Vec;
 
 use crate::error::FormatError;
 
+/// Most dimensions a dataspace can have (`H5S_MAX_RANK`).
+pub const MAX_RANK: u8 = 32;
+
 /// Type of dataspace.
 #[derive(Debug, Clone, PartialEq)]
 pub enum DataspaceType {
@@ -67,6 +70,12 @@ impl Dataspace {
         let version = data[0];
         let rank = data[1];
         let flags = data[2];
+        // H5O__sdspace_decode's checks.
+        if rank > MAX_RANK {
+            return Err(FormatError::InvalidDataspace(
+                "simple dataspace dimensionality is too large",
+            ));
+        }
 
         let (space_type, header_size) = match version {
             1 => {
@@ -88,6 +97,11 @@ impl Dataspace {
                     2 => DataspaceType::Null,
                     _ => return Err(FormatError::InvalidDataspaceType(type_byte)),
                 };
+                if st != DataspaceType::Simple && rank > 0 {
+                    return Err(FormatError::InvalidDataspace(
+                        "invalid rank for scalar or NULL dataspace",
+                    ));
+                }
                 (st, 4usize)
             }
             _ => return Err(FormatError::InvalidDataspaceVersion(version)),
@@ -107,8 +121,13 @@ impl Dataspace {
         // Read max dimensions if flags bit 0 is set
         let max_dimensions = if flags & 0x01 != 0 {
             let mut max_dims = Vec::with_capacity(rank as usize);
-            for _ in 0..rank {
+            for &dim in &dimensions {
                 let val = read_length(data, pos, length_size)?;
+                if dim > val {
+                    return Err(FormatError::InvalidDataspace(
+                        "dataspace dimension size is greater than its maximum size",
+                    ));
+                }
                 max_dims.push(val);
                 pos += ls;
             }
@@ -176,7 +195,6 @@ impl Dataspace {
         match self.space_type {
             DataspaceType::Null => Ok(0),
             DataspaceType::Scalar => Ok(1),
-            DataspaceType::Simple if self.dimensions.is_empty() => Ok(0),
             DataspaceType::Simple => self
                 .dimensions
                 .iter()
@@ -195,18 +213,14 @@ impl Dataspace {
         match self.space_type {
             DataspaceType::Null => 0,
             DataspaceType::Scalar => 1,
-            DataspaceType::Simple => {
+            // A simple dataspace of rank 0 holds one element, as in libhdf5
-                if self.dimensions.is_empty() {
+            // (the product of no dimensions). Saturate rather than wrap: a
-                    0
+            // wrapped product could under-size a buffer. Size-critical
-                } else {
+            // callers use `checked_num_elements`.
-                    // Saturate rather than wrap: a wrapped product could
+            DataspaceType::Simple => self
-                    // under-size a buffer. Size-critical callers use
+                .dimensions
-                    // `checked_num_elements`.
-                    self.dimensions
                 .iter()
-                        .fold(1u64, |acc, &d| acc.saturating_mul(d))
+                .fold(1u64, |acc, &d| acc.saturating_mul(d)),
-                }
-            }
         }
     }
 }
@@ -352,4 +366,44 @@ mod tests {
         let ds = Dataspace::parse(&data, 8).unwrap();
         assert_eq!(ds.max_dimensions, Some(vec![10]));
     }
+
+    /// A simple dataspace of rank 0 (cve-2020-18494's `/dset1`) holds one
+    /// element in libhdf5, which h5py reads as shape `()`. It was 0.
+    #[test]
+    fn simple_rank_zero_holds_one_element() {
+        let data = build_v2_dataspace(0, 0, 1, &[], None);
+        let ds = Dataspace::parse(&data, 8).unwrap();
+        assert_eq!(ds.space_type, DataspaceType::Simple);
+        assert_eq!(ds.num_elements(), 1);
+        assert_eq!(ds.checked_num_elements().unwrap(), 1);
+    }
+
+    /// `H5O__sdspace_decode`'s checks.
+    #[test]
+    fn refuses_what_libhdf5_refuses() {
+        let too_many = build_v2_dataspace(33, 0, 1, &[1; 33], None);
+        assert!(matches!(
+            Dataspace::parse(&too_many, 8),
+            Err(FormatError::InvalidDataspace(_))
+        ));
+        let scalar_with_rank = build_v2_dataspace(1, 0, 0, &[4], None);
+        assert!(matches!(
+            Dataspace::parse(&scalar_with_rank, 8),
+            Err(FormatError::InvalidDataspace(_))
+        ));
+        let null_with_rank = build_v2_dataspace(1, 0, 2, &[4], None);
+        assert!(matches!(
+            Dataspace::parse(&null_with_rank, 8),
+            Err(FormatError::InvalidDataspace(_))
+        ));
+        let over_max = build_v1_dataspace(2, 0x01, &[5, 20], Some(&[10, 10]));
+        assert!(matches!(
+            Dataspace::parse(&over_max, 8),
+            Err(FormatError::InvalidDataspace(_))
+        ));
+        // 32 dimensions, and a size equal to the maximum or unlimited, are fine.
+        assert!(Dataspace::parse(&build_v2_dataspace(32, 0, 1, &[1; 32], None), 8).is_ok());
+        let at_max = build_v1_dataspace(2, 0x01, &[10, 20], Some(&[10, u64::MAX]));
+        assert!(Dataspace::parse(&at_max, 8).is_ok());
+    }
 }

crates/clawhdf5-format/src/ea_writer.rs

@@ -3,6 +3,7 @@
 #[cfg(not(feature = "std"))]
 extern crate alloc;
 
+use crate::addr::saturating_usize;
 #[cfg(not(feature = "std"))]
 use alloc::{vec, vec::Vec};
 
@@ -247,7 +248,7 @@ pub fn build_extensible_array_at(
 
     // Header (EAHD). The six statistics are, in order: super blocks, their
     // bytes, data blocks, their bytes, max index set, elements realised.
-    let mut out = Vec::with_capacity((cursor - ea_base_address) as usize);
+    let mut out = Vec::with_capacity(saturating_usize(cursor - ea_base_address));
     out.extend_from_slice(b"EAHD");
     out.push(0); // version
     out.push(client_id);

crates/clawhdf5-format/src/error.rs

@@ -12,7 +12,11 @@ use std::string::String;
 use core::fmt;
 
 /// Errors that can occur when parsing HDF5 binary format structures.
+///
+/// Non-exhaustive: new failure modes (new storage backends, new file
+/// features) add variants, so a `match` needs a wildcard arm.
 #[derive(Debug, Clone, PartialEq, Eq)]
+#[non_exhaustive]
 pub enum FormatError {
     /// The HDF5 magic signature was not found at any valid offset.
     SignatureNotFound,
@@ -223,6 +227,33 @@ pub enum FormatError {
         /// The file's actual length in bytes.
         actual_len: u64,
     },
+    /// A link libhdf5 refuses to list: a symbol-table entry with an empty
+    /// name ("invalid link name"). Listing the group fails, as in libhdf5.
+    InvalidLinkName,
+    /// A dataspace message libhdf5 refuses to decode (the reason is
+    /// libhdf5's own error text): more than 32 dimensions, a rank on a
+    /// scalar or null dataspace, a dimension larger than its maximum.
+    InvalidDataspace(&'static str),
+    /// A dataset whose storage libhdf5 refuses when it opens the dataset
+    /// (the reason is libhdf5's own error text): an element count times
+    /// element size that overflows, contiguous storage past the end of the
+    /// file, compact data of the wrong size.
+    InvalidDatasetStorage(&'static str),
+    /// A superblock extension message libhdf5 refuses to decode when it
+    /// opens the file (the reason is libhdf5's own error text): a File Space
+    /// Info message that runs off its end or has a bad page size, a metadata
+    /// cache image outside the file, …
+    InvalidSuperblockExtension(&'static str),
+    /// A metadata cache image block libhdf5 refuses to load (the reason is
+    /// libhdf5's own error text).
+    InvalidCacheImage(&'static str),
+    /// The [`Storage`](crate::storage::Storage) backend failed to serve a
+    /// read (an I/O or network error, or a short read inside the file).
+    Storage(String),
+    /// The operation still needs the whole file as one slice and the
+    /// [`Storage`](crate::storage::Storage) backend has no contiguous view
+    /// (`as_contiguous()` is `None`); the text names the operation.
+    ContiguousStorageRequired(&'static str),
 }
 
 impl fmt::Display for FormatError {
@@ -494,6 +525,31 @@ impl fmt::Display for FormatError {
                      but the file is {actual_len} bytes"
                 )
             }
+            FormatError::InvalidLinkName => {
+                write!(f, "invalid link name: a group entry has an empty name")
+            }
+            FormatError::InvalidDataspace(why) => {
+                write!(f, "invalid dataspace: {why}")
+            }
+            FormatError::InvalidDatasetStorage(why) => {
+                write!(f, "invalid dataset storage: {why}")
+            }
+            FormatError::InvalidSuperblockExtension(why) => {
+                write!(f, "invalid superblock extension: {why}")
+            }
+            FormatError::InvalidCacheImage(why) => {
+                write!(f, "invalid metadata cache image: {why}")
+            }
+            FormatError::Storage(why) => {
+                write!(f, "storage read failed: {why}")
+            }
+            FormatError::ContiguousStorageRequired(what) => {
+                write!(
+                    f,
+                    "{what} needs the whole file in memory, which this storage backend does \
+                     not provide"
+                )
+            }
         }
     }
 }

crates/clawhdf5-format/src/extensible_array.rs

@@ -9,19 +9,23 @@ extern crate alloc;
 #[cfg(not(feature = "std"))]
 use alloc::{format, vec, vec::Vec};
 
+use crate::addr::to_usize;
 use crate::chunk_grid::ChunkGrid;
 use crate::chunked_read::ChunkInfo;
 use crate::error::FormatError;
+use crate::storage::{PAGED_BLOCK_ONE_READ_MAX, Storage, Window, read_exact_at};
 
 /// Verify the Jenkins lookup3 checksum stored immediately after
-/// `data[start..end]`, as every Extensible Array structure carries one.
+/// `data[start..end]`, as every Extensible Array structure carries one. `w`
+/// is a window of the file and `start`/`end` are relative to it.
 ///
 /// A corrupt chunk index yields addresses pointing at the wrong bytes, so a
 /// mismatch is an error: otherwise the damage surfaces as plausible data read
 /// from the wrong chunk.
 #[cfg(feature = "checksum")]
-fn verify_checksum(data: &[u8], start: usize, end: usize) -> Result<(), FormatError> {
+fn verify_checksum(w: &Window<'_>, start: usize, end: usize) -> Result<(), FormatError> {
-    ensure_len(data, end, 4)?;
+    w.ensure(end, 4)?;
+    let data: &[u8] = &w.bytes;
     let stored = u32::from_le_bytes([data[end], data[end + 1], data[end + 2], data[end + 3]]);
     let computed = crate::checksum::jenkins_lookup3(&data[start..end]);
     if computed != stored {
@@ -34,7 +38,7 @@ fn verify_checksum(data: &[u8], start: usize, end: usize) -> Result<(), FormatEr
 }
 
 #[cfg(not(feature = "checksum"))]
-fn verify_checksum(_data: &[u8], _start: usize, _end: usize) -> Result<(), FormatError> {
+fn verify_checksum(_w: &Window<'_>, _start: usize, _end: usize) -> Result<(), FormatError> {
     Ok(())
 }
 
@@ -80,19 +84,6 @@ fn read_offset(data: &[u8], pos: usize, size: u8) -> Result<u64, FormatError> {
     })
 }
 
-fn ensure_len(data: &[u8], offset: usize, needed: usize) -> Result<(), FormatError> {
-    if offset
-        .checked_add(needed)
-        .is_none_or(|end| end > data.len())
-    {
-        return Err(FormatError::UnexpectedEof {
-            expected: offset.saturating_add(needed),
-            available: data.len(),
-        });
-    }
-    Ok(())
-}
-
 fn is_undefined_addr(addr: u64, offset_size: u8) -> bool {
     match offset_size {
         2 => addr == 0xFFFF,
@@ -130,6 +121,16 @@ impl ExtensibleArrayHeader {
         offset: usize,
         offset_size: u8,
         length_size: u8,
+    ) -> Result<Self, FormatError> {
+        Self::parse_in(file_data, offset as u64, offset_size, length_size)
+    }
+
+    /// [`Self::parse`] over any [`Storage`]: one read of the header.
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
     ) -> Result<Self, FormatError> {
         // EAHD: signature(4) + version(1) + client_id(1) + element_size(1) +
         //   max_nelmts_bits(1) + idx_blk_elmts(1) + min_dblk_nelmts(1) +
@@ -137,9 +138,10 @@ impl ExtensibleArrayHeader {
         //   6 stats fields (each length_size) + index_block_address(offset_size) + checksum(4)
         let min_size =
             4 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 6 * length_size as usize + offset_size as usize + 4;
-        ensure_len(file_data, offset, min_size)?;
+        let w = Window::read(file, offset, min_size)?;
+        w.ensure(0, min_size)?;
 
-        let d = &file_data[offset..];
+        let d: &[u8] = &w.bytes;
         if &d[0..4] != b"EAHD" {
             return Err(FormatError::ChunkedReadError(
                 "invalid Extensible Array header signature".into(),
@@ -172,7 +174,7 @@ impl ExtensibleArrayHeader {
         pos += ls; // skip max_idx_set (6th stats field)
         let index_block_address = read_offset(d, pos, offset_size)?;
         pos += offset_size as usize;
-        verify_checksum(file_data, offset, offset + pos)?;
+        verify_checksum(&w, 0, pos)?;
 
         Ok(ExtensibleArrayHeader {
             client_id,
@@ -193,11 +195,11 @@ impl ExtensibleArrayHeader {
     }
 }
 
-/// Read a single element from the extensible array element data.
+/// Read a single element at offset `pos` of the window `w`.
 /// Returns (chunk_info, bytes_consumed) or None if unallocated.
 #[allow(clippy::too_many_arguments)]
 fn read_element(
-    data: &[u8],
+    w: &Window<'_>,
     pos: usize,
     client_id: u8,
     element_size: u8,
@@ -207,15 +209,11 @@ fn read_element(
     grid: &ChunkGrid,
 ) -> Result<(Option<ChunkInfo>, usize), FormatError> {
     let os = offset_size as usize;
+    let data: &[u8] = &w.bytes;
 
     if client_id == 0 {
         // Non-filtered: just address
-        if pos + os > data.len() {
+        w.ensure(pos, os)?;
-            return Err(FormatError::UnexpectedEof {
-                expected: pos + os,
-                available: data.len(),
-            });
-        }
         if is_undefined(data, pos, offset_size) {
             return Ok((None, os));
         }
@@ -243,15 +241,7 @@ fn read_element(
         }
         let chunk_size_bytes = es - os - 4;
         let elem_total = os + chunk_size_bytes + 4;
-        if pos
+        w.ensure(pos, elem_total)?;
-            .checked_add(elem_total)
-            .is_none_or(|end| end > data.len())
-        {
-            return Err(FormatError::UnexpectedEof {
-                expected: pos.saturating_add(elem_total),
-                available: data.len(),
-            });
-        }
         if is_undefined(data, pos, offset_size) {
             return Ok((None, elem_total));
         }
@@ -315,9 +305,9 @@ fn page_nelmts(header: &ExtensibleArrayHeader) -> Option<usize> {
 /// paged. The bitmap lives in the super block, not here — a paged data block
 /// stores only its prefix, then one slot per page.
 #[allow(clippy::too_many_arguments)]
-fn read_data_block_elements(
+fn read_data_block_elements<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file: &S,
-    db_offset: usize,
+    db_offset: u64,
     nelmts: usize,
     header: &ExtensibleArrayHeader,
     offset_size: u8,
@@ -330,21 +320,28 @@ fn read_data_block_elements(
     // EADB: signature(4) + version(1) + client_id(1) + header_address(offset_size)
     //       + block offset(arr_off_size)
     let db_header_size = 4 + 1 + 1 + offset_size as usize + arr_off_size(header);
-    ensure_len(file_data, db_offset, db_header_size)?;
+    let prefix = read_exact_at(file, db_offset, db_header_size)?;
 
-    if &file_data[db_offset..db_offset + 4] != b"EADB" {
+    if &prefix[0..4] != b"EADB" {
         return Err(FormatError::ChunkedReadError(
             "invalid Extensible Array data block signature".into(),
         ));
     }
 
-    let mut pos = db_offset + db_header_size;
+    // Positions below are relative to the data block.
+    let mut pos = db_header_size;
     let page = page_nelmts(header).ok_or_else(|| {
         FormatError::Overflow("Extensible Array page element count overflows usize".into())
     })?;
+    let elem_bytes = if header.client_id == 0 {
+        offset_size as usize
+    } else {
+        header.element_size as usize
+    };
 
     let mut chunks = Vec::new();
-    let read_run = |from: usize,
+    let read_run = |w: &Window<'_>,
+                    from: usize,
                     count: usize,
                     first_index: usize,
                     chunks: &mut Vec<ChunkInfo>|
@@ -352,7 +349,7 @@ fn read_data_block_elements(
         let mut p = from;
         for i in 0..count {
             let (info, consumed) = read_element(
-                file_data,
+                w,
                 p,
                 header.client_id,
                 header.element_size,
@@ -370,18 +367,19 @@ fn read_data_block_elements(
     };
 
     if nelmts <= page {
-        // Prefix and elements are covered by one checksum.
+        // Prefix and elements are covered by one checksum. One window holds
-        let elem_bytes = if header.client_id == 0 {
+        // all of it (or ends at the end of the file), so its bounds checks
-            offset_size as usize
+        // are the whole-file ones.
-        } else {
-            header.element_size as usize
-        };
         let end = nelmts
             .checked_mul(elem_bytes)
             .and_then(|b| pos.checked_add(b))
             .ok_or_else(|| FormatError::Overflow("Extensible Array data block span".into()))?;
-        verify_checksum(file_data, db_offset, end)?;
+        // The checksum's bounds check comes first: make it before reading.
-        read_run(pos, nelmts, start_index, &mut chunks)?;
+        #[cfg(feature = "checksum")]
+        Window::check_extent(file, db_offset, end, 4)?;
+        let w = Window::read(file, db_offset, end.saturating_add(4))?;
+        verify_checksum(&w, 0, end)?;
+        read_run(&w, pos, nelmts, start_index, &mut chunks)?;
         return Ok(chunks);
     }
 
@@ -389,18 +387,32 @@ fn read_data_block_elements(
     // each holding `page` elements followed by a checksum. Pages whose bit is
     // clear were never written; their slot still occupies the file, so stride
     // over it rather than reading zeros as addresses.
-    verify_checksum(file_data, db_offset, pos)?;
+    let npages = nelmts.div_ceil(page);
-    pos += 4;
+    // The whole data block in one window when it is small: every position
-    let elem_bytes = if header.client_id == 0 {
+    // checked below lies inside it (or past the end of the file). A larger
-        offset_size as usize
+    // block is read as its prefix, then each page in use on its own.
+    let block_len = pos
+        .saturating_add(4)
+        .saturating_add(npages.saturating_mul(page.saturating_mul(elem_bytes).saturating_add(4)));
+    let whole = if block_len <= PAGED_BLOCK_ONE_READ_MAX {
+        Some(Window::read(file, db_offset, block_len)?)
     } else {
-        header.element_size as usize
+        None
     };
+    let head_w;
+    let head = match &whole {
+        Some(w) => w,
+        None => {
+            head_w = Window::read(file, db_offset, pos + 4)?;
+            &head_w
+        }
+    };
+    verify_checksum(head, 0, pos)?;
+    pos += 4;
     let page_stride = page
         .checked_mul(elem_bytes)
         .and_then(|b| b.checked_add(4))
         .ok_or_else(|| FormatError::Overflow("Extensible Array page stride".into()))?;
-    let npages = nelmts.div_ceil(page);
     for p in 0..npages {
         // One bit per page across the whole super block, packed contiguously
         // and MSB-first within each byte, as H5VM_bit_get reads it.
@@ -410,10 +422,20 @@ fn read_data_block_elements(
             .is_some_and(|byte| byte & (0x80 >> (bit % 8)) != 0);
         if initialised {
             let count = core::cmp::min(page, nelmts - p * page);
+            // `w` holds the page from `base` on (positions below are
+            // relative to it, and `pos` to the data block).
+            let page_w;
+            let (w, base) = match &whole {
+                Some(w) => (w, 0),
+                None => {
+                    page_w = Window::read(file, db_offset.saturating_add(pos as u64), page_stride)?;
+                    (&page_w, pos)
+                }
+            };
             // Each page carries its own checksum, over a full page's worth of
             // slots even when the last one holds fewer live elements.
-            verify_checksum(file_data, pos, pos + page * elem_bytes)?;
+            verify_checksum(w, pos - base, pos - base + page * elem_bytes)?;
-            read_run(pos, count, start_index + p * page, &mut chunks)?;
+            read_run(w, pos - base, count, start_index + p * page, &mut chunks)?;
         }
         pos = pos
             .checked_add(page_stride)
@@ -435,6 +457,32 @@ pub fn read_extensible_array_chunks(
     chunk_dimensions: &[u32],
     element_size: u32,
     offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<ChunkInfo>, FormatError> {
+    read_extensible_array_chunks_in(
+        &file_data,
+        header,
+        dataset_dims,
+        max_dims,
+        chunk_dimensions,
+        element_size,
+        offset_size,
+        length_size,
+    )
+}
+
+/// [`read_extensible_array_chunks`] over any [`Storage`]: one read of the
+/// index block's prefix, one of the whole index block, and the same for
+/// every super block and data block it references.
+#[allow(clippy::too_many_arguments)]
+pub fn read_extensible_array_chunks_in<S: Storage + ?Sized>(
+    file: &S,
+    header: &ExtensibleArrayHeader,
+    dataset_dims: &[u64],
+    max_dims: Option<&[u64]>,
+    chunk_dimensions: &[u32],
+    element_size: u32,
+    offset_size: u8,
     _length_size: u8,
 ) -> Result<Vec<ChunkInfo>, FormatError> {
     let os = offset_size as usize;
@@ -451,19 +499,20 @@ pub fn read_extensible_array_chunks(
     // Parse index block (EAIB): signature(4) + version(1) + client_id(1)
     // + header address(offset_size), then the inline elements, then the
     // direct data block addresses, then the super block addresses.
-    let ib_offset = header.index_block_address as usize;
+    // Positions below are relative to the index block.
+    let ib_offset = header.index_block_address;
     let ib_header_size = 4 + 1 + 1 + os;
-    ensure_len(file_data, ib_offset, ib_header_size)?;
+    let prefix = read_exact_at(file, ib_offset, ib_header_size)?;
 
-    if &file_data[ib_offset..ib_offset + 4] != b"EAIB" {
+    if &prefix[0..4] != b"EAIB" {
         return Err(FormatError::ChunkedReadError(
             "invalid Extensible Array index block signature".into(),
         ));
     }
-    let mut pos = ib_offset + ib_header_size;
+    let mut pos = ib_header_size;
 
     let mut chunks = Vec::new();
-    let total_elements = header.num_elements as usize;
+    let total_elements = to_usize(header.num_elements)?;
 
     let dmin = header.min_dblk_nelmts as usize;
     if dmin == 0 || !dmin.is_power_of_two() {
@@ -520,13 +569,19 @@ pub fn read_extensible_array_chunks(
                 .and_then(|n| n.checked_mul(os).and_then(|b| p.checked_add(b)))
         })
         .ok_or_else(|| FormatError::Overflow("Extensible Array index block span".into()))?;
-    verify_checksum(file_data, ib_offset, ib_end)?;
+    // The whole index block in one window: every position read below is
+    // before `ib_end`.
+    // The checksum's bounds check comes first: make it before reading.
+    #[cfg(feature = "checksum")]
+    Window::check_extent(file, ib_offset, ib_end, 4)?;
+    let w = Window::read(file, ib_offset, ib_end.saturating_add(4))?;
+    verify_checksum(&w, 0, ib_end)?;
 
     // 1. Elements stored inline in the index block.
     let n_inline = (header.idx_blk_elmts as usize).min(total_elements);
     for i in 0..n_inline {
         let (info, consumed) = read_element(
-            file_data,
+            &w,
             pos,
             header.client_id,
             header.element_size,
@@ -550,8 +605,8 @@ pub fn read_extensible_array_chunks(
         if global_index >= total_elements {
             return Ok(chunks);
         }
-        ensure_len(file_data, pos, os)?;
+        w.ensure(pos, os)?;
-        let addr = read_offset(file_data, pos, offset_size)?;
+        let addr = read_offset(&w.bytes, pos, offset_size)?;
         pos += os;
         if !is_undefined_addr(addr, offset_size) {
             if dblk_nelmts > page_nelmts(header).unwrap_or(usize::MAX) {
@@ -562,8 +617,8 @@ pub fn read_extensible_array_chunks(
                 ));
             }
             chunks.extend(read_data_block_elements(
-                file_data,
+                file,
-                addr as usize,
+                addr,
                 dblk_nelmts,
                 header,
                 offset_size,
@@ -583,16 +638,16 @@ pub fn read_extensible_array_chunks(
         if global_index >= total_elements {
             break;
         }
-        ensure_len(file_data, pos, os)?;
+        w.ensure(pos, os)?;
-        let sb_addr = read_offset(file_data, pos, offset_size)?;
+        let sb_addr = read_offset(&w.bytes, pos, offset_size)?;
         pos += os;
         let (ndblks, dblk_nelmts) = sblk_info(u, dmin).ok_or_else(|| {
             FormatError::Overflow("Extensible Array super block layout overflows usize".into())
         })?;
         if !is_undefined_addr(sb_addr, offset_size) {
             chunks.extend(read_super_block(
-                file_data,
+                file,
-                sb_addr as usize,
+                sb_addr,
                 ndblks,
                 dblk_nelmts,
                 header,
@@ -617,9 +672,9 @@ pub fn read_extensible_array_chunks(
 /// + block offset + the page-init bitmap for every data block it owns
 /// + one address per data block + checksum.
 #[allow(clippy::too_many_arguments)]
-fn read_super_block(
+fn read_super_block<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file: &S,
-    sb_offset: usize,
+    sb_offset: u64,
     ndblks: usize,
     dblk_nelmts: usize,
     header: &ExtensibleArrayHeader,
@@ -630,9 +685,9 @@ fn read_super_block(
 ) -> Result<Vec<ChunkInfo>, FormatError> {
     let os = offset_size as usize;
     let sb_header_size = 4 + 1 + 1 + os + arr_off_size(header);
-    ensure_len(file_data, sb_offset, sb_header_size)?;
+    let prefix = read_exact_at(file, sb_offset, sb_header_size)?;
 
-    if &file_data[sb_offset..sb_offset + 4] != b"EASB" {
+    if &prefix[0..4] != b"EASB" {
         return Err(FormatError::ChunkedReadError(
             "invalid Extensible Array super block signature".into(),
         ));
@@ -654,29 +709,38 @@ fn read_super_block(
     let bitmap_bytes = per_dblk_bitmap
         .checked_mul(ndblks)
         .ok_or_else(|| FormatError::Overflow("Extensible Array page bitmap size".into()))?;
-    let bitmap_start = sb_offset + sb_header_size;
+    // Positions below are relative to the super block, whose bytes (up to
-    ensure_len(file_data, bitmap_start, bitmap_bytes)?;
+    // its checksum) are all in one window.
-    let bitmap = &file_data[bitmap_start..bitmap_start + bitmap_bytes];
+    let bitmap_start = sb_header_size;
-
+    // The bitmap's bounds check, then (with checksums) the checksum's, come
+    // before anything else is read from the block: make them before reading
+    // it, so size fields stretching it past the end of the file cost no read.
+    Window::check_extent(file, sb_offset, bitmap_start, bitmap_bytes)?;
     let mut pos = bitmap_start + bitmap_bytes;
-    let mut chunks = Vec::new();
-    let mut global_idx = start_index;
 
     // One checksum covers the prefix, the bitmap and every data block address.
     let sb_end = ndblks
         .checked_mul(os)
         .and_then(|b| pos.checked_add(b))
         .ok_or_else(|| FormatError::Overflow("Extensible Array super block span".into()))?;
-    verify_checksum(file_data, sb_offset, sb_end)?;
+    #[cfg(feature = "checksum")]
+    Window::check_extent(file, sb_offset, sb_end, 4)?;
+    let w = Window::read(file, sb_offset, sb_end.saturating_add(4))?;
+    w.ensure(bitmap_start, bitmap_bytes)?;
+    let bitmap = &w.bytes[bitmap_start..bitmap_start + bitmap_bytes];
+
+    let mut chunks = Vec::new();
+    let mut global_idx = start_index;
+    verify_checksum(&w, 0, sb_end)?;
 
     for i in 0..ndblks {
-        ensure_len(file_data, pos, os)?;
+        w.ensure(pos, os)?;
-        let addr = read_offset(file_data, pos, offset_size)?;
+        let addr = read_offset(&w.bytes, pos, offset_size)?;
         pos += os;
         if !is_undefined_addr(addr, offset_size) {
             chunks.extend(read_data_block_elements(
-                file_data,
+                file,
-                addr as usize,
+                addr,
                 dblk_nelmts,
                 header,
                 offset_size,
@@ -887,11 +951,11 @@ mod tests {
         assert_eq!(chunks[1].offsets, vec![20]);
     }
 
-    /// Build a synthetic EA with inline elements + one direct data block.
+    /// A synthetic EA with inline elements + one direct data block: the
-    #[test]
+    /// file, with the header at 0x100 (8-byte offsets and lengths, 4 chunks
-    fn read_inline_plus_data_blocks() {
+    /// of 10 elements from 0x1000 on).
+    fn build_inline_plus_data_blocks() -> Vec<u8> {
         let os: u8 = 8;
-        let ls: u8 = 8;
         let osv = os as usize;
         let chunk_byte_size = 10u64 * 8; // 10 elements × 8 bytes
         let idx_blk_elmts = 2u8;
@@ -981,8 +1045,17 @@ mod tests {
             dbpos += osv;
         }
         stamp_checksum(&mut file_data, aedb_offset, dbpos);
+        file_data
+    }
 
-        let header = ExtensibleArrayHeader::parse(&file_data, aehd_offset, os, ls).unwrap();
+    /// Build a synthetic EA with inline elements + one direct data block.
+    #[test]
+    fn read_inline_plus_data_blocks() {
+        let (os, ls) = (8u8, 8u8);
+        let chunk_byte_size = 10u64 * 8;
+        let base_addr = 0x1000u64;
+        let file_data = build_inline_plus_data_blocks();
+        let header = ExtensibleArrayHeader::parse(&file_data, 0x100, os, ls).unwrap();
         let ds_dims = vec![40u64];
         let chunk_dims = vec![10u32];
         let chunks = read_extensible_array_chunks(
@@ -1018,7 +1091,8 @@ mod tests {
     fn read_element_unallocated() {
         let data = vec![0xFFu8; 16];
         let grid = ChunkGrid::fixed_array(&[50], None, &[10]).unwrap();
-        let (info, consumed) = read_element(&data, 0, 0, 8, 8, 80, 0, &grid).unwrap();
+        let (info, consumed) =
+            read_element(&Window::whole(&data), 0, 0, 8, 8, 80, 0, &grid).unwrap();
         assert!(info.is_none());
         assert_eq!(consumed, 8);
     }
@@ -1038,8 +1112,17 @@ mod tests {
         data[12..16].copy_from_slice(&0u32.to_le_bytes());
 
         let grid = ChunkGrid::fixed_array(&[50], None, &[10]).unwrap();
-        let (info, consumed) =
+        let (info, consumed) = read_element(
-            read_element(&data, 0, 1, elem_size as u8, os, 80, 2, &grid).unwrap();
+            &Window::whole(&data),
+            0,
+            1,
+            elem_size as u8,
+            os,
+            80,
+            2,
+            &grid,
+        )
+        .unwrap();
         let ci = info.unwrap();
         assert_eq!(ci.address, 0x2000);
         assert_eq!(ci.chunk_size, 120);
@@ -1047,4 +1130,38 @@ mod tests {
         assert_eq!(ci.offsets, vec![20]);
         assert_eq!(consumed, elem_size);
     }
+
+    /// The Storage path reads exactly what the slice path reads: the array
+    /// whole, cut at every length through its structures, and with a byte
+    /// damaged in each of them, through a read_at-only CountingStorage.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::storage::CountingStorage;
+        let full = build_inline_plus_data_blocks();
+        let mut files = Vec::new();
+        for cut in 0x100..0x340 {
+            files.push(full[..cut].to_vec());
+        }
+        for at in [0x104, 0x150, 0x204, 0x216, 0x230, 0x304, 0x318] {
+            let mut damaged = full.clone();
+            damaged[at] ^= 1;
+            files.push(damaged);
+        }
+        files.push(full);
+        let mut compared = 0;
+        for f in files {
+            let storage = CountingStorage::new(f.clone());
+            let want = ExtensibleArrayHeader::parse(&f, 0x100, 8, 8);
+            let got = ExtensibleArrayHeader::parse_in(&storage, 0x100, 8, 8);
+            assert_eq!(format!("{got:?}"), format!("{want:?}"));
+            let Ok(h) = want else { continue };
+            for dims in [&[40u64][..], &[25]] {
+                let want = read_extensible_array_chunks(&f, &h, dims, None, &[10], 8, 8, 8);
+                let got = read_extensible_array_chunks_in(&storage, &h, dims, None, &[10], 8, 8, 8);
+                assert_eq!(format!("{got:?}"), format!("{want:?}"), "{} bytes", f.len());
+                compared += 1;
+            }
+        }
+        assert!(compared > 100);
+    }
 }

crates/clawhdf5-format/src/file_writer.rs

@@ -3,10 +3,12 @@
 //! Produces valid HDF5 files with v3 superblock, v2 object headers,
 //! link messages, contiguous datasets, inline and dense attributes.
 
+use crate::addr::saturating_usize;
 #[cfg(not(feature = "std"))]
 use alloc::{format, vec, vec::Vec};
 
 use crate::attribute::AttributeMessage;
+use crate::btree_v2_write::{BTreeV2Params, build_btree_v2};
 use crate::chunked_write::{
     ChunkOptions, PrecompressedChunks, build_chunked_data_from_precompressed, precompress_chunks,
 };
@@ -74,14 +76,57 @@ const DENSE_LINK_THRESHOLD: usize = 8;
 
 // ---- OH builders ----
 
+/// An object's attributes as its header stores them: inline Attribute
+/// messages, or (`dense`) the Attribute Info message of dense storage; with
+/// `track_order`, their creation order tracked and indexed.
+#[derive(Clone, Copy)]
+pub(crate) struct AttrStorage<'a> {
+    pub(crate) attrs: &'a [AttributeMessage],
+    pub(crate) dense: Option<&'a DenseAttrBlob>,
+    pub(crate) track_order: bool,
+}
+
+impl AttrStorage<'_> {
+    /// Add the attribute messages to the header being built. Tracking
+    /// creation order, as libhdf5 does it: the header's flags say so, an
+    /// Attribute Info message is written even for inline attributes (it
+    /// holds the next creation order), and each inline attribute's message
+    /// carries its creation order.
+    fn add_to(&self, w: &mut ObjectHeaderWriter) {
+        if self.track_order {
+            w.track_attr_order();
+        }
+        if let Some(blob) = self.dense {
+            w.add_message(MessageType::AttributeInfo, blob.attr_info_message.clone());
+            return;
+        }
+        if self.track_order {
+            w.add_message(
+                MessageType::AttributeInfo,
+                serialize_attribute_info(
+                    u64::MAX,
+                    u64::MAX,
+                    Some((self.attrs.len() as u16, u64::MAX)),
+                ),
+            );
+        }
+        for (i, attr) in self.attrs.iter().enumerate() {
+            w.add_message_with_order(
+                MessageType::Attribute,
+                attr.serialize(LENGTH_SIZE),
+                i as u16,
+            );
+        }
+    }
+}
+
 #[allow(clippy::too_many_arguments)]
 pub(crate) fn build_chunked_dataset_oh(
     dt: &Datatype,
     ds: &Dataspace,
     layout_message: &[u8],
     pipeline_message: Option<&[u8]>,
-    attrs: &[AttributeMessage],
+    attrs: AttrStorage<'_>,
-    dense_blob: Option<&DenseAttrBlob>,
     fill_message: &[u8],
     refcount: u32,
 ) -> Result<Vec<u8>, FormatError> {
@@ -93,13 +138,7 @@ pub(crate) fn build_chunked_dataset_oh(
     if let Some(pm) = pipeline_message {
         w.add_message(MessageType::FilterPipeline, pm.to_vec());
     }
-    if let Some(blob) = dense_blob {
+    attrs.add_to(&mut w);
-        w.add_message(MessageType::AttributeInfo, blob.attr_info_message.clone());
-    } else {
-        for attr in attrs {
-            w.add_message(MessageType::Attribute, attr.serialize(LENGTH_SIZE));
-        }
-    }
     add_refcount(&mut w, refcount);
     w.serialize()
 }
@@ -110,8 +149,7 @@ pub(crate) fn build_dataset_oh(
     ds: &Dataspace,
     data_addr: u64,
     data_size: u64,
-    attrs: &[AttributeMessage],
+    attrs: AttrStorage<'_>,
-    dense_blob: Option<&DenseAttrBlob>,
     fill_message: &[u8],
     refcount: u32,
 ) -> Result<Vec<u8>, FormatError> {
@@ -131,13 +169,7 @@ pub(crate) fn build_dataset_oh(
     dl.extend_from_slice(&data_addr.to_le_bytes());
     dl.extend_from_slice(&data_size.to_le_bytes());
     w.add_message(MessageType::DataLayout, dl);
-    if let Some(blob) = dense_blob {
+    attrs.add_to(&mut w);
-        w.add_message(MessageType::AttributeInfo, blob.attr_info_message.clone());
-    } else {
-        for attr in attrs {
-            w.add_message(MessageType::Attribute, attr.serialize(LENGTH_SIZE));
-        }
-    }
     add_refcount(&mut w, refcount);
     w.serialize()
 }
@@ -147,8 +179,7 @@ pub(crate) fn build_compact_dataset_oh(
     dt: &Datatype,
     ds: &Dataspace,
     data: &[u8],
-    attrs: &[AttributeMessage],
+    attrs: AttrStorage<'_>,
-    dense_blob: Option<&DenseAttrBlob>,
     fill_message: &[u8],
     refcount: u32,
 ) -> Result<Vec<u8>, FormatError> {
@@ -163,13 +194,7 @@ pub(crate) fn build_compact_dataset_oh(
     dl.extend_from_slice(&(data.len() as u16).to_le_bytes());
     dl.extend_from_slice(data);
     w.add_message(MessageType::DataLayout, dl);
-    if let Some(blob) = dense_blob {
+    attrs.add_to(&mut w);
-        w.add_message(MessageType::AttributeInfo, blob.attr_info_message.clone());
-    } else {
-        for attr in attrs {
-            w.add_message(MessageType::Attribute, attr.serialize(LENGTH_SIZE));
-        }
-    }
     add_refcount(&mut w, refcount);
     w.serialize()
 }
@@ -181,8 +206,7 @@ pub(crate) fn build_group_oh(
     links: &[LinkMessage],
     link_info: &[u8],
     dense_links: bool,
-    attrs: &[AttributeMessage],
+    attrs: AttrStorage<'_>,
-    dense_blob: Option<&DenseAttrBlob>,
     refcount: u32,
 ) -> Result<Vec<u8>, FormatError> {
     let mut w = ObjectHeaderWriter::new();
@@ -197,13 +221,7 @@ pub(crate) fn build_group_oh(
             w.add_message(MessageType::Link, link.serialize(OFFSET_SIZE));
         }
     }
-    if let Some(blob) = dense_blob {
+    attrs.add_to(&mut w);
-        w.add_message(MessageType::AttributeInfo, blob.attr_info_message.clone());
-    } else {
-        for attr in attrs {
-            w.add_message(MessageType::Attribute, attr.serialize(LENGTH_SIZE));
-        }
-    }
     add_refcount(&mut w, refcount);
     w.serialize()
 }
@@ -319,7 +337,7 @@ pub(crate) fn build_single_block_fractal_heap(
 
     // An object must fit one direct block: the writer has no huge-object
     // path, and libhdf5 cannot read an object that overruns its block.
-    let max_managed = max_direct_block_size as usize - dblock_header_size;
+    let max_managed = saturating_usize(max_direct_block_size) - dblock_header_size;
     if let Some(big) = serialized.iter().find(|s| s.len() > max_managed) {
         return Err(FormatError::SerializationError(format!(
             "a {}-byte message cannot go in dense storage: a fractal heap \
@@ -375,7 +393,7 @@ pub(crate) fn build_single_block_fractal_heap(
     let dblock_addr = frhp_addr + frhp_size as u64;
     let btree_addr = dblock_addr + starting_block_size;
 
-    let data_space = starting_block_size as usize - dblock_header_size;
+    let data_space = saturating_usize(starting_block_size) - dblock_header_size;
     let free_space = data_space - total_data_size;
 
     // Build fractal heap header
@@ -411,7 +429,7 @@ pub(crate) fn build_single_block_fractal_heap(
     debug_assert_eq!(frhp.len(), frhp_size);
 
     // Build direct block: header (with checksum) + data + padding
-    let mut dblock = Vec::with_capacity(starting_block_size as usize);
+    let mut dblock = Vec::with_capacity(saturating_usize(starting_block_size));
     dblock.extend_from_slice(b"FHDB");
     dblock.push(0); // version
     write_offset(&mut dblock, frhp_addr, OFFSET_SIZE);
@@ -429,12 +447,12 @@ pub(crate) fn build_single_block_fractal_heap(
     }
 
     // Pad to full block size
-    dblock.resize(starting_block_size as usize, 0);
+    dblock.resize(saturating_usize(starting_block_size), 0);
 
     // Checksum: computed over entire block with checksum field zeroed
     let dblock_checksum = crate::checksum::jenkins_lookup3(&dblock);
     dblock[cksum_pos..cksum_pos + 4].copy_from_slice(&dblock_checksum.to_le_bytes());
-    debug_assert_eq!(dblock.len(), starting_block_size as usize);
+    debug_assert_eq!(dblock.len(), saturating_usize(starting_block_size));
 
     // Build heap IDs
     let heap_ids: Vec<Vec<u8>> = obj_offsets
@@ -689,7 +707,7 @@ impl HeapIndirectBlock {
                     let cksum_pos = out.len();
                     out.extend_from_slice(&[0u8; 4]); // checksum placeholder
                     out.extend_from_slice(&b.data);
-                    out.resize(d + b.size as usize, 0);
+                    out.resize(d + saturating_usize(b.size), 0);
                     let cksum = crate::checksum::jenkins_lookup3(&out[d..]);
                     out[cksum_pos..cksum_pos + 4].copy_from_slice(&cksum.to_le_bytes());
                     child += b.size;
@@ -723,7 +741,7 @@ impl HeapPacker<'_> {
         nrows: Option<usize>,
     ) -> Result<HeapIndirectBlock, FormatError> {
         let geom = self.geom;
-        let width = geom.width as usize;
+        let width = saturating_usize(geom.width);
         let mut slots = Vec::new();
         let mut off = heap_offset;
         let mut row = 0usize;
@@ -746,7 +764,8 @@ impl HeapPacker<'_> {
                     // A child whose biggest direct block cannot hold the
                     // next object is skipped whole, not walked.
                     let biggest = geom.row_size(child_rows.min(geom.max_direct_rows()) - 1);
-                    if self.objects[self.next].len() > (biggest as usize - geom.dblock_header_size)
+                    if self.objects[self.next].len()
+                        > (saturating_usize(biggest) - geom.dblock_header_size)
                     {
                         slots.push(HeapSlot::Empty);
                         off += size;
@@ -777,7 +796,7 @@ impl HeapPacker<'_> {
     /// objects as fit; leave it unallocated if not even the next one does.
     fn fill_direct(&mut self, heap_offset: u64, size: u64) -> HeapSlot {
         let header = self.geom.dblock_header_size;
-        let capacity = size as usize - header;
+        let capacity = saturating_usize(size) - header;
         let mut data = Vec::new();
         while let Some(obj) = self.objects.get(self.next) {
             if data.len() + obj.len() > capacity {
@@ -890,11 +909,31 @@ fn write_frhp(p: WriteFrhp) -> Vec<u8> {
     frhp
 }
 
+/// libhdf5 numbers the attributes of an object that tracks their creation
+/// order with a 2-byte counter.
+fn check_tracked_attr_count(track_order: bool, n: usize) -> Result<(), FormatError> {
+    if track_order && n > usize::from(u16::MAX) {
+        return Err(FormatError::SerializationError(format!(
+            "{n} attributes on one object with creation order tracked: libhdf5 \
+             numbers at most {} (set fewer, or turn off track_order)",
+            u16::MAX
+        )));
+    }
+    Ok(())
+}
+
 /// Build dense attribute storage for a set of attributes.
+///
+/// With `track_order` the Attribute Info message tracks creation order (an
+/// attribute's creation order is its position in `attrs`) and a type-9
+/// creation-order index follows the name index, as libhdf5 writes for h5py's
+/// `track_order=True`. libhdf5 numbers at most 65 535 attributes.
 pub(crate) fn build_dense_attrs(
     attrs: &[AttributeMessage],
     base_address: u64,
+    track_order: bool,
 ) -> Result<DenseAttrBlob, FormatError> {
+    check_tracked_attr_count(track_order, attrs.len())?;
     // Dense attrs use v3 attribute messages (adds character set encoding byte).
     let serialized: Vec<Vec<u8>> = attrs.iter().map(|a| a.serialize_v3(LENGTH_SIZE)).collect();
 
@@ -910,31 +949,55 @@ pub(crate) fn build_dense_attrs(
     let heap_id_length = heap.heap_id_length;
     let heap_ids = &heap.heap_ids;
 
-    // Build B-tree v2 type 8 records (17 bytes each)
+    // Build B-tree v2 type 8 records (17 bytes each), in the index's key
+    // order: libhdf5 compares the name hash, then — for names whose hashes
+    // collide — the names themselves (`strcmp`).
     let record_size: u16 = heap_id_length + 1 + 4 + 4;
-    let mut records: Vec<(u32, u32, Vec<u8>)> = Vec::with_capacity(attrs.len());
+    let mut order: Vec<usize> = (0..attrs.len()).collect();
-    for (i, heap_id) in heap_ids.iter().enumerate() {
+    order.sort_by(|&a, &b| {
+        name_hashes[a]
+            .cmp(&name_hashes[b])
+            .then_with(|| attrs[a].name.as_bytes().cmp(attrs[b].name.as_bytes()))
+    });
+    let records: Vec<Vec<u8>> = order
+        .into_iter()
+        .map(|i| {
             let mut rec = Vec::with_capacity(record_size as usize);
-        rec.extend_from_slice(heap_id);
+            rec.extend_from_slice(&heap_ids[i]);
             rec.push(0); // msg_flags
             rec.extend_from_slice(&(i as u32).to_le_bytes()); // creation_order
             rec.extend_from_slice(&name_hashes[i].to_le_bytes()); // hash
-        records.push((name_hashes[i], i as u32, rec));
+            rec
-    }
+        })
-    records.sort_by(|a, b| a.0.cmp(&b.0).then(a.1.cmp(&b.1)));
+        .collect();
-
-    let records: Vec<Vec<u8>> = records.into_iter().map(|(_, _, rec)| rec).collect();
     let bthd_addr = btree_addr;
     let mut blob = heap.blob;
-    blob.extend_from_slice(&single_leaf_v2_btree(
+    blob.extend_from_slice(&dense_v2_btree(8, record_size, &records, bthd_addr)?);
-        8,
-        record_size,
-        &records,
-        bthd_addr,
-        "attributes on one object",
-    )?);
 
-    let attr_info = serialize_attribute_info(frhp_addr, bthd_addr);
+    let order = if track_order {
+        // Type 9 records: heap ID, message flags, creation order (the key).
+        let records: Vec<Vec<u8>> = heap_ids
+            .iter()
+            .enumerate()
+            .map(|(i, heap_id)| {
+                let mut rec = heap_id.clone();
+                rec.push(0); // msg_flags
+                rec.extend_from_slice(&(i as u32).to_le_bytes());
+                rec
+            })
+            .collect();
+        let corder_addr = base_address + blob.len() as u64;
+        blob.extend_from_slice(&dense_v2_btree(
+            9,
+            heap_id_length + 1 + 4,
+            &records,
+            corder_addr,
+        )?);
+        Some((attrs.len() as u16, corder_addr))
+    } else {
+        None
+    };
+    let attr_info = serialize_attribute_info(frhp_addr, bthd_addr, order);
 
     Ok(DenseAttrBlob {
         attr_info_message: attr_info,
@@ -953,69 +1016,57 @@ pub(crate) struct DenseLinkBlob {
     pub(crate) blob: Vec<u8>,
 }
 
-/// A v2 B-tree of `btree_type` holding `records` (already in key order) in a
+/// libhdf5's node size for the dense link and attribute indexes it creates
-/// single leaf, laid out at `addr`: the header, then the leaf. `what` names
+/// (`H5G_NAME_BT2_NODE_SIZE`, `H5A_NAME_BT2_NODE_SIZE`, and the
-/// the records in the error for too many ("links in one group").
+/// creation-order indexes'), with their split and merge percentages.
-fn single_leaf_v2_btree(
+const DENSE_BT2_NODE_SIZE: u32 = 512;
+const DENSE_BT2_SPLIT_PERCENT: u8 = 100;
+const DENSE_BT2_MERGE_PERCENT: u8 = 40;
+
+/// A dense-storage v2 B-tree of `btree_type` holding `records` (already in
+/// key order), laid out at `addr`: the header, then its nodes.
+///
+/// Up to 65 535 records go in one leaf node sized to hold them (the layout
+/// the writer has always used, kept so those files do not change). A leaf's
+/// record count is a 2-byte field, and libhdf5 sizes a leaf's capacity from
+/// the node size: a node with room for more than 65 535 records makes it
+/// overflow that count when it adds one, so the node is capped at a full
+/// leaf. More records get libhdf5's own 512-byte nodes, with internal nodes
+/// above the leaves.
+fn dense_v2_btree(
     btree_type: u8,
     record_size: u16,
     records: &[Vec<u8>],
     addr: u64,
-    what: &str,
 ) -> Result<Vec<u8>, FormatError> {
-    let os = OFFSET_SIZE as usize;
+    let rs = usize::from(record_size);
-    let ls = LENGTH_SIZE as usize;
+    let n = records.len();
-    // The root node's record count is a 2-byte field; more records need
+    let node_size = if n <= usize::from(u16::MAX) {
-    // internal nodes, which the writer does not build.
+        let btlf_size = 4 + 1 + 1 + n * rs + 4;
-    let num_records = u16::try_from(records.len()).map_err(|_| {
+        let max_node = 4 + 1 + 1 + usize::from(u16::MAX) * rs + 4;
-        FormatError::SerializationError(format!(
+        u32::try_from(btlf_size.next_power_of_two().max(512).min(max_node))
-            "{} {what}: at most {} can be written \
+            .map_err(|_| FormatError::Overflow("B-tree v2 node size".into()))?
-             (a deeper B-tree index is not implemented)",
+    } else {
-            records.len(),
+        DENSE_BT2_NODE_SIZE
-            u16::MAX
+    };
-        ))
+    let flat: Vec<u8> = records
-    })?;
+        .iter()
-    let bthd_size = 4 + 1 + 1 + 4 + 2 + 2 + 1 + 1 + os + 2 + ls + 4;
+        .inspect(|r| debug_assert_eq!(r.len(), rs))
-    let btlf_size = 4 + 1 + 1 + (records.len() * record_size as usize) + 4;
+        .flat_map(|r| r.iter().copied())
-    // libhdf5 sizes a leaf's capacity from the node size, and a leaf's
+        .collect();
-    // record count is a 2-byte field: a node with room for more than
+    build_btree_v2(
-    // 65 535 records makes it overflow that count when it adds one (the
+        BTreeV2Params {
-    // group can then no longer be listed). Cap the node at a full leaf.
+            tree_type: btree_type,
-    let max_node = btlf_size - records.len() * record_size as usize
+            node_size,
-        + usize::from(u16::MAX) * record_size as usize;
+            record_size,
-    let node_size = btlf_size.next_power_of_two().max(512).min(max_node) as u32;
+            split_percent: DENSE_BT2_SPLIT_PERCENT,
-    let btlf_addr = addr + bthd_size as u64;
+            merge_percent: DENSE_BT2_MERGE_PERCENT,
-
+        },
-    let mut out = Vec::with_capacity(bthd_size + node_size as usize);
+        &flat,
-    out.extend_from_slice(b"BTHD");
+        addr,
-    out.push(0); // version
+        OFFSET_SIZE,
-    out.push(btree_type);
+        LENGTH_SIZE,
-    out.extend_from_slice(&node_size.to_le_bytes());
+    )
-    out.extend_from_slice(&record_size.to_le_bytes());
-    out.extend_from_slice(&0u16.to_le_bytes()); // depth = 0 (single leaf)
-    out.push(100); // split_percent
-    out.push(40); // merge_percent
-    write_offset(&mut out, btlf_addr, OFFSET_SIZE);
-    out.extend_from_slice(&num_records.to_le_bytes());
-    write_length(&mut out, records.len() as u64, LENGTH_SIZE);
-    let checksum = crate::checksum::jenkins_lookup3(&out);
-    out.extend_from_slice(&checksum.to_le_bytes());
-    debug_assert_eq!(out.len(), bthd_size);
-
-    let mut btlf = Vec::with_capacity(node_size as usize);
-    btlf.extend_from_slice(b"BTLF");
-    btlf.push(0); // version
-    btlf.push(btree_type);
-    for rec in records {
-        debug_assert_eq!(rec.len(), record_size as usize);
-        btlf.extend_from_slice(rec);
-    }
-    // The checksum follows the records, not the end of the node.
-    let checksum = crate::checksum::jenkins_lookup3(&btlf);
-    btlf.extend_from_slice(&checksum.to_le_bytes());
-    btlf.resize(node_size as usize, 0);
-    out.extend_from_slice(&btlf);
-    Ok(out)
 }
 
 /// Build dense link storage for a group's links, laid out at `base_address`.
@@ -1039,14 +1090,18 @@ pub(crate) fn build_dense_links(
     let heap = build_single_block_fractal_heap(&serialized, base_address, 32, 7)?;
     let heap_id_length = heap.heap_id_length;
 
-    // Type 5 records: hash(4) + heap_id. The B-tree's key is the name hash,
+    // Type 5 records: hash(4) + heap_id. The B-tree's key is the name hash
-    // so records are sorted by (hash, order).
+    // and, for names whose hashes collide, the name (libhdf5 compares them
+    // with `strcmp`): records out of that order are not found by name.
     let mut by_name: Vec<(u32, usize)> = links
         .iter()
         .enumerate()
         .map(|(i, l)| (crate::checksum::jenkins_lookup3(l.name.as_bytes()), i))
         .collect();
-    by_name.sort_unstable();
+    by_name.sort_unstable_by(|&(ha, a), &(hb, b)| {
+        ha.cmp(&hb)
+            .then_with(|| links[a].name.as_bytes().cmp(links[b].name.as_bytes()))
+    });
     let name_records: Vec<Vec<u8>> = by_name
         .iter()
         .map(|&(hash, i)| {
@@ -1057,12 +1112,11 @@ pub(crate) fn build_dense_links(
         .collect();
     let name_bt_addr = heap.btree_addr;
     let mut blob = heap.blob;
-    blob.extend_from_slice(&single_leaf_v2_btree(
+    blob.extend_from_slice(&dense_v2_btree(
         5,
         4 + heap_id_length,
         &name_records,
         name_bt_addr,
-        "links in one group",
     )?);
 
     let link_info_message = if track_order {
@@ -1082,12 +1136,11 @@ pub(crate) fn build_dense_links(
             })
             .collect();
         let order_bt_addr = base_address + blob.len() as u64;
-        blob.extend_from_slice(&single_leaf_v2_btree(
+        blob.extend_from_slice(&dense_v2_btree(
             6,
             8 + heap_id_length,
             &order_records,
             order_bt_addr,
-            "links in one group",
         )?);
         let next_order = by_order.last().map_or(0, |&(o, _)| o + 1);
         serialize_link_info(
@@ -1147,12 +1200,25 @@ fn encode_managed_id(offset: u64, length: u64, max_heap_size: u16, id_length: u1
     id
 }
 
-fn serialize_attribute_info(fh_addr: u64, btree_name_addr: u64) -> Vec<u8> {
+/// Serialize an Attribute Info message (version 0). `order` — the next
+/// creation order to assign and the creation-order index's address — is
+/// present when creation order is tracked and indexed.
+fn serialize_attribute_info(
+    fh_addr: u64,
+    btree_name_addr: u64,
+    order: Option<(u16, u64)>,
+) -> Vec<u8> {
     let mut data = Vec::new();
     data.push(0); // version
-    data.push(0x00); // flags
+    data.push(if order.is_some() { 0x03 } else { 0x00 }); // flags: tracked, indexed
+    if let Some((next, _)) = order {
+        data.extend_from_slice(&next.to_le_bytes());
+    }
     data.extend_from_slice(&fh_addr.to_le_bytes());
     data.extend_from_slice(&btree_name_addr.to_le_bytes());
+    if let Some((_, corder_addr)) = order {
+        data.extend_from_slice(&corder_addr.to_le_bytes());
+    }
     data
 }
 
@@ -1227,8 +1293,7 @@ pub(crate) fn build_vds_dataset_oh(
     dt: &Datatype,
     ds: &Dataspace,
     global_heap_addr: u64,
-    attrs: &[AttributeMessage],
+    attrs: AttrStorage<'_>,
-    dense_blob: Option<&DenseAttrBlob>,
     fill_message: &[u8],
     refcount: u32,
 ) -> Result<Vec<u8>, FormatError> {
@@ -1243,13 +1308,7 @@ pub(crate) fn build_vds_dataset_oh(
     dl.extend_from_slice(&global_heap_addr.to_le_bytes());
     dl.extend_from_slice(&1u32.to_le_bytes()); // object index 1 in the collection
     w.add_message(MessageType::DataLayout, dl);
-    if let Some(blob) = dense_blob {
+    attrs.add_to(&mut w);
-        w.add_message(MessageType::AttributeInfo, blob.attr_info_message.clone());
-    } else {
-        for attr in attrs {
-            w.add_message(MessageType::Attribute, attr.serialize(LENGTH_SIZE));
-        }
-    }
     add_refcount(&mut w, refcount);
     w.serialize()
 }
@@ -1284,7 +1343,8 @@ fn write_undef_offset(buf: &mut Vec<u8>, offset_size: u8) {
 pub struct FileWriter {
     /// The root group's contents (its name is unused).
     root: GroupBuilder,
-    /// Default for groups that do not call [`GroupBuilder::track_order`].
+    /// Default for groups and datasets that do not set their own
+    /// `track_order`.
     track_order: bool,
     /// Global alignment threshold: datasets with raw data >= this many bytes
     /// will have their data aligned to `alignment_bytes`.
@@ -1319,11 +1379,18 @@ struct DsFlat {
     virtual_sources: Option<Vec<VdsMapping>>,
     /// Number of hard links to the dataset.
     refcount: u32,
+    /// Track (and index) attribute creation order.
+    track_order: bool,
 }
 
 /// Convert a DatasetBuilder into a DsFlat, handling VDS (which does not
 /// require a `data` field).
-fn flatten_ds(db: DatasetBuilder, refcount: u32) -> Result<DsFlat, FormatError> {
+fn flatten_ds(
+    db: DatasetBuilder,
+    refcount: u32,
+    default_track_order: bool,
+) -> Result<DsFlat, FormatError> {
+    let track_order = db.track_order.unwrap_or(default_track_order);
     let dt = db.datatype.ok_or(FormatError::DatasetMissingData)?;
     let shape = db.shape.ok_or(FormatError::DatasetMissingShape)?;
     let is_vds = db.virtual_sources.is_some();
@@ -1373,6 +1440,7 @@ fn flatten_ds(db: DatasetBuilder, refcount: u32) -> Result<DsFlat, FormatError>
         alignment: db.alignment,
         virtual_sources: db.virtual_sources,
         refcount,
+        track_order,
     })
 }
 
@@ -1429,10 +1497,12 @@ impl FileWriter {
         self
     }
 
-    /// Track (and index) link creation order in every group that does not
+    /// Track (and index) creation order — of links and attributes in every
-    /// set its own [`GroupBuilder::track_order`], the root included — as
+    /// group that does not set its own [`GroupBuilder::track_order`], the
-    /// h5py's `track_order=True`: libhdf5 then lists members in the order
+    /// root included, and of attributes on every dataset that does not set
-    /// they were added. Off by default (members are listed by name).
+    /// its own [`DatasetBuilder::track_order`] — as h5py's
+    /// `track_order=True`: libhdf5 then lists members and attributes in the
+    /// order they were added. Off by default (they are listed by name).
     pub fn track_order(&mut self, track: bool) -> &mut Self {
         self.track_order = track;
         self
@@ -1503,7 +1573,7 @@ impl FileWriter {
         let all_ds: Vec<DsFlat> = tree
             .datasets
             .into_iter()
-            .map(|(db, refcount)| flatten_ds(db, refcount))
+            .map(|(db, refcount)| flatten_ds(db, refcount, self.track_order))
             .collect::<Result<_, _>>()?;
         let groups: Vec<GrpFlat> = tree
             .groups
@@ -1520,6 +1590,15 @@ impl FileWriter {
             })
             .collect();
 
+        // Refuse up front what dense storage would refuse after the work.
+        let tracked = groups
+            .iter()
+            .map(|g| (g.track_order, g.attrs.len()))
+            .chain(all_ds.iter().map(|d| (d.track_order, d.attrs.len())));
+        for (track, n) in tracked {
+            check_tracked_attr_count(track, n)?;
+        }
+
         // Every datatype must have an on-disk encoding before anything is laid
         // out: `Datatype::serialize` itself cannot report a failure.
         let group_attrs = groups.iter().flat_map(|g| &g.attrs);
@@ -1574,7 +1653,7 @@ impl FileWriter {
             .map(|(gi, g)| {
                 let dummy_links = g.link_messages(&[], &[]);
                 let attr_blob = group_dense[gi]
-                    .then(|| build_dense_attrs(&g.attrs, 0))
+                    .then(|| build_dense_attrs(&g.attrs, 0, g.track_order))
                     .transpose()?;
                 let li = if group_links_dense[gi] {
                     serialize_link_info(
@@ -1590,8 +1669,11 @@ impl FileWriter {
                     &dummy_links,
                     &li,
                     group_links_dense[gi],
-                    &g.attrs,
+                    AttrStorage {
-                    attr_blob.as_ref(),
+                        attrs: &g.attrs,
+                        dense: attr_blob.as_ref(),
+                        track_order: g.track_order,
+                    },
                     g.refcount,
                 )
                 .map(|oh| oh.len())
@@ -1610,7 +1692,7 @@ impl FileWriter {
         let mut dummy_cursor = 0u64;
         for (i, d) in all_ds.iter().enumerate() {
             let dense_blob = ds_dense[i]
-                .then(|| build_dense_attrs(&d.attrs, 0))
+                .then(|| build_dense_attrs(&d.attrs, 0, d.track_order))
                 .transpose()?;
             if is_vds[i] {
                 // VDS: dummy OH with address 0 to get the OH size. The global
@@ -1619,8 +1701,11 @@ impl FileWriter {
                     &d.dt,
                     &d.ds,
                     0, // dummy address
-                    &d.attrs,
+                    AttrStorage {
-                    dense_blob.as_ref(),
+                        attrs: &d.attrs,
+                        dense: dense_blob.as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1659,8 +1744,11 @@ impl FileWriter {
                     &d.ds,
                     &result.layout_message,
                     result.pipeline_message.as_deref(),
-                    &d.attrs,
+                    AttrStorage {
-                    dense_blob.as_ref(),
+                        attrs: &d.attrs,
+                        dense: dense_blob.as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1674,8 +1762,11 @@ impl FileWriter {
                     &d.dt,
                     &d.ds,
                     &d.raw,
-                    &d.attrs,
+                    AttrStorage {
-                    dense_blob.as_ref(),
+                        attrs: &d.attrs,
+                        dense: dense_blob.as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1690,8 +1781,11 @@ impl FileWriter {
                     &d.ds,
                     0,
                     d.raw.len() as u64,
-                    &d.attrs,
+                    AttrStorage {
-                    dense_blob.as_ref(),
+                        attrs: &d.attrs,
+                        dense: dense_blob.as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1735,7 +1829,7 @@ impl FileWriter {
                 group_link_blob_addrs.push(None);
             }
             if group_dense[gi] {
-                let blob = build_dense_attrs(&g.attrs, cursor2 as u64)?;
+                let blob = build_dense_attrs(&g.attrs, cursor2 as u64, g.track_order)?;
                 cursor2 += blob.blob.len();
                 group_dense_blobs.push(Some(blob));
             } else {
@@ -1752,7 +1846,8 @@ impl FileWriter {
                 let addr = cursor2 as u64;
                 cursor2 += sz;
                 if ds_dense[i] {
-                    let blob = build_dense_attrs(&all_ds[i].attrs, cursor2 as u64)?;
+                    let blob =
+                        build_dense_attrs(&all_ds[i].attrs, cursor2 as u64, all_ds[i].track_order)?;
                     cursor2 += blob.blob.len();
                     ds_dense_blobs.push(Some(blob));
                 } else {
@@ -1776,8 +1871,11 @@ impl FileWriter {
                     &d.dt,
                     &d.ds,
                     heap_addr,
-                    &d.attrs,
+                    AttrStorage {
-                    ds_dense_blobs[i].as_ref(),
+                        attrs: &d.attrs,
+                        dense: ds_dense_blobs[i].as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1804,8 +1902,11 @@ impl FileWriter {
                     &d.ds,
                     &result.layout_message,
                     result.pipeline_message.as_deref(),
-                    &d.attrs,
+                    AttrStorage {
-                    ds_dense_blobs[i].as_ref(),
+                        attrs: &d.attrs,
+                        dense: ds_dense_blobs[i].as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1820,8 +1921,11 @@ impl FileWriter {
                     &d.dt,
                     &d.ds,
                     &d.raw,
-                    &d.attrs,
+                    AttrStorage {
-                    ds_dense_blobs[i].as_ref(),
+                        attrs: &d.attrs,
+                        dense: ds_dense_blobs[i].as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1846,8 +1950,11 @@ impl FileWriter {
                     &d.ds,
                     cursor2 as u64,
                     d.raw.len() as u64,
-                    &d.attrs,
+                    AttrStorage {
-                    ds_dense_blobs[i].as_ref(),
+                        attrs: &d.attrs,
+                        dense: ds_dense_blobs[i].as_ref(),
+                        track_order: d.track_order,
+                    },
                     &d.fill_message,
                     d.refcount,
                 )?;
@@ -1915,8 +2022,11 @@ impl FileWriter {
                 &links,
                 &li,
                 link_blob.is_some(),
-                &g.attrs,
+                AttrStorage {
-                group_dense_blobs[gi].as_ref(),
+                    attrs: &g.attrs,
+                    dense: group_dense_blobs[gi].as_ref(),
+                    track_order: g.track_order,
+                },
                 g.refcount,
             )?;
             debug_assert_eq!(oh.len(), group_oh_sizes[gi]);
@@ -2147,6 +2257,92 @@ mod tests {
         assert_eq!(read_dataset_f64(&bytes, "data"), vec![1.0, 2.0, 3.0]);
     }
 
+    /// Attribute names of the object at `path`, in the order the reader
+    /// lists them.
+    fn attr_names(bytes: &[u8], path: &str) -> Vec<String> {
+        let sig = signature::find_signature(bytes).unwrap();
+        let sb = Superblock::parse(bytes, sig).unwrap();
+        let addr = if path == "/" {
+            sb.root_group_address
+        } else {
+            resolve_path_any(bytes, &sb, path).unwrap()
+        };
+        let hdr =
+            ObjectHeader::parse(bytes, addr as usize, sb.offset_size, sb.length_size).unwrap();
+        crate::attribute::extract_attributes_full(bytes, &hdr, sb.offset_size, sb.length_size)
+            .unwrap()
+            .into_iter()
+            .map(|a| a.name)
+            .collect()
+    }
+
+    #[test]
+    fn tracked_attributes_are_read_in_creation_order() {
+        let set = |names: &[String]| -> Vec<(String, AttrValue)> {
+            names
+                .iter()
+                .enumerate()
+                .map(|(i, n)| (n.clone(), AttrValue::I64(i as i64)))
+                .collect()
+        };
+        let compact: Vec<String> = ["zeta", "alpha", "mid"].map(String::from).to_vec();
+        let dense: Vec<String> = (0..30).rev().map(|i| format!("a{i:02}")).collect();
+        let mut fw = FileWriter::new();
+        fw.track_order(true);
+        for (n, v) in set(&compact) {
+            fw.set_root_attr(&n, v);
+        }
+        let ds = fw.create_dataset("dense");
+        ds.with_i32_data(&[1]);
+        for (n, v) in set(&dense) {
+            ds.set_attr(&n, v);
+        }
+        let ds = fw.create_dataset("untracked");
+        ds.with_i32_data(&[1]).track_order(false);
+        for (n, v) in set(&dense) {
+            ds.set_attr(&n, v);
+        }
+        let mut g = fw.create_group("g");
+        g.track_order(false);
+        for (n, v) in set(&compact) {
+            g.set_attr(&n, v);
+        }
+        fw.add_group(g.finish());
+        let bytes = fw.finish().unwrap();
+        assert_eq!(attr_names(&bytes, "/"), compact);
+        assert_eq!(attr_names(&bytes, "dense"), dense);
+        // Without tracking: storage order (inline: as added; dense: hash).
+        assert_eq!(attr_names(&bytes, "g"), compact);
+        let mut by_hash = dense.clone();
+        by_hash.sort_by_key(|n| crate::checksum::jenkins_lookup3(n.as_bytes()));
+        assert_eq!(attr_names(&bytes, "untracked"), by_hash);
+    }
+
+    #[test]
+    fn too_many_tracked_attributes_is_an_error() {
+        // libhdf5 numbers at most 65 535 attributes on an object that
+        // tracks their creation order (a 2-byte field). (`set_attr` looks
+        // for an earlier value, so 65 536 of them through the builder take
+        // a while; build the messages directly.)
+        let attrs: Vec<AttributeMessage> = (0..65_536)
+            .map(|i| build_attr_message(&format!("a{i}"), &AttrValue::I64(i)))
+            .collect();
+        let err = build_dense_attrs(&attrs, 0, true)
+            .err()
+            .unwrap()
+            .to_string();
+        assert!(err.contains("65536 attributes on one object"), "{err}");
+        assert!(build_dense_attrs(&attrs[1..], 0, true).is_ok());
+        assert!(build_dense_attrs(&attrs, 0, false).is_ok());
+        let mut fw = FileWriter::new();
+        let ds = fw.create_dataset("x");
+        ds.with_i32_data(&[1]).track_order(true);
+        for i in 0..20 {
+            ds.set_attr(&format!("a{i}"), AttrValue::I64(i));
+        }
+        assert!(fw.finish().is_ok());
+    }
+
     #[test]
     fn dense_attrs_root_group_self_roundtrip() {
         let mut fw = FileWriter::new();

crates/clawhdf5-format/src/fill_value.rs

@@ -12,7 +12,8 @@
 #[cfg(not(feature = "std"))]
 use alloc::{format, vec, vec::Vec};
 
-use crate::chunked_read::{alloc_output, checked_byte_len, list_chunks};
+use crate::addr::to_usize;
+use crate::chunked_read::{alloc_output, checked_byte_len, list_chunks_in};
 use crate::data_layout::DataLayout;
 use crate::dataspace::Dataspace;
 use crate::error::FormatError;
@@ -116,9 +117,22 @@ pub fn dataset_fill_value_in(
     messages: &[HeaderMessage],
     offset_size: u8,
     length_size: u8,
+) -> Result<Option<Vec<u8>>, FormatError> {
+    dataset_fill_value_from_storage(&file_data, messages, offset_size, length_size)
+}
+
+/// [`dataset_fill_value_in`] with the file behind any
+/// [`Storage`](crate::storage::Storage) (a `&dyn Storage` too). (The trait
+/// is not imported here: its `len` would shadow the slice method in this
+/// module.)
+pub fn dataset_fill_value_from_storage<S: crate::storage::Storage + ?Sized>(
+    file: &S,
+    messages: &[HeaderMessage],
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Option<Vec<u8>>, FormatError> {
     fill_value_from(messages, |msg| {
-        crate::shared_message::message_data_with_sohm(file_data, msg, offset_size, length_size)
+        crate::shared_message::message_data_with_sohm_in(file, msg, offset_size, length_size)
             .map(|data| data.into_owned())
     })
 }
@@ -199,6 +213,30 @@ pub fn read_full_with_fill<E: From<FormatError>>(
     offset_size: u8,
     length_size: u8,
     read: impl FnOnce() -> Result<Vec<u8>, E>,
+) -> Result<Vec<u8>, E> {
+    read_full_with_fill_in(
+        messages,
+        file_data,
+        layout,
+        dataspace,
+        elem_size,
+        offset_size,
+        length_size,
+        read,
+    )
+}
+
+/// [`read_full_with_fill`] over any [`Storage`](crate::storage::Storage).
+#[allow(clippy::too_many_arguments)]
+pub fn read_full_with_fill_in<E: From<FormatError>, S: crate::storage::Storage + ?Sized>(
+    messages: &[HeaderMessage],
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    elem_size: usize,
+    offset_size: u8,
+    length_size: u8,
+    read: impl FnOnce() -> Result<Vec<u8>, E>,
 ) -> Result<Vec<u8>, E> {
     // A dataset with external raw data also has no data address in this
     // file. It is NOT unallocated — its values live elsewhere — so it must
@@ -209,12 +247,12 @@ pub fn read_full_with_fill<E: From<FormatError>>(
     {
         return Err(FormatError::ExternalDataFilesUnsupported.into());
     }
-    let fill = dataset_fill_value_in(file_data, messages, offset_size, length_size)?;
+    let fill = dataset_fill_value_from_storage(file_data, messages, offset_size, length_size)?;
     if !has_storage(layout) {
         return Ok(filled_dataset(dataspace, elem_size, fill.as_deref())?);
     }
     let mut output = read()?;
-    apply_to_unallocated_chunks(
+    apply_to_unallocated_chunks_in(
         &mut output,
         file_data,
         layout,
@@ -240,6 +278,30 @@ pub fn apply_to_unallocated_chunks(
     fill: Option<&[u8]>,
     offset_size: u8,
     length_size: u8,
+) -> Result<(), FormatError> {
+    apply_to_unallocated_chunks_in(
+        output,
+        file_data,
+        layout,
+        dataspace,
+        elem_size,
+        fill,
+        offset_size,
+        length_size,
+    )
+}
+
+/// [`apply_to_unallocated_chunks`] over any [`Storage`](crate::storage::Storage).
+#[allow(clippy::too_many_arguments)]
+pub fn apply_to_unallocated_chunks_in<S: crate::storage::Storage + ?Sized>(
+    output: &mut [u8],
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    elem_size: usize,
+    fill: Option<&[u8]>,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<(), FormatError> {
     let Some(fill) = fill.filter(|f| f.len() == elem_size && !is_default(Some(f))) else {
         return Ok(());
@@ -247,7 +309,7 @@ pub fn apply_to_unallocated_chunks(
     if !matches!(layout, DataLayout::Chunked { .. }) || elem_size == 0 {
         return Ok(());
     }
-    let (chunks, chunk_dims) = list_chunks(
+    let (chunks, chunk_dims) = list_chunks_in(
         file_data,
         layout,
         dataspace,
@@ -256,7 +318,11 @@ pub fn apply_to_unallocated_chunks(
         length_size,
     )?;
     let rank = chunk_dims.len();
-    let ds_dims: Vec<usize> = dataspace.dimensions.iter().map(|&d| d as usize).collect();
+    let ds_dims: Vec<usize> = dataspace
+        .dimensions
+        .iter()
+        .map(|&d| to_usize(d))
+        .collect::<Result<_, _>>()?;
     if rank == 0 || ds_dims.len() != rank || chunk_dims.contains(&0) {
         return Ok(());
     }
@@ -288,7 +354,7 @@ pub fn apply_to_unallocated_chunks(
         let mut cell = 0usize;
         let mut in_range = true;
         for d in 0..rank {
-            let coord = chunk.offsets[d] as usize / chunk_dims[d];
+            let coord = to_usize(chunk.offsets[d])? / chunk_dims[d];
             if coord >= grid[d] {
                 in_range = false;
                 break;
@@ -439,4 +505,37 @@ mod tests {
             .collect();
         assert_eq!(filled, [2, 3, 7, 8]);
     }
+
+    /// Fill values, shared ones in the SOHM heap included, resolve
+    /// identically through a read_at-only CountingStorage.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::object_header::ObjectHeader;
+        use crate::storage::CountingStorage;
+        let file: &[u8] = include_bytes!("../tests/fixtures/shared_fill_value.h5");
+        let sb = crate::superblock::Superblock::parse(file, 0).unwrap();
+        let (os, ls) = (sb.offset_size, sb.length_size);
+        let storage = CountingStorage::new(file.to_vec());
+        let mut shared = 0;
+        let children =
+            crate::group_v2::resolve_group_children(file, &sb, sb.root_group_address).unwrap();
+        assert!(children.len() >= 3);
+        for child in children {
+            let h =
+                ObjectHeader::parse(file, child.object_header_address as usize, os, ls).unwrap();
+            shared += h
+                .messages
+                .iter()
+                .filter(|m| {
+                    m.msg_type == MessageType::FillValue
+                        && crate::shared_message::is_shared(m.flags)
+                })
+                .count();
+            let want = dataset_fill_value_in(file, &h.messages, os, ls);
+            assert_eq!(want, Ok(Some((-7i32).to_le_bytes().to_vec())));
+            let got = dataset_fill_value_from_storage(&storage, &h.messages, os, ls);
+            assert_eq!(got, want, "{}", child.name);
+        }
+        assert!(shared >= 2);
+    }
 }

crates/clawhdf5-format/src/filter_pipeline.rs

@@ -27,7 +27,8 @@ pub const FILTER_LZF: u16 = 32000;
 pub const FILTER_BLOSC: u16 = 32001;
 /// Bitshuffle, optionally with LZ4 or Zstandard (hdf5plugin's `Bitshuffle`).
 pub const FILTER_BITSHUFFLE: u16 = 32008;
-/// ZFP lossy floating-point compression (hdf5plugin's `Zfp`). Not supported.
+/// ZFP lossy (and lossless) compression of numeric arrays (H5Z-ZFP;
+/// hdf5plugin's `Zfp`). Read-only, with the `zfp` feature.
 pub const FILTER_ZFP: u16 = 32013;
 /// Blosc 2 (hdf5plugin's `Blosc2`).
 pub const FILTER_BLOSC2: u16 = 32026;

crates/clawhdf5-format/src/filter_registry.rs

@@ -5,7 +5,8 @@
 //! * **Built-in filters** — a static table of the filters compiled into this
 //!   build: the HDF5 standard filters (deflate, shuffle, Fletcher32, szip,
 //!   N-Bit, scale-offset) and the plugin filters whose cargo features are
-//!   enabled (LZ4, Zstandard, pcodec, LZF, bitshuffle, bzip2, blosc).
+//!   enabled (LZ4, Zstandard, pcodec, LZF, bitshuffle, bzip2, blosc,
+//!   blosc2, zfp).
 //!   [`builtin_filters`] lists them.
 //! * **Registered filters** (`std` only) — codecs the application supplies
 //!   for any other ID with [`register_filter`] (a [`FilterCodec`], or just a
@@ -161,12 +162,12 @@ pub fn known_filter(id: u16) -> Option<(&'static str, Option<&'static str>)> {
         32001 => ("Blosc", Some("blosc")),
         32004 => ("LZ4", Some("lz4")),
         32008 => ("bitshuffle", Some("bitshuffle")),
-        32013 => ("ZFP", None),
+        32013 => ("ZFP", Some("zfp")),
         32015 => ("Zstandard", Some("zstd")),
         32019 => ("JPEG", None),
         32022 => ("BitGroom", None),
         32023 => ("Granular BitRound", None),
-        32026 => ("Blosc2", None),
+        32026 => ("Blosc2", Some("blosc2")),
         _ => return None,
     })
 }
@@ -187,6 +188,22 @@ pub fn is_filter_available(id: u16) -> bool {
     }
 }
 
+/// Whether chunks filtered with `id` may be decoded by a codec the
+/// application registered (whose stored sizes this crate cannot bound).
+pub(crate) fn may_be_registered(id: u16) -> bool {
+    if builtin_filter(id).is_some_and(|b| !b.is_shared()) {
+        return false;
+    }
+    #[cfg(feature = "std")]
+    {
+        registered(id).is_some()
+    }
+    #[cfg(not(feature = "std"))]
+    {
+        false
+    }
+}
+
 #[cfg(feature = "std")]
 mod custom {
     use super::FilterCodec;
@@ -452,12 +469,14 @@ pub(crate) mod tests {
     #[test]
     fn unsupported_filter_error_names_the_filter() {
         let msg = FormatError::UnsupportedFilter(32026).to_string();
+        assert!(msg.contains("Blosc2") && msg.contains("`blosc2`"), "{msg}");
+        let msg = FormatError::UnsupportedFilter(32013).to_string();
+        assert!(msg.contains("ZFP") && msg.contains("`zfp`"), "{msg}");
+        let msg = FormatError::UnsupportedFilter(32019).to_string();
         assert!(
-            msg.contains("Blosc2") && msg.contains("not implemented"),
+            msg.contains("JPEG") && msg.contains("not implemented"),
             "{msg}"
         );
-        let msg = FormatError::UnsupportedFilter(32013).to_string();
-        assert!(msg.contains("ZFP"), "{msg}");
         let msg = FormatError::UnsupportedFilter(32000).to_string();
         assert!(msg.contains("LZF") && msg.contains("`lzf`"), "{msg}");
         assert_eq!(

crates/clawhdf5-format/src/filters_bitshuffle.rs

@@ -218,12 +218,20 @@ pub(crate) fn bitshuffle_decode(
 }
 
 /// Decode Zstandard frames into exactly `dst`, failing if they hold more.
+///
+/// ruzstd reserves a frame's declared window (by default up to 100 MiB)
+/// before decoding it, so the window is capped at what the output could
+/// need: twice `dst` (window sizes are rounded up), and at least 128 KiB.
+/// The encoders behind these filters (c-blosc, c-blosc2, bitshuffle)
+/// compress each block in one call with its size known, so libzstd's
+/// window never exceeds the block.
 #[cfg(any(feature = "bitshuffle", feature = "blosc"))]
 pub(crate) fn zstd_decode_into(
     decoder: &mut ruzstd::decoding::FrameDecoder,
     frames: &[u8],
     dst: &mut [u8],
 ) -> Result<usize, FormatError> {
+    decoder.set_max_window_size((2 * dst.len()).max(1 << 17) as u64);
     decoder
         .decode_all(frames, dst)
         .map_err(|e| FormatError::DecompressionError(format!("zstd: {e}")))

crates/clawhdf5-format/src/filters_blosc.rs

@@ -49,7 +49,7 @@ fn le32(b: &[u8], at: usize) -> Result<usize, FormatError> {
 
 /// The codec inside a Blosc frame (flags bits 5-7).
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]
-enum Codec {
+pub(crate) enum Codec {
     BloscLz,
     Lz4,
     Snappy,
@@ -58,7 +58,7 @@ enum Codec {
 }
 
 impl Codec {
-    fn from_flags(flags: u8) -> Result<Codec, FormatError> {
+    pub(crate) fn from_flags(flags: u8) -> Result<Codec, FormatError> {
         match flags >> 5 {
             0 => Ok(Codec::BloscLz),
             1 => Ok(Codec::Lz4),
@@ -71,7 +71,7 @@ impl Codec {
 }
 
 /// Decode one codec stream into exactly `dst`.
-fn decode_stream(
+pub(crate) fn decode_stream(
     codec: Codec,
     src: &[u8],
     dst: &mut [u8],

crates/clawhdf5-format/src/filters_bzip2.rs

@@ -4,6 +4,7 @@
 //! the compression level). Decoded with the `bzip2` crate's default backend,
 //! `libbz2-rs-sys`, a pure-Rust port of libbzip2.
 
+use crate::addr::saturating_usize;
 use crate::error::FormatError;
 use crate::filter_registry::FilterContext;
 
@@ -28,7 +29,7 @@ pub(crate) fn bzip2_decode(input: &[u8], ctx: &FilterContext<'_>) -> Result<Vec<
     loop {
         let (in_before, out_before) = (dec.total_in(), dec.total_out());
         let status = dec
-            .decompress_vec(&input[in_before as usize..], &mut out)
+            .decompress_vec(&input[saturating_usize(in_before)..], &mut out)
             .map_err(|e| err(&e.to_string()))?;
         if out.len() > limit {
             return Err(err("output exceeds the chunk size"));
@@ -43,7 +44,7 @@ pub(crate) fn bzip2_decode(input: &[u8], ctx: &FilterContext<'_>) -> Result<Vec<
                 .max(1);
             out.try_reserve_exact(grow)
                 .map_err(|_| err("cannot allocate the output buffer"))?;
-        } else if dec.total_in() as usize >= input.len()
+        } else if saturating_usize(dec.total_in()) >= input.len()
             || (dec.total_in(), dec.total_out()) == (in_before, out_before)
         {
             return Err(err("truncated stream"));
@@ -61,7 +62,7 @@ pub(crate) fn bzip2_encode(input: &[u8], ctx: &FilterContext<'_>) -> Result<Vec<
     // bzip2's worst case is about 1% + 600 bytes over the input.
     let mut out = Vec::with_capacity(input.len() + input.len() / 100 + 600);
     loop {
-        let consumed = enc.total_in() as usize;
+        let consumed = saturating_usize(enc.total_in());
         let status = enc
             .compress_vec(&input[consumed..], &mut out, Action::Finish)
             .map_err(|e| cerr(e.to_string()))?;

crates/clawhdf5-format/src/fixed_array.rs

@@ -6,19 +6,23 @@ extern crate alloc;
 #[cfg(not(feature = "std"))]
 use alloc::{format, vec, vec::Vec};
 
+use crate::addr::to_usize;
 use crate::chunk_grid::ChunkGrid;
 use crate::chunked_read::ChunkInfo;
 use crate::error::FormatError;
+use crate::storage::{PAGED_BLOCK_ONE_READ_MAX, Storage, Window, len_usize, read_exact_at};
 
 /// Verify the Jenkins lookup3 checksum stored immediately after
-/// `data[start..end]`, as every Fixed Array structure carries one.
+/// `data[start..end]`, as every Fixed Array structure carries one. `w` is
+/// a window of the file and `start`/`end` are relative to it.
 ///
 /// A corrupt chunk index silently yields addresses pointing at the wrong
 /// bytes, so a mismatch has to be an error rather than a shrug: without this
 /// the damage surfaces as plausible-looking data from the wrong chunk.
 #[cfg(feature = "checksum")]
-fn verify_checksum(data: &[u8], start: usize, end: usize) -> Result<(), FormatError> {
+fn verify_checksum(w: &Window<'_>, start: usize, end: usize) -> Result<(), FormatError> {
-    ensure_len(data, end, 4)?;
+    w.ensure(end, 4)?;
+    let data = &w.bytes;
     let stored = u32::from_le_bytes([data[end], data[end + 1], data[end + 2], data[end + 3]]);
     let computed = crate::checksum::jenkins_lookup3(&data[start..end]);
     if computed != stored {
@@ -31,7 +35,7 @@ fn verify_checksum(data: &[u8], start: usize, end: usize) -> Result<(), FormatEr
 }
 
 #[cfg(not(feature = "checksum"))]
-fn verify_checksum(_data: &[u8], _start: usize, _end: usize) -> Result<(), FormatError> {
+fn verify_checksum(_w: &Window<'_>, _start: usize, _end: usize) -> Result<(), FormatError> {
     Ok(())
 }
 
@@ -73,19 +77,6 @@ fn read_length(data: &[u8], pos: usize, size: u8) -> Result<u64, FormatError> {
     read_offset(data, pos, size)
 }
 
-fn ensure_len(data: &[u8], offset: usize, needed: usize) -> Result<(), FormatError> {
-    if offset
-        .checked_add(needed)
-        .is_none_or(|end| end > data.len())
-    {
-        return Err(FormatError::UnexpectedEof {
-            expected: offset.saturating_add(needed),
-            available: data.len(),
-        });
-    }
-    Ok(())
-}
-
 fn is_undefined(data: &[u8], pos: usize, size: u8) -> bool {
     let s = size as usize;
     if pos + s > data.len() {
@@ -101,13 +92,24 @@ impl FixedArrayHeader {
         offset: usize,
         offset_size: u8,
         length_size: u8,
+    ) -> Result<Self, FormatError> {
+        Self::parse_in(file_data, offset as u64, offset_size, length_size)
+    }
+
+    /// [`Self::parse`] over any [`Storage`]: one read of the header.
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
     ) -> Result<Self, FormatError> {
         // FAHD signature(4) + version(1) + client_id(1) + element_size(1) +
         // max_nelmts_bits(1) + num_elements(length_size) + data_block_addr(offset_size) + checksum(4)
         let min_size = 4 + 1 + 1 + 1 + 1 + length_size as usize + offset_size as usize + 4;
-        ensure_len(file_data, offset, min_size)?;
+        let w = Window::read(file, offset, min_size)?;
+        w.ensure(0, min_size)?;
 
-        let d = &file_data[offset..];
+        let d: &[u8] = &w.bytes;
         if &d[0..4] != b"FAHD" {
             return Err(FormatError::ChunkedReadError(
                 "invalid Fixed Array header signature".into(),
@@ -130,7 +132,7 @@ impl FixedArrayHeader {
         pos += length_size as usize;
         let data_block_address = read_offset(d, pos, offset_size)?;
         pos += offset_size as usize;
-        verify_checksum(file_data, offset, offset + pos)?;
+        verify_checksum(&w, 0, pos)?;
 
         Ok(FixedArrayHeader {
             client_id,
@@ -156,15 +158,39 @@ pub fn read_fixed_array_chunks(
     chunk_dimensions: &[u32],
     element_size: u32,
     offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<ChunkInfo>, FormatError> {
+    read_fixed_array_chunks_in(
+        &file_data,
+        header,
+        dataset_dims,
+        max_dims,
+        chunk_dimensions,
+        element_size,
+        offset_size,
+        length_size,
+    )
+}
+
+/// [`read_fixed_array_chunks`] over any [`Storage`]: one read of the data
+/// block's prefix, one of the whole data block (pages included).
+#[allow(clippy::too_many_arguments)]
+pub fn read_fixed_array_chunks_in<S: Storage + ?Sized>(
+    file: &S,
+    header: &FixedArrayHeader,
+    dataset_dims: &[u64],
+    max_dims: Option<&[u64]>,
+    chunk_dimensions: &[u32],
+    element_size: u32,
+    offset_size: u8,
     _length_size: u8,
 ) -> Result<Vec<ChunkInfo>, FormatError> {
-    let db_offset = header.data_block_address as usize;
+    let file_len = len_usize(file);
+    let db_offset = to_usize(header.data_block_address)?;
 
     // Parse data block header: FADB(4) + version(1) + client_id(1) + header_address(offset_size)
     let db_header_size = 4 + 1 + 1 + offset_size as usize;
-    ensure_len(file_data, db_offset, db_header_size)?;
+    let d = read_exact_at(file, db_offset as u64, db_header_size)?;
-
-    let d = &file_data[db_offset..];
     if &d[0..4] != b"FADB" {
         return Err(FormatError::ChunkedReadError(
             "invalid Fixed Array data block signature".into(),
@@ -174,11 +200,11 @@ pub fn read_fixed_array_chunks(
     // Elements start immediately after the data block prefix.
     let elements_start = db_offset + db_header_size;
 
-    let num_elements = header.num_elements as usize;
+    let num_elements = to_usize(header.num_elements)?;
     // A chunk index cannot describe more elements than the file has bytes (each
     // element occupies at least `offset_size` bytes). Reject a corrupt count
     // before it can drive a huge loop or overflow an offset computation.
-    if num_elements > file_data.len() {
+    if num_elements > file_len {
         return Err(FormatError::ChunkedReadError(
             "Fixed Array element count exceeds file size".into(),
         ));
@@ -208,11 +234,15 @@ pub fn read_fixed_array_chunks(
         chunk_dimensions.iter().map(|&d| d as u64).product::<u64>() * element_size as u64;
 
     let mut chunks = Vec::new();
-    let push_element =
+    // `rel` is relative to the data block, whose bytes are in `w`.
-        |i: usize, abs: usize, chunks: &mut Vec<ChunkInfo>| -> Result<(), FormatError> {
+    let push_element = |w: &Window<'_>,
+                        i: usize,
+                        rel: usize,
+                        chunks: &mut Vec<ChunkInfo>|
+     -> Result<(), FormatError> {
         if let Some((address, chunk_size, filter_mask)) = parse_fa_element(
-                file_data,
+            w,
-                abs,
+            rel,
             header.client_id,
             offset_size,
             header.element_size,
@@ -246,10 +276,16 @@ pub fn read_fixed_array_chunks(
 
     if !is_paged {
         // Non-paged: prefix, then `num_elements` elements packed directly,
-        // then a checksum over both.
+        // then a checksum over both. One window holds all of it (or ends at
-        verify_checksum(file_data, db_offset, elem_at(elements_start, num_elements)?)?;
+        // the end of the file), so its bounds checks are the whole-file ones.
+        let end = elem_at(elements_start, num_elements)?;
+        // The checksum's bounds check comes first: make it before reading.
+        #[cfg(feature = "checksum")]
+        Window::check_extent(file, db_offset as u64, end - db_offset, 4)?;
+        let w = Window::read(file, db_offset as u64, end.saturating_add(4) - db_offset)?;
+        verify_checksum(&w, 0, end - db_offset)?;
         for i in 0..num_elements {
-            push_element(i, elem_at(elements_start, i)?, &mut chunks)?;
+            push_element(&w, i, elem_at(elements_start, i)? - db_offset, &mut chunks)?;
         }
         return Ok(chunks);
     }
@@ -272,22 +308,40 @@ pub fn read_fixed_array_chunks(
         .and_then(|x| x.checked_add(4))
         .ok_or_else(stride_overflow)?;
 
-    if bitmap_start + bitmap_size > file_data.len() {
+    if bitmap_start + bitmap_size > file_len {
         return Err(FormatError::UnexpectedEof {
             expected: bitmap_start + bitmap_size,
-            available: file_data.len(),
+            available: file_len,
         });
     }
+    // The whole data block in one window when it is small: every page slot
+    // is at most `page_stride` bytes, so every position checked below lies
+    // inside it (or past the end of the file). A larger block is read as its
+    // prefix and bitmap, then each page in use on its own.
+    let block_len = (pages_start - db_offset).saturating_add(npages.saturating_mul(page_stride));
+    let whole = if block_len <= PAGED_BLOCK_ONE_READ_MAX {
+        Some(Window::read(file, db_offset as u64, block_len)?)
+    } else {
+        None
+    };
+    let head_w;
+    let head = match &whole {
+        Some(w) => w,
+        None => {
+            head_w = Window::read(file, db_offset as u64, pages_start - db_offset)?;
+            &head_w
+        }
+    };
     // The prefix and page bitmap are covered by their own checksum, and each
     // initialised page by one of its own.
-    verify_checksum(file_data, db_offset, bitmap_start + bitmap_size)?;
+    verify_checksum(head, 0, bitmap_start + bitmap_size - db_offset)?;
 
     for p in 0..npages {
         let page_first = p * page_nelmts; // < num_elements, cannot overflow
         let page_count = core::cmp::min(page_nelmts, num_elements - page_first);
 
         // Check the page-init bit (MSB-first within each byte).
-        let bit_byte = file_data[bitmap_start + p / 8];
+        let bit_byte = head.bytes[bitmap_start + p / 8 - db_offset];
         let bit_mask = 1u8 << (7 - (p % 8));
         if bit_byte & bit_mask == 0 {
             continue; // entire page unallocated
@@ -297,21 +351,33 @@ pub fn read_fixed_array_chunks(
             .checked_mul(page_stride)
             .and_then(|o| pages_start.checked_add(o))
             .ok_or_else(stride_overflow)?;
-        verify_checksum(file_data, page_off, elem_at(page_off, page_count)?)?;
+        let page_end = elem_at(page_off, page_count)?;
+        // `w` holds the page from `base` on (positions below are relative
+        // to it).
+        let page_w;
+        let (w, base) = match &whole {
+            Some(w) => (w, db_offset),
+            None => {
+                page_w =
+                    Window::read(file, page_off as u64, page_end.saturating_add(4) - page_off)?;
+                (&page_w, page_off)
+            }
+        };
+        verify_checksum(w, page_off - base, page_end - base)?;
         for e in 0..page_count {
-            push_element(page_first + e, elem_at(page_off, e)?, &mut chunks)?;
+            push_element(w, page_first + e, elem_at(page_off, e)? - base, &mut chunks)?;
         }
     }
 
     Ok(chunks)
 }
 
-/// Parse a single Fixed Array element at absolute file offset `abs`.
+/// Parse a single Fixed Array element at offset `abs` of the window `w`.
 ///
 /// Returns `Some((address, chunk_size, filter_mask))` for an allocated chunk, or
 /// `None` if the element is undefined (an unallocated chunk, address all-`0xFF`).
 fn parse_fa_element(
-    file_data: &[u8],
+    w: &Window<'_>,
     abs: usize,
     client_id: u8,
     offset_size: u8,
@@ -321,12 +387,8 @@ fn parse_fa_element(
     let os = offset_size as usize;
     if client_id == 0 {
         // Non-filtered: element is just the chunk address.
-        if abs + os > file_data.len() {
+        w.ensure(abs, os)?;
-            return Err(FormatError::UnexpectedEof {
+        let file_data: &[u8] = &w.bytes;
-                expected: abs + os,
-                available: file_data.len(),
-            });
-        }
         if is_undefined(file_data, abs, offset_size) {
             return Ok(None);
         }
@@ -341,17 +403,14 @@ fn parse_fa_element(
             ));
         }
         let chunk_size_bytes = es - os - 4;
-        if abs + es > file_data.len() {
+        w.ensure(abs, es)?;
-            return Err(FormatError::UnexpectedEof {
+        let file_data: &[u8] = &w.bytes;
-                expected: abs + es,
-                available: file_data.len(),
-            });
-        }
         if is_undefined(file_data, abs, offset_size) {
             return Ok(None);
         }
         let address = read_offset(file_data, abs, offset_size)?;
-        let chunk_size = read_variable_length(&file_data[abs + os..], chunk_size_bytes)?;
+        let chunk_size =
+            read_variable_length(&file_data[abs + os..abs + es - 4], chunk_size_bytes)?;
         let fm_off = abs + os + chunk_size_bytes;
         let filter_mask = u32::from_le_bytes([
             file_data[fm_off],
@@ -813,4 +872,126 @@ mod tests {
             .collect();
         assert_eq!(got, expect);
     }
+
+    /// A fixed array (header at 0x100, data block at 0x200) of `n` chunks,
+    /// filtered or not, paged when `n` exceeds `1 << page_bits`; every
+    /// page initialised except page 1.
+    fn build_fixed_array(n: usize, filtered: bool, page_bits: u8) -> Vec<u8> {
+        let os = 8usize;
+        let es = if filtered { os + 4 + 4 } else { os };
+        let (fahd, db) = (0x100usize, 0x200usize);
+        let mut f = vec![0u8; 0x2000];
+        f[fahd..fahd + 4].copy_from_slice(b"FAHD");
+        f[fahd + 5] = u8::from(filtered);
+        f[fahd + 6] = es as u8;
+        f[fahd + 7] = page_bits;
+        f[fahd + 8..fahd + 16].copy_from_slice(&(n as u64).to_le_bytes());
+        f[fahd + 16..fahd + 24].copy_from_slice(&(db as u64).to_le_bytes());
+        stamp_checksum(&mut f, fahd, fahd + 24);
+        f[db..db + 4].copy_from_slice(b"FADB");
+        f[db + 5] = u8::from(filtered);
+        f[db + 6..db + 14].copy_from_slice(&(fahd as u64).to_le_bytes());
+        let elems = db + 6 + os;
+        let write = |f: &mut Vec<u8>, at: usize, i: usize| {
+            let addr = if i == 2 {
+                u64::MAX
+            } else {
+                0x1000 + i as u64 * 0x100
+            };
+            f[at..at + os].copy_from_slice(&addr.to_le_bytes());
+            if filtered {
+                f[at + os..at + os + 4].copy_from_slice(&(100 + i as u32).to_le_bytes());
+                f[at + os + 4..at + os + 8].copy_from_slice(&(i as u32 & 1).to_le_bytes());
+            }
+        };
+        let page = 1usize << page_bits;
+        if n <= page {
+            for i in 0..n {
+                write(&mut f, elems + i * es, i);
+            }
+            stamp_checksum(&mut f, db, elems + n * es);
+        } else {
+            let npages = n.div_ceil(page);
+            let bitmap = npages.div_ceil(8);
+            for p in 0..npages {
+                if p != 1 {
+                    f[elems + p / 8] |= 0x80 >> (p % 8);
+                }
+            }
+            stamp_checksum(&mut f, db, elems + bitmap);
+            let pages_start = elems + bitmap + 4;
+            for p in (0..npages).filter(|&p| p != 1) {
+                let at = pages_start + p * (page * es + 4);
+                let count = page.min(n - p * page);
+                for e in 0..count {
+                    write(&mut f, at + e * es, p * page + e);
+                }
+                stamp_checksum(&mut f, at, at + count * es);
+            }
+        }
+        f
+    }
+
+    /// Non-paged and paged, filtered and unfiltered arrays, cut at every
+    /// length through the data block and with a damaged byte, read
+    /// identically through a `read_at`-only storage.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::storage::CountingStorage;
+        for (n, filtered, bits) in [(3, false, 10), (3, true, 10), (11, false, 2), (11, true, 2)] {
+            let full = build_fixed_array(n, filtered, bits);
+            let es = if filtered { 16 } else { 8 };
+            let dims = [n as u64 * 20];
+            let h = FixedArrayHeader::parse(&full, 0x100, 8, 8).unwrap();
+            let chunks = read_fixed_array_chunks(&full, &h, &dims, None, &[20], 8, 8, 8).unwrap();
+            // Chunk 2 is unallocated, and so is page 1 of a paged array.
+            let expect = if n > 4 { n - 1 - 4 } else { n - 1 };
+            assert_eq!(chunks.len(), expect);
+            let mut files = Vec::new();
+            for cut in (0x100..0x200 + 40 + n * (es + 4) + 16).step_by(3) {
+                files.push(full[..cut].to_vec());
+            }
+            for at in [0x104, 0x210, 0x21a, 0x230] {
+                let mut damaged = full.clone();
+                damaged[at] ^= 1;
+                files.push(damaged);
+            }
+            files.push(full);
+            for f in files {
+                let storage = CountingStorage::new(f.clone());
+                let want = FixedArrayHeader::parse(&f, 0x100, 8, 8);
+                let got = FixedArrayHeader::parse_in(&storage, 0x100, 8, 8);
+                assert_eq!(format!("{got:?}"), format!("{want:?}"));
+                let Ok(h) = want else { continue };
+                let want = read_fixed_array_chunks(&f, &h, &dims, None, &[20], 8, 8, 8);
+                let got = read_fixed_array_chunks_in(&storage, &h, &dims, None, &[20], 8, 8, 8);
+                assert_eq!(format!("{got:?}"), format!("{want:?}"), "{} bytes", f.len());
+            }
+        }
+    }
+
+    /// A header whose element count stretches its data block (one checksum
+    /// over the whole block) far past the end of a 16 MiB file: the
+    /// checksum's bounds check fails before the block is read, with the
+    /// slice read's error.
+    #[cfg(feature = "checksum")]
+    #[test]
+    fn oversized_block_fails_before_reading() {
+        use crate::storage::CountingStorage;
+        let mut f = build_fixed_array(3, false, 10);
+        f.resize(16 << 20, 0);
+        let mut h = FixedArrayHeader::parse(&f, 0x100, 8, 8).unwrap();
+        h.max_nelmts_bits = 30;
+        h.num_elements = 4 << 20;
+        let dims = [h.num_elements * 20];
+        let want = read_fixed_array_chunks(&f, &h, &dims, None, &[20], 8, 8, 8);
+        assert!(
+            matches!(want, Err(FormatError::UnexpectedEof { .. })),
+            "{want:?}"
+        );
+        let storage = CountingStorage::new(f);
+        let got = read_fixed_array_chunks_in(&storage, &h, &dims, None, &[20], 8, 8, 8);
+        assert_eq!(format!("{got:?}"), format!("{want:?}"));
+        assert!(storage.bytes_read() < 64, "{} bytes", storage.bytes_read());
+    }
 }

crates/clawhdf5-format/src/fractal_heap.rs

@@ -6,9 +6,11 @@ use alloc::{format, vec::Vec};
 #[cfg(feature = "checksum")]
 use byteorder::{ByteOrder, LittleEndian};
 
-use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records};
+use crate::addr::to_usize;
+use crate::btree_v2::{BTreeV2Header, find_btree_v2_records_in};
 use crate::error::FormatError;
 use crate::filter_pipeline::FilterPipeline;
+use crate::storage::{Storage, Window, len_usize, read_exact_at};
 
 /// Parsed fractal heap header (signature "FRHP").
 #[derive(Debug, Clone)]
@@ -138,6 +140,36 @@ impl FractalHeapHeader {
         offset_size: u8,
         length_size: u8,
     ) -> Result<FractalHeapHeader, FormatError> {
+        Self::parse_in(file_data, offset as u64, offset_size, length_size)
+    }
+
+    /// [`Self::parse`] over any [`Storage`]: one read of the header (two
+    /// when it holds an I/O filter pipeline).
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
+    ) -> Result<FractalHeapHeader, FormatError> {
+        // Every field up to the checksum, without and with the filter
+        // information; the window holds all of it (or ends at the end of
+        // the file), so its bounds checks are the whole-file ones.
+        let (os, ls) = (usize::from(offset_size), usize::from(length_size));
+        let unfiltered_len = 26 + 12 * ls + 3 * os;
+        let mut w = Window::read(file, offset, unfiltered_len)?;
+        if w.bytes.len() == unfiltered_len {
+            let filter_len = usize::from(u16::from_le_bytes([w.bytes[7], w.bytes[8]]));
+            if filter_len > 0 {
+                w = Window::read(file, offset, unfiltered_len + ls + 4 + filter_len)?;
+            }
+        }
+        let ensure_len = |_: &[u8], pos: usize, needed: usize| w.ensure(pos, needed);
+        let read_offset = |_: &[u8], pos: usize, size: u8| {
+            w.ensure(pos, usize::from(size))?;
+            read_offset(&w.bytes, pos, size)
+        };
+        let file_data: &[u8] = &w.bytes;
+        let offset = 0usize;
         ensure_len(file_data, offset, 5)?;
         if &file_data[offset..offset + 4] != b"FRHP" {
             return Err(FormatError::InvalidFractalHeapSignature);
@@ -148,9 +180,6 @@ impl FractalHeapHeader {
             return Err(FormatError::InvalidFractalHeapVersion(version));
         }
 
-        let os = offset_size as usize;
-        let ls = length_size as usize;
-
         let mut pos = offset + 5;
         ensure_len(file_data, pos, 2)?;
         let heap_id_length = u16::from_le_bytes([file_data[pos], file_data[pos + 1]]);
@@ -354,6 +383,17 @@ impl FractalHeapHeader {
         id_bytes: &[u8],
         offset_size: u8,
     ) -> Result<Vec<u8>, FormatError> {
+        self.read_managed_object_in(file_data, id_bytes, offset_size)
+    }
+
+    /// [`Self::read_managed_object`] over any [`Storage`].
+    pub fn read_managed_object_in<S: Storage + ?Sized>(
+        &self,
+        file_data: &S,
+        id_bytes: &[u8],
+        offset_size: u8,
+    ) -> Result<Vec<u8>, FormatError> {
+        crate::lookup_stats::heap_object_read();
         let Some(&first) = id_bytes.first() else {
             return Err(FormatError::UnexpectedEof {
                 expected: 1,
@@ -383,7 +423,11 @@ impl FractalHeapHeader {
     }
 
     /// Read a huge object (heap ID type 1).
-    fn read_huge_object(&self, file_data: &[u8], id: &[u8]) -> Result<Vec<u8>, FormatError> {
+    fn read_huge_object<S: Storage + ?Sized>(
+        &self,
+        file: &S,
+        id: &[u8],
+    ) -> Result<Vec<u8>, FormatError> {
         let os = usize::from(self.offset_size);
         let ls = usize::from(self.length_size);
         // (address, stored length, filter mask, decoded length); the last two
@@ -414,18 +458,17 @@ impl FractalHeapHeader {
             let key_len = (usize::from(self.heap_id_length).saturating_sub(1)).min(8);
             ensure_len(id, 1, key_len)?;
             let key = le_uint(&id[1..1 + key_len]);
-            self.find_huge_record(file_data, key)?
+            self.find_huge_record(file, key)?
         };
 
         let start = usize::try_from(addr).map_err(|_| heap_error("huge object address"))?;
         let len = usize::try_from(stored_len).map_err(|_| heap_error("huge object length"))?;
-        ensure_len(file_data, start, len)?;
+        let stored = read_exact_at(file, start as u64, len)?;
-        let stored = &file_data[start..start + len];
         match &self.filter_pipeline {
-            None => Ok(stored.to_vec()),
+            None => Ok(stored.into_owned()),
             Some(pipeline) => {
                 let mem = usize::try_from(mem_len).map_err(|_| heap_error("huge object size"))?;
-                let out = crate::filters::decompress_chunk_masked(stored, pipeline, mem, 1, mask)?;
+                let out = crate::filters::decompress_chunk_masked(&stored, pipeline, mem, 1, mask)?;
                 if out.len() != mem {
                     return Err(heap_error("filtered huge object decoded to the wrong size"));
                 }
@@ -436,9 +479,9 @@ impl FractalHeapHeader {
 
     /// Look up huge object `key` in the huge-object v2 B-tree, returning
     /// (address, stored length, filter mask, decoded length).
-    fn find_huge_record(
+    fn find_huge_record<S: Storage + ?Sized>(
         &self,
-        file_data: &[u8],
+        file: &S,
         key: u64,
     ) -> Result<(u64, u64, u32, u64), FormatError> {
         if is_undefined(self.huge_btree_address, self.offset_size) {
@@ -446,9 +489,9 @@ impl FractalHeapHeader {
                 "huge object ID but the heap has no huge-object index",
             ));
         }
-        let hdr = BTreeV2Header::parse(
+        let hdr = BTreeV2Header::parse_in(
-            file_data,
+            file,
-            self.huge_btree_address as usize,
+            self.huge_btree_address,
             self.offset_size,
             self.length_size,
         )?;
@@ -463,8 +506,12 @@ impl FractalHeapHeader {
         if hdr.tree_type != expected_type || usize::from(hdr.record_size) < rec_len {
             return Err(heap_error("unexpected huge-object B-tree record type"));
         }
-        let records =
+        // Records are ordered by ID (the last field): descend to the ones
-            collect_btree_v2_records(file_data, &hdr, self.offset_size, self.length_size)?;
+        // equal to `key` instead of reading the whole index.
+        let id_at = rec_len - ls;
+        let records = find_btree_v2_records_in(file, &hdr, self.offset_size, &mut |r| {
+            le_uint(&r[id_at..id_at + ls]).cmp(&key)
+        })?;
         for rec in &records {
             let d = &rec.data;
             if d.len() < rec_len {
@@ -513,9 +560,9 @@ impl FractalHeapHeader {
     }
 
     /// Read a managed object (heap ID type 0).
-    fn read_heap_managed(
+    fn read_heap_managed<S: Storage + ?Sized>(
         &self,
-        file_data: &[u8],
+        file_data: &S,
         id_bytes: &[u8],
         offset_size: u8,
     ) -> Result<Vec<u8>, FormatError> {
@@ -533,24 +580,24 @@ impl FractalHeapHeader {
             self.read_from_direct_block(
                 file_data,
                 DirectBlock {
-                    addr: self.root_block_address as usize,
+                    addr: to_usize(self.root_block_address)?,
                     size: self.starting_block_size,
                     heap_offset: 0,
                     filtered_size: self.root_direct_block_filtered_size,
                     filter_mask: self.root_direct_block_filter_mask,
                 },
                 heap_offset,
-                obj_len as usize,
+                to_usize(obj_len)?,
             )
         } else {
             // Root is an indirect block — limit recursion to 64 levels
             self.read_from_indirect_block(
                 file_data,
-                self.root_block_address as usize,
+                to_usize(self.root_block_address)?,
                 self.current_rows_in_root_indirect_block,
                 0, // block offset
                 heap_offset,
-                obj_len as usize,
+                to_usize(obj_len)?,
                 offset_size,
                 64, // max recursion depth
             )
@@ -563,27 +610,27 @@ impl FractalHeapHeader {
     /// header), so we just add it to the block address minus the block's heap
     /// offset. A filtered heap stores each direct block (header included)
     /// through its filter pipeline, so the block is decoded first.
-    fn read_from_direct_block(
+    fn read_from_direct_block<S: Storage + ?Sized>(
         &self,
-        file_data: &[u8],
+        file: &S,
         block: DirectBlock,
         target_offset: u64,
         length: usize,
     ) -> Result<Vec<u8>, FormatError> {
         if target_offset < block.heap_offset {
             return Err(FormatError::UnexpectedEof {
-                expected: block.heap_offset as usize,
+                expected: to_usize(block.heap_offset)?,
-                available: target_offset as usize,
+                available: to_usize(target_offset)?,
             });
         }
-        let local_offset = (target_offset - block.heap_offset) as usize;
+        let local_offset = to_usize(target_offset - block.heap_offset)?;
         if let Some(pipeline) = &self.filter_pipeline {
             let stored_len = usize::try_from(block.filtered_size)
                 .map_err(|_| heap_error("direct block size"))?;
             let size = usize::try_from(block.size).map_err(|_| heap_error("direct block size"))?;
-            ensure_len(file_data, block.addr, stored_len)?;
+            let stored = read_exact_at(file, block.addr as u64, stored_len)?;
             let decoded = crate::filters::decompress_chunk_masked(
-                &file_data[block.addr..block.addr + stored_len],
+                &stored,
                 pipeline,
                 size,
                 1,
@@ -597,17 +644,16 @@ impl FractalHeapHeader {
             .checked_add(local_offset)
             .ok_or(FormatError::UnexpectedEof {
                 expected: usize::MAX,
-                available: file_data.len(),
+                available: len_usize(file),
             })?;
-        ensure_len(file_data, pos, length)?;
+        Ok(read_exact_at(file, pos as u64, length)?.into_owned())
-        Ok(file_data[pos..pos + length].to_vec())
     }
 
     /// Read an object by traversing an indirect block to find the right direct block.
     #[allow(clippy::too_many_arguments)]
-    fn read_from_indirect_block(
+    fn read_from_indirect_block<S: Storage + ?Sized>(
         &self,
-        file_data: &[u8],
+        file: &S,
         iblock_addr: usize,
         nrows: u16,
         iblock_heap_offset: u64,
@@ -621,29 +667,169 @@ impl FractalHeapHeader {
                 "fractal heap: maximum recursion depth exceeded".into(),
             ));
         }
-        // Parse indirect block header
-        ensure_len(file_data, iblock_addr, 4)?;
-        if &file_data[iblock_addr..iblock_addr + 4] != b"FHIB" {
-            return Err(FormatError::InvalidFractalHeapSignature);
-        }
-
         let block_offset_bytes = (self.max_heap_size as usize).div_ceil(8);
         let iblock_header = 5 + offset_size as usize + block_offset_bytes;
-        let mut pos = iblock_addr + iblock_header;
-
-        let tw = self.table_width as u64;
         let nrows_usize = nrows as usize;
-        let mut current_heap_offset = iblock_heap_offset;
 
         // Rows below max_direct_rows hold direct blocks; rows at/above hold
         // child indirect blocks. (NOT the FRHP "starting rows" field.)
         let start_indirect = self.max_direct_rows();
         let max_direct_rows = nrows_usize.min(start_indirect);
 
+        // The block up to its last child entry. The walk below reads
+        // entries in order and stops at the one covering the target, which
+        // the geometry alone locates, so the first window ends there: a
+        // header claiming a huge table costs a read of the entries in front
+        // of the target, not of the rest of the file. Only when that entry
+        // is unallocated (or none covers the target) does the walk go on,
+        // over the whole block. Either window holds what it was asked for or
+        // ends at the end of the file, so its bounds checks are the
+        // whole-file ones.
+        let direct_entry = usize::from(offset_size)
+            + if self.filter_pipeline.is_some() {
+                usize::from(self.length_size) + 4
+            } else {
+                0
+            };
+        let direct_entries = max_direct_rows.saturating_mul(usize::from(self.table_width));
+        let entries_len = |n: usize| {
+            n.min(direct_entries)
+                .saturating_mul(direct_entry)
+                .saturating_add(
+                    n.saturating_sub(direct_entries)
+                        .saturating_mul(usize::from(offset_size)),
+                )
+        };
+        let all_entries = direct_entries.saturating_add(
+            nrows_usize
+                .saturating_sub(start_indirect)
+                .saturating_mul(usize::from(self.table_width)),
+        );
+        let block_len = iblock_header.saturating_add(entries_len(all_entries));
+        let target_entry = self.indirect_entry_for(nrows_usize, iblock_heap_offset, target_offset);
+        let first_len = target_entry.map_or(block_len, |i| {
+            iblock_header
+                .saturating_add(entries_len(i.saturating_add(1)))
+                .min(block_len)
+        });
+        let mut next = self.walk_indirect_block(
+            &Window::read(file, iblock_addr as u64, first_len)?,
+            nrows_usize,
+            iblock_heap_offset,
+            target_offset,
+            offset_size,
+            target_entry.map_or(usize::MAX, |i| i.saturating_add(1)),
+        )?;
+        if next.is_none() && first_len < block_len {
+            next = self.walk_indirect_block(
+                &Window::read(file, iblock_addr as u64, block_len)?,
+                nrows_usize,
+                iblock_heap_offset,
+                target_offset,
+                offset_size,
+                usize::MAX,
+            )?;
+        }
+        match next {
+            Some(IndirectChild::Direct(block)) => {
+                self.read_from_direct_block(file, block, target_offset, length)
+            }
+            Some(IndirectChild::Indirect {
+                addr,
+                nrows,
+                heap_offset,
+            }) => self.read_from_indirect_block(
+                file,
+                addr,
+                nrows,
+                heap_offset,
+                target_offset,
+                length,
+                offset_size,
+                depth_remaining - 1,
+            ),
+            None => Err(FormatError::UnexpectedEof {
+                expected: to_usize(target_offset)?.saturating_add(length),
+                available: len_usize(file),
+            }),
+        }
+    }
+
+    /// Which child entry of an indirect block (numbered in walk order:
+    /// direct rows, then indirect rows) covers `target_offset`, from the
+    /// doubling-table geometry alone — the entry
+    /// [`Self::walk_indirect_block`] stops at if it is allocated. `None`
+    /// when no entry does.
+    fn indirect_entry_for(&self, nrows: usize, heap_offset: u64, target: u64) -> Option<usize> {
+        // The walk adds block sizes with saturation; in u128 the same test
+        // is `cur <= target < cur + size` without it (a target of u64::MAX
+        // is never inside a saturated range).
+        if target == u64::MAX {
+            return None;
+        }
+        let (tw, target) = (u128::from(self.table_width), u128::from(target));
+        let mut cur = u128::from(heap_offset);
+        let mut before = 0usize;
+        for row in 0..nrows {
+            if target < cur {
+                return None;
+            }
+            // Direct and indirect rows alike span this row's block size per
+            // entry.
+            let size = u128::from(self.block_size_for_row(row));
+            let span = size * tw;
+            if size > 0 && target < cur + span {
+                let col = usize::try_from((target - cur) / size).ok()?;
+                return before.checked_add(col);
+            }
+            cur += span;
+            before = before.saturating_add(self.table_width as usize);
+        }
+        None
+    }
+
+    /// Walk an indirect block's child entries in order, in the window `w`
+    /// (the block from its signature on), and return the allocated child
+    /// covering `target_offset`, or `None` when no entry among the first
+    /// `limit` does.
+    fn walk_indirect_block(
+        &self,
+        w: &Window<'_>,
+        nrows: usize,
+        iblock_heap_offset: u64,
+        target_offset: u64,
+        offset_size: u8,
+        limit: usize,
+    ) -> Result<Option<IndirectChild>, FormatError> {
+        let ensure_len = |_: &[u8], pos: usize, needed: usize| w.ensure(pos, needed);
+        let read_offset = |_: &[u8], pos: usize, size: u8| {
+            w.ensure(pos, usize::from(size))?;
+            read_offset(&w.bytes, pos, size)
+        };
+        let file_data: &[u8] = &w.bytes;
+        let block_offset_bytes = (self.max_heap_size as usize).div_ceil(8);
+        let iblock_header = 5 + offset_size as usize + block_offset_bytes;
+        let tw = self.table_width as u64;
+        let mut current_heap_offset = iblock_heap_offset;
+        let start_indirect = self.max_direct_rows();
+        let max_direct_rows = nrows.min(start_indirect);
+        let mut walked = 0usize;
+
+        // Parse indirect block header
+        ensure_len(file_data, 0, 4)?;
+        if &file_data[..4] != b"FHIB" {
+            return Err(FormatError::InvalidFractalHeapSignature);
+        }
+        let mut pos = iblock_header;
+
         for row in 0..max_direct_rows {
             let block_size = self.block_size_for_row(row);
 
             for _col in 0..tw {
+                if walked == limit {
+                    return Ok(None);
+                }
+                walked += 1;
                 let child_addr = read_offset(file_data, pos, offset_size)?;
                 pos += offset_size as usize;
 
@@ -670,18 +856,13 @@ impl FractalHeapHeader {
                     && target_offset >= current_heap_offset
                     && target_offset < block_end
                 {
-                    return self.read_from_direct_block(
+                    return Ok(Some(IndirectChild::Direct(DirectBlock {
-                        file_data,
+                        addr: to_usize(child_addr)?,
-                        DirectBlock {
-                            addr: child_addr as usize,
                         size: block_size,
                         heap_offset: current_heap_offset,
                         filtered_size,
                         filter_mask,
-                        },
+                    })));
-                        target_offset,
-                        length,
-                    );
                 }
                 current_heap_offset = block_end;
             }
@@ -690,11 +871,15 @@ impl FractalHeapHeader {
         // Rows at and above `start_indirect` hold child indirect blocks. A
         // child in row r spans exactly that row's block size of heap space,
         // so it has as many rows as a table of that total size needs.
-        for row in start_indirect..nrows_usize {
+        for row in start_indirect..nrows {
             let child_space = self.block_size_for_row(row);
             let child_nrows = self.rows_for_size(child_space);
 
             for _col in 0..tw {
+                if walked == limit {
+                    return Ok(None);
+                }
+                walked += 1;
                 let child_addr = read_offset(file_data, pos, offset_size)?;
                 pos += offset_size as usize;
 
@@ -703,25 +888,16 @@ impl FractalHeapHeader {
                     && target_offset >= current_heap_offset
                     && target_offset < block_end
                 {
-                    return self.read_from_indirect_block(
+                    return Ok(Some(IndirectChild::Indirect {
-                        file_data,
+                        addr: to_usize(child_addr)?,
-                        child_addr as usize,
+                        nrows: child_nrows,
-                        child_nrows,
+                        heap_offset: current_heap_offset,
-                        current_heap_offset,
+                    }));
-                        target_offset,
-                        length,
-                        offset_size,
-                        depth_remaining - 1,
-                    );
                 }
                 current_heap_offset = block_end;
             }
         }
-
+        Ok(None)
-        Err(FormatError::UnexpectedEof {
-            expected: target_offset as usize + length,
-            available: file_data.len(),
-        })
     }
 
     /// Number of rows in the doubling table whose block size is at most the
@@ -765,6 +941,16 @@ impl FractalHeapHeader {
 
 /// A managed direct block's location, extent and (for a filtered heap) its
 /// stored size and filter mask.
+/// The child of an indirect block that covers a heap offset.
+enum IndirectChild {
+    Direct(DirectBlock),
+    Indirect {
+        addr: usize,
+        nrows: u16,
+        heap_offset: u64,
+    },
+}
+
 struct DirectBlock {
     addr: usize,
     size: u64,
@@ -1024,4 +1210,152 @@ mod tests {
         let id = [0x40u8, 0, 0, 0, 0, 0, 0];
         assert!(hdr.read_managed_object(&file_data, &id, 8).is_err());
     }
+
+    /// Headers, and managed (in a direct root and through an indirect
+    /// root), huge and tiny objects read identically through a
+    /// `read_at`-only storage, for every truncation of the file.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::storage::CountingStorage;
+        let (mut file, header_end) = build_simple_heap(8, 8);
+        // An indirect root block at 600: row 0 holds the direct block at
+        // 256, then three undefined blocks.
+        file[600..604].copy_from_slice(b"FHIB");
+        let mut at = 600 + 5 + 8 + 2;
+        for addr in [256u64, u64::MAX, u64::MAX, u64::MAX] {
+            file[at..at + 8].copy_from_slice(&addr.to_le_bytes());
+            at += 8;
+        }
+        file[900..905].copy_from_slice(b"huge!");
+        let managed_id = |offset: u64, len: u64| {
+            let payload = offset | (len << 16);
+            let mut id = vec![0u8];
+            id.extend_from_slice(&payload.to_le_bytes()[..6]);
+            id
+        };
+        let mut huge = vec![0x10u8];
+        huge.extend_from_slice(&900u64.to_le_bytes());
+        huge.extend_from_slice(&5u64.to_le_bytes());
+        let ids = [
+            managed_id(15, 13),
+            managed_id(15, 200),
+            managed_id(130, 4),
+            huge,
+            vec![0x22, b'a', b'b', b'c', 0, 0, 0],
+        ];
+        let mut cuts: Vec<usize> = (0..=header_end + 1).collect();
+        cuts.extend([256, 260, 271, 280, 600, 610, 620, 640, 900, 903, file.len()]);
+        for cut in cuts {
+            let f = &file[..cut];
+            let storage = CountingStorage::new(f.to_vec());
+            let want = FractalHeapHeader::parse(f, 0, 8, 8);
+            let got = FractalHeapHeader::parse_in(&storage, 0, 8, 8);
+            assert_eq!(format!("{got:?}"), format!("{want:?}"), "cut {cut}");
+            let Ok(direct) = want else { continue };
+            let mut indirect = direct.clone();
+            indirect.root_block_address = 600;
+            indirect.current_rows_in_root_indirect_block = 1;
+            let mut huge_ids = direct.clone();
+            huge_ids.heap_id_length = 17;
+            for hdr in [&direct, &indirect, &huge_ids] {
+                for id in &ids {
+                    assert_eq!(
+                        hdr.read_managed_object_in(&storage, id, 8),
+                        hdr.read_managed_object(f, id, 8),
+                        "cut {cut}"
+                    );
+                }
+            }
+        }
+    }
+
+    /// A header claiming a huge doubling table (width 0xFFFF, 0xFFFF rows in
+    /// the root indirect block) in a 16 MiB file: reading an object from the
+    /// table's first block reads the entries up to it, not the rest of the
+    /// file, and gives what the slice read gives. When the covering entry is
+    /// unallocated the walk goes on over the whole block, still identically.
+    #[test]
+    fn huge_table_claims_read_only_what_the_walk_needs() {
+        use crate::storage::CountingStorage;
+        let (mut file, _) = build_simple_heap(8, 8);
+        file.resize(16 << 20, 0);
+        file[600..604].copy_from_slice(b"FHIB");
+        let first_entry = 600 + 5 + 8 + 2;
+        file[first_entry..first_entry + 8].copy_from_slice(&256u64.to_le_bytes());
+        let mut hdr = FractalHeapHeader::parse(&file, 0, 8, 8).unwrap();
+        hdr.table_width = 0xFFFF;
+        hdr.root_block_address = 600;
+        hdr.current_rows_in_root_indirect_block = 0xFFFF;
+        let managed_id = |offset: u64, len: u64| {
+            let payload = offset | (len << 16);
+            let mut id = vec![0u8];
+            id.extend_from_slice(&payload.to_le_bytes()[..6]);
+            id
+        };
+        let storage = CountingStorage::new(file.clone());
+        let id = managed_id(15, 13);
+        let want = hdr.read_managed_object(&file, &id, 8);
+        assert!(want.is_ok(), "{want:?}");
+        storage.reset();
+        assert_eq!(hdr.read_managed_object_in(&storage, &id, 8), want);
+        assert!(
+            storage.bytes_read() < 1024,
+            "{} bytes in {} reads",
+            storage.bytes_read(),
+            storage.reads()
+        );
+        // The second entry (heap offsets 128..256) is unallocated (zero is
+        // not the undefined address, so make it all ones).
+        file[first_entry + 8..first_entry + 16].fill(0xFF);
+        let storage = CountingStorage::new(file.clone());
+        let id = managed_id(130, 4);
+        let want = hdr.read_managed_object(&file, &id, 8);
+        assert_eq!(hdr.read_managed_object_in(&storage, &id, 8), want);
+    }
+
+    /// A huge object found through the huge-object B-tree reads the
+    /// B-tree through Storage: the same result (here an error, there is no
+    /// B-tree at that address) as from the slice.
+    #[test]
+    fn huge_object_btree_reads_through_storage() {
+        use crate::storage::CountingStorage;
+        let (file, _) = build_simple_heap(8, 8);
+        let mut hdr = FractalHeapHeader::parse(&file, 0, 8, 8).unwrap();
+        hdr.huge_btree_address = 700;
+        let id = [0x10, 1, 0, 0, 0, 0, 0];
+        let want = hdr.read_managed_object(&file, &id, 8);
+        assert!(want.is_err());
+        let storage = CountingStorage::new(file);
+        assert_eq!(hdr.read_managed_object_in(&storage, &id, 8), want);
+    }
+
+    /// A header with an I/O filter pipeline (read in a second, longer
+    /// window) parses identically through a `read_at`-only storage, for
+    /// every truncation.
+    #[test]
+    fn filtered_header_parses_identically_through_storage() {
+        use crate::storage::CountingStorage;
+        let (simple, header_end) = build_simple_heap(8, 8);
+        let pipeline = [2u8, 1, 1, 0, 0, 0, 1, 0, 6, 0, 0, 0]; // deflate, level 6
+        let mut header = simple[..header_end - 4].to_vec();
+        header[7..9].copy_from_slice(&(pipeline.len() as u16).to_le_bytes());
+        header.extend_from_slice(&100u64.to_le_bytes()); // root block's stored size
+        header.extend_from_slice(&0u32.to_le_bytes()); // its filter mask
+        header.extend_from_slice(&pipeline);
+        let sum = crate::checksum::jenkins_lookup3(&header);
+        header.extend_from_slice(&sum.to_le_bytes());
+        let mut file = header.clone();
+        file.resize(256, 0);
+        let hdr = FractalHeapHeader::parse(&file, 0, 8, 8).unwrap();
+        assert!(hdr.filter_pipeline.is_some());
+        for cut in 0..=file.len() {
+            let f = &file[..cut];
+            let storage = CountingStorage::new(f.to_vec());
+            assert_eq!(
+                format!("{:?}", FractalHeapHeader::parse_in(&storage, 0, 8, 8)),
+                format!("{:?}", FractalHeapHeader::parse(f, 0, 8, 8)),
+                "cut {cut}"
+            );
+        }
+    }
 }

crates/clawhdf5-format/src/gather.rs

@@ -13,6 +13,7 @@ use alloc::{vec, vec::Vec};
 use crate::data_read::NativeElement;
 use crate::error::FormatError;
 use crate::selection::Selection;
+use crate::storage::{ExtentBytes, ExtentReq, Storage, raw_batches};
 
 /// Row-major element strides of `dims` (the last dimension has stride 1).
 fn strides(dims: &[u64]) -> Vec<u64> {
@@ -261,6 +262,259 @@ pub(crate) fn gather<T: NativeElement>(
     Ok(out)
 }
 
+/// Largest gap between two of a selection's runs that [`gather_storage`]
+/// reads through rather than asking for the runs separately: skipping a
+/// few KiB costs a remote backend far less than another request (and a
+/// local one less than another call and allocation).
+pub(crate) const GATHER_GAP_BYTES: usize = 4 << 10;
+
+/// Largest single read [`gather_storage`] makes of a selection's runs: runs
+/// are merged into reads up to this size, and a longer run is split.
+pub(crate) const GATHER_SPAN_BYTES: usize = 8 << 20;
+
+/// Call `emit(first_element, element_count)` for each run of a validated
+/// hyperslab or point selection (in output order; see [`hyperslab_runs`]),
+/// or the error for a hyperslab of the wrong rank or a point outside `dims`
+/// (runs before that point have been emitted).
+fn selection_runs(
+    dims: &[u64],
+    selection: &Selection,
+    emit: &mut dyn FnMut(u64, u64),
+) -> Result<(), FormatError> {
+    match selection {
+        Selection::Hyperslab {
+            start,
+            stride,
+            count,
+            block,
+        } => {
+            let rank = dims.len();
+            if [start.len(), stride.len(), count.len(), block.len()] != [rank; 4] {
+                return Err(FormatError::SelectionOutOfBounds(
+                    "hyperslab rank does not match dataset rank".into(),
+                ));
+            }
+            hyperslab_runs(dims, start, stride, count, block, emit);
+        }
+        Selection::Points(points) => {
+            let strides = strides(dims);
+            let mut coalesce = Coalesce {
+                start: 0,
+                len: 0,
+                emit,
+            };
+            for p in points {
+                if p.len() != dims.len() || p.iter().zip(dims).any(|(c, n)| c >= n) {
+                    return Err(FormatError::SelectionOutOfBounds(
+                        "selection addresses elements outside the dataset".into(),
+                    ));
+                }
+                let at = p
+                    .iter()
+                    .zip(&strides)
+                    .fold(0u64, |acc, (c, s)| acc.wrapping_add(c.wrapping_mul(*s)));
+                coalesce.push(at, 1);
+            }
+            coalesce.flush();
+        }
+        Selection::None | Selection::All => {}
+    }
+    Ok(())
+}
+
+/// One read of [`gather_storage`]: bytes `[start, end)` of the dataset,
+/// which hold the output's bytes up to `out_end` (from where the previous
+/// span's end left off).
+#[derive(Clone, Copy)]
+struct Span {
+    start: usize,
+    end: usize,
+    out_end: usize,
+}
+
+/// [`gather`] of bytes (`T = u8`) from a dataset that is not in memory: the
+/// dataset's `src_len` bytes start at `base` in `file`, which must hold all
+/// of them (the caller checks). Same checks and errors as [`gather`].
+///
+/// The selection's runs are walked twice. The first walk checks them and
+/// plans the reads: runs in increasing order with at most
+/// [`GATHER_GAP_BYTES`] between them are read as one span (the gap is read
+/// and dropped), up to [`GATHER_SPAN_BYTES`] per span. So a strided
+/// selection is a few large reads, not one per element, and nothing is
+/// allocated per run. The spans are fetched batch by batch (one
+/// [`Storage::read_ranges`] call per [`crate::storage::RAW_BATCH_BYTES`])
+/// while the second walk copies each run out of its span.
+pub(crate) fn gather_storage<S: Storage + ?Sized>(
+    file: &S,
+    base: u64,
+    src_len: usize,
+    dims: &[u64],
+    elem_size: usize,
+    selection: &Selection,
+) -> Result<Vec<u8>, FormatError> {
+    if elem_size == 0 {
+        return Err(FormatError::DataSizeMismatch {
+            expected: 1,
+            actual: elem_size,
+        });
+    }
+    let n_elements = match selection {
+        Selection::None => 0,
+        Selection::Hyperslab { count, block, .. } => count
+            .iter()
+            .zip(block)
+            .try_fold(1u64, |acc, (&c, &b)| acc.checked_mul(c.checked_mul(b)?))
+            .ok_or_else(|| FormatError::Overflow("hyperslab count x block overflows".into()))?,
+        Selection::Points(points) => points.len() as u64,
+        Selection::All => {
+            return Err(FormatError::SelectionOutOfBounds(
+                "gather does not take Selection::All".into(),
+            ));
+        }
+    };
+    let out_bytes = crate::chunked_read::checked_byte_len(n_elements, elem_size)?;
+    let outside = || {
+        FormatError::SelectionOutOfBounds("selection addresses elements outside the dataset".into())
+    };
+
+    // First walk: check every run and plan the spans.
+    let mut spans: Vec<Span> = Vec::new();
+    let mut total = 0usize;
+    let mut failed = false;
+    selection_runs(dims, selection, &mut |first: u64, n: u64| {
+        if failed {
+            return;
+        }
+        let range = usize::try_from(first)
+            .ok()
+            .and_then(|f| f.checked_mul(elem_size))
+            .zip(
+                usize::try_from(n)
+                    .ok()
+                    .and_then(|n| n.checked_mul(elem_size)),
+            )
+            .and_then(|(at, len)| Some((at, len, at.checked_add(len)?)));
+        let Some((mut at, mut len)) = range
+            .filter(|&(_, len, end)| end <= src_len && len <= out_bytes - total)
+            .map(|(at, len, _)| (at, len))
+        else {
+            failed = true;
+            return;
+        };
+        while len > 0 {
+            let room = match spans.last_mut() {
+                Some(s)
+                    if at >= s.end
+                        && at - s.end <= GATHER_GAP_BYTES
+                        && at - s.start < GATHER_SPAN_BYTES =>
+                {
+                    let take = len.min(GATHER_SPAN_BYTES - (at - s.start));
+                    s.end = at + take;
+                    s.out_end += take;
+                    take
+                }
+                _ => {
+                    let take = len.min(GATHER_SPAN_BYTES);
+                    spans.push(Span {
+                        start: at,
+                        end: at + take,
+                        out_end: total + take,
+                    });
+                    take
+                }
+            };
+            total += room;
+            at += room;
+            len -= room;
+        }
+    })?;
+    if failed || total != out_bytes {
+        return Err(outside());
+    }
+
+    // The spans' reads, and the batches they are fetched in.
+    let reqs: Vec<ExtentReq> = spans
+        .iter()
+        .map(|s| ExtentReq {
+            addr: base + s.start as u64,
+            len: s.end - s.start,
+            fetch: Some(s.end - s.start),
+        })
+        .collect();
+    let batches = raw_batches(reqs.len(), false, |i| reqs[i].len);
+
+    // Second walk: copy each run out of its span, fetching each batch of
+    // spans when the walk reaches it (and dropping the previous one).
+    let mut out = crate::bulk_alloc::vec_for_bulk(out_bytes);
+    let mut span = 0usize;
+    let mut batch = 0usize;
+    let mut fetched: Option<ExtentBytes<'_>> = None;
+    let mut error: Option<FormatError> = None;
+    selection_runs(dims, selection, &mut |first: u64, n: u64| {
+        if error.is_some() {
+            return;
+        }
+        // Checked by the first walk (these cannot saturate or wrap).
+        let mut at = crate::addr::saturating_usize(first).wrapping_mul(elem_size);
+        let mut len = crate::addr::saturating_usize(n).wrapping_mul(elem_size);
+        while len > 0 {
+            while spans.get(span).is_some_and(|s| s.out_end <= out.len()) {
+                span += 1;
+            }
+            if fetched.is_none() || span >= batches[batch].end {
+                fetched = None;
+                while batches.get(batch).is_some_and(|b| span >= b.end) {
+                    batch += 1;
+                }
+                let (Some(b), Some(_)) = (batches.get(batch).cloned(), spans.get(span)) else {
+                    // The second walk emitted more than the first.
+                    error = Some(outside());
+                    return;
+                };
+                match ExtentBytes::fetch(file, &reqs[b.clone()], b.start) {
+                    Ok(f) => fetched = Some(f),
+                    Err(e) => {
+                        error = Some(e);
+                        return;
+                    }
+                }
+            }
+            let s = spans[span];
+            let take = len.min(s.out_end - out.len());
+            let bytes = match fetched
+                .as_ref()
+                .map(|f| f.get(span, &reqs[span]))
+                .unwrap_or_else(|| Err(outside()))
+            {
+                Ok(b) => b,
+                Err(e) => {
+                    error = Some(e);
+                    return;
+                }
+            };
+            match at
+                .checked_sub(s.start)
+                .and_then(|o| bytes.get(o..o.checked_add(take)?))
+            {
+                Some(b) => out.extend_from_slice(b),
+                None => {
+                    error = Some(outside());
+                    return;
+                }
+            }
+            at += take;
+            len -= take;
+        }
+    })?;
+    if let Some(e) = error {
+        return Err(e);
+    }
+    if out.len() != out_bytes {
+        return Err(outside());
+    }
+    Ok(out)
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;

crates/clawhdf5-format/src/global_heap.rs

@@ -1,9 +1,12 @@
 //! HDF5 Global Heap collection parsing.
 
 #[cfg(not(feature = "std"))]
-use alloc::{format, string::String, vec::Vec};
+use alloc::{borrow::Cow, format, string::String, vec::Vec};
+#[cfg(feature = "std")]
+use std::borrow::Cow;
 
 use crate::error::FormatError;
+use crate::storage::{Storage, len_usize, read_exact_at};
 
 /// Magic signature for global heap collections.
 const GCOL_SIGNATURE: [u8; 4] = *b"GCOL";
@@ -28,19 +31,20 @@ pub struct GlobalHeapObject {
     pub data: Vec<u8>,
 }
 
-fn ensure_len(data: &[u8], offset: usize, needed: usize) -> Result<(), FormatError> {
+/// Checks that `[offset, offset + needed)` ends by `data_len`.
+fn ensure_len(data_len: usize, offset: usize, needed: usize) -> Result<(), FormatError> {
     match offset.checked_add(needed) {
-        Some(end) if end <= data.len() => Ok(()),
+        Some(end) if end <= data_len => Ok(()),
         _ => Err(FormatError::UnexpectedEof {
             expected: offset.saturating_add(needed),
-            available: data.len(),
+            available: data_len,
         }),
     }
 }
 
 fn read_length(data: &[u8], offset: usize, length_size: u8) -> Result<u64, FormatError> {
     let s = length_size as usize;
-    ensure_len(data, offset, s)?;
+    ensure_len(data.len(), offset, s)?;
     let slice = &data[offset..offset + s];
     Ok(match length_size {
         2 => u16::from_le_bytes([slice[0], slice[1]]) as u64,
@@ -95,7 +99,17 @@ impl GlobalHeapCollection {
         offset: usize,
         length_size: u8,
     ) -> Result<GlobalHeapCollection, FormatError> {
-        let index = Self::parse_index(file_data, offset, length_size)?;
+        Self::parse_in(file_data, offset as u64, length_size)
+    }
+
+    /// [`Self::parse`] over any [`Storage`]: one read of the header, one of
+    /// the collection.
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        length_size: u8,
+    ) -> Result<GlobalHeapCollection, FormatError> {
+        let (bytes, base, index) = Self::read_collection(file, offset, length_size)?;
         Ok(GlobalHeapCollection {
             collection_size: index.collection_size,
             objects: index
@@ -104,7 +118,7 @@ impl GlobalHeapCollection {
                 .map(|o| GlobalHeapObject {
                     index: o.index,
                     reference_count: o.reference_count,
-                    data: file_data[o.offset..o.offset + o.size].to_vec(),
+                    data: bytes[o.offset - base..o.offset - base + o.size].to_vec(),
                 })
                 .collect(),
         })
@@ -122,43 +136,72 @@ impl GlobalHeapCollection {
         offset: usize,
         length_size: u8,
     ) -> Result<GlobalHeapIndex, FormatError> {
+        Self::parse_index_in(file_data, offset as u64, length_size)
+    }
+
+    /// [`Self::parse_index`] over any [`Storage`]: one read of the header,
+    /// one of the collection. The object offsets are file offsets.
+    pub fn parse_index_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        length_size: u8,
+    ) -> Result<GlobalHeapIndex, FormatError> {
+        Ok(Self::read_collection(file, offset, length_size)?.2)
+    }
+
+    /// Read the collection at `offset` and index its objects: the
+    /// collection's bytes, its offset as a `usize`, and the index (with
+    /// file offsets).
+    pub(crate) fn read_collection<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        length_size: u8,
+    ) -> Result<(Cow<'_, [u8]>, usize, GlobalHeapIndex), FormatError> {
+        let file_len = len_usize(file);
         // signature(4) + version(1) + reserved(3) + collection_size(length_size),
         // padded to a multiple of 8 as libhdf5 lays it out (`H5HG_SIZEOF_HDR`).
         // With 8-byte lengths the padding is 0; with 4-byte lengths it is 4,
         // and reading without it put every object 4 bytes early.
         let header_size = pad8(8 + length_size as usize);
-        ensure_len(file_data, offset, header_size)?;
+        let header = read_exact_at(file, offset, header_size)?;
+        let offset = usize::try_from(offset).map_err(|_| FormatError::UnexpectedEof {
+            expected: usize::MAX,
+            available: file_len,
+        })?;
 
-        if file_data[offset..offset + 4] != GCOL_SIGNATURE {
+        if header[..4] != GCOL_SIGNATURE {
             return Err(FormatError::InvalidGlobalHeapSignature);
         }
 
-        let version = file_data[offset + 4];
+        let version = header[4];
         if version != 1 {
             return Err(FormatError::InvalidGlobalHeapVersion(version));
         }
 
-        let collection_size = read_length(file_data, offset + 8, length_size)?;
+        let collection_size = read_length(&header, 8, length_size)?;
         let collection_end = usize::try_from(collection_size)
             .ok()
             .and_then(|size| offset.checked_add(size))
             .ok_or(FormatError::UnexpectedEof {
                 expected: usize::MAX,
-                available: file_data.len(),
+                available: file_len,
             })?;
-        if collection_end > file_data.len() {
+        if collection_end > file_len {
             return Err(FormatError::UnexpectedEof {
                 expected: collection_end,
-                available: file_data.len(),
+                available: file_len,
             });
         }
+        let collection = read_exact_at(file, offset as u64, collection_end - offset)?;
+        // Positions below are file offsets; `file_data(p)` is the byte at `p`.
+        let file_data = |p: usize| collection[p - offset];
 
         let mut pos = offset + header_size;
         let mut objects = Vec::new();
 
         // Parse objects until we hit index 0 (free space) or run out of space
         while pos + 2 <= collection_end {
-            let object_index = u16::from_le_bytes([file_data[pos], file_data[pos + 1]]);
+            let object_index = u16::from_le_bytes([file_data(pos), file_data(pos + 1)]);
 
             if object_index == 0 {
                 // Free space marker — done
@@ -168,10 +211,11 @@ impl GlobalHeapCollection {
             // object_index(2) + reference_count(2) + reserved(4) +
             // object_size(length_size), padded to 8 (`H5HG_SIZEOF_OBJHDR`).
             let obj_header_size = pad8(8 + length_size as usize);
-            ensure_len(&file_data[..collection_end], pos, obj_header_size)?;
+            ensure_len(collection_end, pos, obj_header_size)?;
 
-            let reference_count = u16::from_le_bytes([file_data[pos + 2], file_data[pos + 3]]);
+            let reference_count = u16::from_le_bytes([file_data(pos + 2), file_data(pos + 3)]);
-            let object_size = usize::try_from(read_length(file_data, pos + 8, length_size)?)
+            let object_size =
+                usize::try_from(read_length(&collection[pos - offset..], 8, length_size)?)
                     .map_err(|_| FormatError::Overflow("global heap object size".into()))?;
 
             pos += obj_header_size;
@@ -197,10 +241,11 @@ impl GlobalHeapCollection {
             pos = pos.saturating_add(pad8(object_size));
         }
 
-        Ok(GlobalHeapIndex {
+        let index = GlobalHeapIndex {
             collection_size,
             objects,
-        })
+        };
+        Ok((collection, offset, index))
     }
 
     /// Get an object by its index.
@@ -226,7 +271,7 @@ mod tests {
         let mut obj_size_total = 0usize;
         for (_, _, data) in objects {
             let obj_header = pad8(8 + ls);
-            obj_size_total += obj_header + pad8(data.len());
+            obj_size_total += obj_header + pad8(<[u8]>::len(data));
         }
         // Free space marker (2 bytes for index 0)
         obj_size_total += 2;
@@ -251,15 +296,17 @@ mod tests {
             buf.extend_from_slice(&ref_count.to_le_bytes());
             buf.extend_from_slice(&[0u8; 4]); // reserved
             match length_size {
-                4 => buf.extend_from_slice(&(data.len() as u32).to_le_bytes()),
+                // `<[u8]>::len`: with `Storage` in scope `data.len()` on a
-                8 => buf.extend_from_slice(&(data.len() as u64).to_le_bytes()),
+                // `&&[u8]` resolves to `Storage::len` (a `u64`).
+                4 => buf.extend_from_slice(&(<[u8]>::len(data) as u32).to_le_bytes()),
+                8 => buf.extend_from_slice(&(<[u8]>::len(data) as u64).to_le_bytes()),
                 _ => panic!("unsupported"),
             }
             buf.resize(buf.len() + (pad8(8 + ls) - (8 + ls)), 0);
             buf.extend_from_slice(data);
             // Pad to 8 bytes
-            let padded = pad8(data.len());
+            let padded = pad8(<[u8]>::len(data));
-            buf.resize(buf.len() + (padded - data.len()), 0);
+            buf.resize(buf.len() + (padded - <[u8]>::len(data)), 0);
         }
 
         // Free space marker
@@ -327,4 +374,44 @@ mod tests {
         assert_eq!(coll.objects.len(), 1);
         assert_eq!(coll.objects[0].data, b"test");
     }
+
+    /// Collections, and every truncation of them, index and parse
+    /// identically through a `read_at`-only storage: two reads each.
+    #[test]
+    fn storage_parse_matches_slice_parse() {
+        use crate::storage::CountingStorage;
+        let objs: &[(u16, u16, &[u8])] = &[(1, 1, b"hello"), (2, 3, b"a longer object")];
+        for ls in [4u8, 8] {
+            let coll = build_collection(objs, ls);
+            let mut corrupt = coll.clone();
+            corrupt[8] = 200; // collection size past the end of the file
+            let mut overrun = coll.clone();
+            let size_at = pad8(8 + ls as usize) + 8;
+            overrun[size_at] = 250; // first object runs past the collection
+            for full in [coll, corrupt, overrun] {
+                for at in [0usize, 5] {
+                    for cut in 0..=full.len() {
+                        let mut f = vec![0u8; at];
+                        f.extend_from_slice(&full[..cut]);
+                        let storage = CountingStorage::new(f.clone());
+                        let want = GlobalHeapCollection::parse(&f, at, ls);
+                        let got = GlobalHeapCollection::parse_in(&storage, at as u64, ls);
+                        assert_eq!(format!("{got:?}"), format!("{want:?}"));
+                        let want = GlobalHeapCollection::parse_index(&f, at, ls);
+                        let got = GlobalHeapCollection::parse_index_in(&storage, at as u64, ls);
+                        assert_eq!(format!("{got:?}"), format!("{want:?}"));
+                    }
+                }
+            }
+        }
+        let storage = CountingStorage::new(build_collection(objs, 8));
+        assert_eq!(
+            GlobalHeapCollection::parse_in(&storage, 0, 8)
+                .unwrap()
+                .objects
+                .len(),
+            2
+        );
+        assert_eq!(storage.reads(), 2);
+    }
 }

crates/clawhdf5-format/src/group_v1.rs

@@ -3,11 +3,13 @@
 #[cfg(not(feature = "std"))]
 use alloc::{string::String, vec::Vec};
 
-use crate::btree_v1::collect_symbol_table_nodes;
+use crate::addr::checked_addr;
+use crate::btree_v1::collect_symbol_table_nodes_in;
 use crate::error::FormatError;
 use crate::local_heap::LocalHeap;
 use crate::message_type::MessageType;
 use crate::object_header::ObjectHeader;
+use crate::storage::Storage;
 use crate::symbol_table::{SymbolTableMessage, SymbolTableNode};
 
 /// A resolved group entry (child name + object header address).
@@ -21,23 +23,56 @@ pub struct GroupEntry {
     pub cache_type: u32,
 }
 
-/// Given a SymbolTableMessage, resolve all group children.
+/// Given a SymbolTableMessage, resolve all group children: the group's
+/// listing.
+///
+/// An entry with an empty name fails the listing with
+/// [`FormatError::InvalidLinkName`], as it fails libhdf5's link iteration
+/// (`H5G__ent_to_link`: "invalid link name"). Looking a name up
+/// ([`resolve_path`], and the path resolution in
+/// [`crate::group_v2::resolve_path_any`]) still works in such a group, as it
+/// does in libhdf5.
 pub fn resolve_v1_group_entries(
     file_data: &[u8],
     sym_table_msg: &SymbolTableMessage,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<GroupEntry>, FormatError> {
+    resolve_v1_group_entries_in(file_data, sym_table_msg, offset_size, length_size)
+}
+
+/// [`resolve_v1_group_entries`] over any [`Storage`].
+pub fn resolve_v1_group_entries_in<S: Storage + ?Sized>(
+    file_data: &S,
+    sym_table_msg: &SymbolTableMessage,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<GroupEntry>, FormatError> {
+    let entries = v1_group_entries(file_data, sym_table_msg, offset_size, length_size)?;
+    if entries.iter().any(|e| e.name.is_empty()) {
+        return Err(FormatError::InvalidLinkName);
+    }
+    Ok(entries)
+}
+
+/// Every entry of a v1 group, empty names included — for looking a name up,
+/// which never matches an empty name.
+pub(crate) fn v1_group_entries<S: Storage + ?Sized>(
+    file_data: &S,
+    sym_table_msg: &SymbolTableMessage,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<GroupEntry>, FormatError> {
     // Parse local heap
-    let heap = LocalHeap::parse(
+    let heap = LocalHeap::parse_in(
         file_data,
-        sym_table_msg.local_heap_address as usize,
+        checked_addr(sym_table_msg.local_heap_address)?,
         offset_size,
         length_size,
     )?;
 
     // Collect all SNOD addresses from B-tree
-    let snod_addrs = collect_symbol_table_nodes(
+    let snod_addrs = collect_symbol_table_nodes_in(
         file_data,
         sym_table_msg.btree_address,
         offset_size,
@@ -47,15 +82,15 @@ pub fn resolve_v1_group_entries(
     let mut entries = Vec::new();
     let mut heap_checked = false;
     for snod_addr in snod_addrs {
-        let snod = SymbolTableNode::parse(file_data, snod_addr as usize, offset_size)?;
+        let snod = SymbolTableNode::parse_in(file_data, checked_addr(snod_addr)?, offset_size)?;
         for entry in &snod.entries {
             // Like libhdf5, look at the heap's free list only once a name is
             // needed: an empty group with a damaged heap still lists.
             if !heap_checked {
-                heap.validate_free_list(file_data, length_size)?;
+                heap.validate_free_list_in(file_data, length_size)?;
                 heap_checked = true;
             }
-            let name = heap.read_string(file_data, entry.link_name_offset)?;
+            let name = heap.read_string_in(file_data, entry.link_name_offset)?;
             entries.push(GroupEntry {
                 name,
                 object_header_address: entry.object_header_address,
@@ -79,6 +114,17 @@ pub fn find_v1_soft_link(
     name: &str,
     offset_size: u8,
     length_size: u8,
+) -> Result<Option<String>, FormatError> {
+    find_v1_soft_link_in(file_data, sym_table_msg, name, offset_size, length_size)
+}
+
+/// [`find_v1_soft_link`] over any [`Storage`].
+pub fn find_v1_soft_link_in<S: Storage + ?Sized>(
+    file_data: &S,
+    sym_table_msg: &SymbolTableMessage,
+    name: &str,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Option<String>, FormatError> {
     let mut found = None;
     for_each_v1_soft_link(
@@ -101,6 +147,16 @@ pub fn v1_soft_links(
     sym_table_msg: &SymbolTableMessage,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<(String, String)>, FormatError> {
+    v1_soft_links_in(file_data, sym_table_msg, offset_size, length_size)
+}
+
+/// [`v1_soft_links`] over any [`Storage`].
+pub fn v1_soft_links_in<S: Storage + ?Sized>(
+    file_data: &S,
+    sym_table_msg: &SymbolTableMessage,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<(String, String)>, FormatError> {
     let mut links = Vec::new();
     for_each_v1_soft_link(
@@ -119,21 +175,21 @@ pub fn v1_soft_links(
 
 /// Visit the soft links of a v1 group whose name passes `wanted`, with their
 /// target paths, until `visit` returns false.
-fn for_each_v1_soft_link(
+fn for_each_v1_soft_link<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     sym_table_msg: &SymbolTableMessage,
     offset_size: u8,
     length_size: u8,
     wanted: impl Fn(&str) -> bool,
     mut visit: impl FnMut(&str, String) -> bool,
 ) -> Result<(), FormatError> {
-    let heap = LocalHeap::parse(
+    let heap = LocalHeap::parse_in(
         file_data,
-        sym_table_msg.local_heap_address as usize,
+        checked_addr(sym_table_msg.local_heap_address)?,
         offset_size,
         length_size,
     )?;
-    let snod_addrs = collect_symbol_table_nodes(
+    let snod_addrs = collect_symbol_table_nodes_in(
         file_data,
         sym_table_msg.btree_address,
         offset_size,
@@ -141,16 +197,16 @@ fn for_each_v1_soft_link(
     )?;
     let mut heap_checked = false;
     for snod_addr in snod_addrs {
-        let snod = SymbolTableNode::parse(file_data, snod_addr as usize, offset_size)?;
+        let snod = SymbolTableNode::parse_in(file_data, checked_addr(snod_addr)?, offset_size)?;
         for entry in &snod.entries {
             if entry.cache_type != CACHE_TYPE_SOFT_LINK {
                 continue;
             }
             if !heap_checked {
-                heap.validate_free_list(file_data, length_size)?;
+                heap.validate_free_list_in(file_data, length_size)?;
                 heap_checked = true;
             }
-            let name = heap.read_string(file_data, entry.link_name_offset)?;
+            let name = heap.read_string_in(file_data, entry.link_name_offset)?;
             if !wanted(&name) {
                 continue;
             }
@@ -160,7 +216,7 @@ fn for_each_v1_soft_link(
                 entry.scratch_pad[2],
                 entry.scratch_pad[3],
             ]);
-            let target = heap.read_string(file_data, u64::from(value_offset))?;
+            let target = heap.read_string_in(file_data, u64::from(value_offset))?;
             if !visit(&name, target) {
                 return Ok(());
             }
@@ -198,6 +254,17 @@ pub fn resolve_path(
     path: &str,
     offset_size: u8,
     length_size: u8,
+) -> Result<u64, FormatError> {
+    resolve_path_in(file_data, root_sym_table, path, offset_size, length_size)
+}
+
+/// [`resolve_path`] over any [`Storage`].
+pub fn resolve_path_in<S: Storage + ?Sized>(
+    file_data: &S,
+    root_sym_table: &SymbolTableMessage,
+    path: &str,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<u64, FormatError> {
     let components: Vec<&str> = path.split('/').filter(|s| !s.is_empty()).collect();
     if components.is_empty() {
@@ -207,8 +274,7 @@ pub fn resolve_path(
     let mut current_sym_table = root_sym_table.clone();
 
     for (i, component) in components.iter().enumerate() {
-        let entries =
+        let entries = v1_group_entries(file_data, &current_sym_table, offset_size, length_size)?;
-            resolve_v1_group_entries(file_data, &current_sym_table, offset_size, length_size)?;
 
         let found = entries.iter().find(|e| e.name == *component);
         match found {
@@ -218,9 +284,9 @@ pub fn resolve_path(
                     return Ok(entry.object_header_address);
                 }
                 // Not last — must be a group, parse its object header to get symbol table
-                let obj_header = ObjectHeader::parse(
+                let obj_header = ObjectHeader::parse_in(
                     file_data,
-                    entry.object_header_address as usize,
+                    checked_addr(entry.object_header_address)?,
                     offset_size,
                     length_size,
                 )?;
@@ -425,6 +491,22 @@ mod tests {
         assert_eq!(entries[1].object_header_address, 0x2000);
     }
 
+    /// cve-2021-46244 `/BAG_root`: a symbol-table entry with an empty name.
+    /// libhdf5 fails the group's listing ("invalid link name"); a lookup of
+    /// the other names still works.
+    #[test]
+    fn empty_entry_name_fails_the_listing_not_a_lookup() {
+        let (file, msg) = build_synthetic_group(&[("", 0x1000, 0), ("elevation", 0x2000, 0)], 8, 8);
+        assert_eq!(
+            resolve_v1_group_entries(&file, &msg, 8, 8).unwrap_err(),
+            FormatError::InvalidLinkName
+        );
+        assert_eq!(
+            resolve_path(&file, &msg, "elevation", 8, 8).unwrap(),
+            0x2000
+        );
+    }
+
     #[test]
     fn resolve_path_single_level() {
         let (file, msg) =

crates/clawhdf5-format/src/group_v2.rs

@@ -6,7 +6,14 @@
 #[cfg(not(feature = "std"))]
 use alloc::{string::String, vec::Vec};
 
-use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records};
+#[cfg(not(feature = "std"))]
+use alloc::collections::BTreeSet;
+#[cfg(feature = "std")]
+use std::collections::BTreeSet;
+
+use crate::addr::checked_addr;
+use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records_in, find_btree_v2_records_in};
+use crate::checksum::jenkins_lookup3;
 use crate::error::FormatError;
 use crate::fractal_heap::FractalHeapHeader;
 use crate::group_v1::{self, GroupEntry};
@@ -14,6 +21,7 @@ use crate::link_info::LinkInfoMessage;
 use crate::link_message::{LinkMessage, LinkTarget};
 use crate::message_type::MessageType;
 use crate::object_header::ObjectHeader;
+use crate::storage::Storage;
 use crate::superblock::Superblock;
 use crate::symbol_table::SymbolTableMessage;
 
@@ -25,6 +33,16 @@ pub fn resolve_v2_group_entries(
     object_header: &ObjectHeader,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<GroupEntry>, FormatError> {
+    resolve_v2_group_entries_in(file_data, object_header, offset_size, length_size)
+}
+
+/// [`resolve_v2_group_entries`] over any [`Storage`].
+pub fn resolve_v2_group_entries_in<S: Storage + ?Sized>(
+    file_data: &S,
+    object_header: &ObjectHeader,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<GroupEntry>, FormatError> {
     // Look for Link Info message to determine storage type
     let link_info = find_link_info(object_header, offset_size)?;
@@ -84,8 +102,8 @@ fn resolve_compact_entries(
 }
 
 /// Visit every link in dense storage (fractal heap + B-tree v2 name index).
-fn for_each_dense_link(
+fn for_each_dense_link<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     link_info: &LinkInfoMessage,
     fh_addr: u64,
     offset_size: u8,
@@ -93,14 +111,20 @@ fn for_each_dense_link(
     mut visit: impl FnMut(LinkMessage),
 ) -> Result<(), FormatError> {
     // Parse fractal heap
-    let fh = FractalHeapHeader::parse(file_data, fh_addr as usize, offset_size, length_size)?;
+    let fh =
+        FractalHeapHeader::parse_in(file_data, checked_addr(fh_addr)?, offset_size, length_size)?;
 
     // Parse B-tree v2 for name index
     let btree_addr = link_info
         .btree_name_index_address
         .ok_or_else(|| FormatError::PathNotFound(String::from("no B-tree v2 name index")))?;
-    let btree_hdr = BTreeV2Header::parse(file_data, btree_addr as usize, offset_size, length_size)?;
+    let btree_hdr = BTreeV2Header::parse_in(
-    let records = collect_btree_v2_records(file_data, &btree_hdr, offset_size, length_size)?;
+        file_data,
+        checked_addr(btree_addr)?,
+        offset_size,
+        length_size,
+    )?;
+    let records = collect_btree_v2_records_in(file_data, &btree_hdr, offset_size, length_size)?;
 
     for record in &records {
         // For type 5 (name index): hash(4) + heap_id(heap_id_length)
@@ -117,7 +141,7 @@ fn for_each_dense_link(
         let id_bytes = &record.data[id_offset..id_offset + fh.heap_id_length as usize];
 
         // Read managed object from fractal heap
-        let link_data = fh.read_managed_object(file_data, id_bytes, offset_size)?;
+        let link_data = fh.read_managed_object_in(file_data, id_bytes, offset_size)?;
         if let Some(link) = parse_link(&link_data, offset_size)? {
             visit(link);
         }
@@ -126,8 +150,8 @@ fn for_each_dense_link(
 }
 
 /// Resolve entries from dense storage (fractal heap + B-tree v2).
-fn resolve_dense_entries(
+fn resolve_dense_entries<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     link_info: &LinkInfoMessage,
     fh_addr: u64,
     offset_size: u8,
@@ -156,17 +180,16 @@ fn resolve_dense_entries(
     Ok(entries)
 }
 
-/// The soft or external link called `name` in this group, if there is one.
+/// The soft link called `name` in a v1 (symbol table) group, if there is
-/// Hard links are what `resolve_group_entries` returns; this is consulted only
+/// one. Hard links are what `resolve_group_entries` returns; this is
-/// when a path component isn't among them.
+/// consulted only when a path component isn't among them.
-fn find_symbolic_link(
+fn find_v1_symbolic_link<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     object_header: &ObjectHeader,
     name: &str,
     offset_size: u8,
     length_size: u8,
 ) -> Result<Option<LinkTarget>, FormatError> {
-    if is_v1_group(object_header) {
     let Some(sym_msg) = object_header
         .messages
         .iter()
@@ -175,16 +198,56 @@ fn find_symbolic_link(
         return Ok(None);
     };
     let stm = SymbolTableMessage::parse(&sym_msg.data, offset_size)?;
-        return group_v1::find_v1_soft_link(file_data, &stm, name, offset_size, length_size)
+    group_v1::find_v1_soft_link_in(file_data, &stm, name, offset_size, length_size)
-            .map(|target| target.map(|target_path| LinkTarget::Soft { target_path }));
+        .map(|target| target.map(|target_path| LinkTarget::Soft { target_path }))
-    }
+}
-    if !is_v2_group(object_header) {
+
-        return Ok(None);
+/// B-tree v2 record type of a dense group's link name index.
-    }
+const LINK_NAME_INDEX: u8 = 5;
-    let is_symbolic = |t: &LinkTarget| !matches!(t, LinkTarget::Hard { .. });
+
+/// The links called `name` in a v2 group (a valid group has at most one),
+/// in storage order: header message order for a compact group, name index
+/// order for a dense one.
+///
+/// In dense storage the link name index (a v2 B-tree of lookup3 name
+/// hashes, record type 5) is descended to the records with the name's hash,
+/// and only their links are read from the heap — O(log n) instead of every
+/// link. libhdf5 orders records with equal hashes by name; all of them are
+/// read and compared here, so that order does not matter. An index of
+/// another type is scanned in full.
+fn links_named<S: Storage + ?Sized>(
+    file_data: &S,
+    object_header: &ObjectHeader,
+    name: &str,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<LinkMessage>, FormatError> {
+    let mut found = Vec::new();
     let link_info = find_link_info(object_header, offset_size)?;
-    let mut found = None;
+    let Some(fh_addr) = link_info.fractal_heap_address else {
-    if let Some(fh_addr) = link_info.fractal_heap_address {
+        for msg in &object_header.messages {
+            if msg.msg_type == MessageType::Link
+                && let Some(link) = parse_link(&msg.data, offset_size)?
+                && link.name == name
+            {
+                found.push(link);
+            }
+        }
+        return Ok(found);
+    };
+
+    let fh =
+        FractalHeapHeader::parse_in(file_data, checked_addr(fh_addr)?, offset_size, length_size)?;
+    let btree_addr = link_info
+        .btree_name_index_address
+        .ok_or_else(|| FormatError::PathNotFound(String::from("no B-tree v2 name index")))?;
+    let btree_hdr = BTreeV2Header::parse_in(
+        file_data,
+        checked_addr(btree_addr)?,
+        offset_size,
+        length_size,
+    )?;
+    if btree_hdr.tree_type != LINK_NAME_INDEX {
         for_each_dense_link(
             file_data,
             &link_info,
@@ -192,26 +255,176 @@ fn find_symbolic_link(
             offset_size,
             length_size,
             |link| {
-                if link.name == name && is_symbolic(&link.link_target) {
+                if link.name == name {
-                    found = Some(link.link_target);
+                    found.push(link);
                 }
             },
         )?;
-    } else {
+        return Ok(found);
-        for msg in &object_header.messages {
+    }
-            if msg.msg_type == MessageType::Link {
+
-                let Some(link) = parse_link(&msg.data, offset_size)? else {
+    // Record: hash(4) + heap ID.
+    let hash = jenkins_lookup3(name.as_bytes());
+    let records = find_btree_v2_records_in(file_data, &btree_hdr, offset_size, &mut |r| {
+        match r.get(..4) {
+            Some(h) => u32::from_le_bytes([h[0], h[1], h[2], h[3]]).cmp(&hash),
+            // Too short to hold a hash (a corrupt record size): never a match.
+            None => core::cmp::Ordering::Less,
+        }
+    })?;
+    let id_len = usize::from(fh.heap_id_length);
+    for record in &records {
+        let Some(id_bytes) = record.data.get(4..4 + id_len) else {
             continue;
         };
-                if link.name == name && is_symbolic(&link.link_target) {
+        let link_data = fh.read_managed_object_in(file_data, id_bytes, offset_size)?;
-                    found = Some(link.link_target);
+        if let Some(link) = parse_link(&link_data, offset_size)?
-                }
+            && link.name == name
-            }
+        {
+            found.push(link);
         }
     }
     Ok(found)
 }
 
+/// The link called `name` in a v2 group, if any.
+///
+/// A valid group has at most one; libhdf5 cannot create two. If a damaged
+/// or hand-made group has several, the first wins and the rest are
+/// ignored, whatever their kind and even if the first cannot be followed.
+/// That is libhdf5's rule for a compact group (`H5G__compact_lookup` stops
+/// at the first Link message of that name; h5py then fails to open a
+/// dangling first link although a later one resolves). For a dense group
+/// "first" is first in name index order; libhdf5 binary-searches the index
+/// and may land on another of several exact duplicates. The listing
+/// ([`resolve_group_children`]), [`resolve_child`] and path resolution all
+/// apply this rule, so they agree.
+fn first_link_named<S: Storage + ?Sized>(
+    file_data: &S,
+    object_header: &ObjectHeader,
+    name: &str,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Option<LinkMessage>, FormatError> {
+    Ok(
+        links_named(file_data, object_header, name, offset_size, length_size)?
+            .into_iter()
+            .next(),
+    )
+}
+
+/// The link [`resolve_path_any`] follows for one path component `name` of
+/// the group with header `object_header`: a hard link (as `Hard`), else a
+/// soft or external link of that name, else `None`. Fails with
+/// `PathNotFound` if the object is not a group.
+fn lookup_link<S: Storage + ?Sized>(
+    file_data: &S,
+    object_header: &ObjectHeader,
+    name: &str,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Option<LinkTarget>, FormatError> {
+    if is_v1_group(object_header) {
+        let entries = resolve_group_entries(file_data, object_header, offset_size, length_size)?;
+        if let Some(e) = entries
+            .iter()
+            .find(|e| e.name == name && e.object_header_address != u64::MAX)
+        {
+            return Ok(Some(LinkTarget::Hard {
+                object_header_address: e.object_header_address,
+            }));
+        }
+        return find_v1_symbolic_link(file_data, object_header, name, offset_size, length_size);
+    }
+    if !is_v2_group(object_header) {
+        return Err(FormatError::PathNotFound(String::from(
+            "object header is not a group",
+        )));
+    }
+    Ok(
+        first_link_named(file_data, object_header, name, offset_size, length_size)?
+            .map(|link| link.link_target)
+            .filter(|t| {
+                !matches!(
+                    t,
+                    LinkTarget::Hard {
+                        object_header_address: u64::MAX
+                    }
+                )
+            }),
+    )
+}
+
+/// The object header address of the child called `name` of the group at
+/// `group_address`: the address [`resolve_group_children`] lists under that
+/// name, or `PathNotFound` if it lists none.
+///
+/// A dense group's child is found through its link name index (see
+/// [`links_named`]) and only the named link is read and, if it is a soft
+/// link, followed — not every link in the group. A v1 group is listed.
+pub fn resolve_child(
+    file_data: &[u8],
+    superblock: &Superblock,
+    group_address: u64,
+    name: &str,
+) -> Result<u64, FormatError> {
+    resolve_child_core(file_data, superblock, group_address, name)
+}
+
+/// [`resolve_child`] over any [`Storage`]. One with the whole file in memory
+/// is read as the slice, by code compiled in this crate (see
+/// [`crate::storage`], "Slice entry points").
+#[inline]
+pub fn resolve_child_in<S: Storage + ?Sized>(
+    file_data: &S,
+    superblock: &Superblock,
+    group_address: u64,
+    name: &str,
+) -> Result<u64, FormatError> {
+    match file_data.as_contiguous() {
+        Some(all) => resolve_child(all, superblock, group_address, name),
+        None => resolve_child_core(file_data, superblock, group_address, name),
+    }
+}
+
+fn resolve_child_core<S: Storage + ?Sized>(
+    file_data: &S,
+    superblock: &Superblock,
+    group_address: u64,
+    name: &str,
+) -> Result<u64, FormatError> {
+    let os = superblock.offset_size;
+    let ls = superblock.length_size;
+    let not_found = || FormatError::PathNotFound(String::from(name));
+    let header = ObjectHeader::parse_in(file_data, checked_addr(group_address)?, os, ls)?;
+    if !is_v2_group(&header) || is_v1_group(&header) {
+        return resolve_group_children_in(file_data, superblock, group_address)?
+            .into_iter()
+            .find(|e| e.name == name)
+            .map(|e| e.object_header_address)
+            .ok_or_else(not_found);
+    }
+    // The first link of that name only, as the listing (see
+    // `first_link_named`).
+    match first_link_named(file_data, &header, name, os, ls)?.map(|l| l.link_target) {
+        Some(LinkTarget::Hard {
+            object_header_address,
+        }) => Ok(object_header_address),
+        Some(LinkTarget::Soft { target_path }) => {
+            match resolve_path_from_in(file_data, superblock, group_address, &target_path) {
+                // Left out of the listing: dangling, cyclic, or in another file.
+                Err(
+                    FormatError::PathNotFound(_)
+                    | FormatError::NestingDepthExceeded
+                    | FormatError::ExternalLinkUnsupported { .. },
+                ) => Err(not_found()),
+                other => other,
+            }
+        }
+        Some(LinkTarget::External { .. }) | None => Err(not_found()),
+    }
+}
+
 /// Find and parse the Link Info message from an object header.
 fn find_link_info(
     object_header: &ObjectHeader,
@@ -254,6 +467,29 @@ pub fn resolve_path_any(
     file_data: &[u8],
     superblock: &Superblock,
     path: &str,
+) -> Result<u64, FormatError> {
+    resolve_path_any_core(file_data, superblock, path)
+}
+
+/// [`resolve_path_any`] over any [`Storage`]. One with the whole file in memory
+/// is read as the slice, by code compiled in this crate (see
+/// [`crate::storage`], "Slice entry points").
+#[inline]
+pub fn resolve_path_any_in<S: Storage + ?Sized>(
+    file_data: &S,
+    superblock: &Superblock,
+    path: &str,
+) -> Result<u64, FormatError> {
+    match file_data.as_contiguous() {
+        Some(all) => resolve_path_any(all, superblock, path),
+        None => resolve_path_any_core(file_data, superblock, path),
+    }
+}
+
+fn resolve_path_any_core<S: Storage + ?Sized>(
+    file_data: &S,
+    superblock: &Superblock,
+    path: &str,
 ) -> Result<u64, FormatError> {
     resolve_path_following_links(
         file_data,
@@ -272,6 +508,16 @@ pub fn resolve_path_from(
     superblock: &Superblock,
     group_address: u64,
     path: &str,
+) -> Result<u64, FormatError> {
+    resolve_path_from_in(file_data, superblock, group_address, path)
+}
+
+/// [`resolve_path_from`] over any [`Storage`].
+pub fn resolve_path_from_in<S: Storage + ?Sized>(
+    file_data: &S,
+    superblock: &Superblock,
+    group_address: u64,
+    path: &str,
 ) -> Result<u64, FormatError> {
     let start = if path.starts_with('/') {
         superblock.root_group_address
@@ -295,10 +541,33 @@ pub fn resolve_group_children(
     file_data: &[u8],
     superblock: &Superblock,
     group_address: u64,
+) -> Result<Vec<GroupEntry>, FormatError> {
+    resolve_group_children_core(file_data, superblock, group_address)
+}
+
+/// [`resolve_group_children`] over any [`Storage`]. One with the whole file in memory
+/// is read as the slice, by code compiled in this crate (see
+/// [`crate::storage`], "Slice entry points").
+#[inline]
+pub fn resolve_group_children_in<S: Storage + ?Sized>(
+    file_data: &S,
+    superblock: &Superblock,
+    group_address: u64,
+) -> Result<Vec<GroupEntry>, FormatError> {
+    match file_data.as_contiguous() {
+        Some(all) => resolve_group_children(all, superblock, group_address),
+        None => resolve_group_children_core(file_data, superblock, group_address),
+    }
+}
+
+fn resolve_group_children_core<S: Storage + ?Sized>(
+    file_data: &S,
+    superblock: &Superblock,
+    group_address: u64,
 ) -> Result<Vec<GroupEntry>, FormatError> {
     let os = superblock.offset_size;
     let ls = superblock.length_size;
-    let header = ObjectHeader::parse(file_data, group_address as usize, os, ls)?;
+    let header = ObjectHeader::parse_in(file_data, checked_addr(group_address)?, os, ls)?;
 
     let mut entries = Vec::new();
     let mut soft = Vec::new();
@@ -309,13 +578,19 @@ pub fn resolve_group_children(
             .find(|m| m.msg_type == MessageType::SymbolTable)
             .ok_or_else(|| FormatError::PathNotFound(String::from("no symbol table message")))?;
         let stm = SymbolTableMessage::parse(&sym_msg.data, os)?;
-        let all = group_v1::resolve_v1_group_entries(file_data, &stm, os, ls)?;
+        let all = group_v1::resolve_v1_group_entries_in(file_data, &stm, os, ls)?;
         if all.iter().any(group_v1::is_v1_soft_link) {
-            soft = group_v1::v1_soft_links(file_data, &stm, os, ls)?;
+            soft = group_v1::v1_soft_links_in(file_data, &stm, os, ls)?;
         }
         entries.extend(all.into_iter().filter(|e| !group_v1::is_v1_soft_link(e)));
     } else if is_v2_group(&header) {
-        let mut visit = |link: LinkMessage| match link.link_target {
+        // Only the first link of each name counts (see `first_link_named`).
+        let mut seen = BTreeSet::new();
+        let mut visit = |link: LinkMessage| {
+            if !seen.insert(link.name.clone()) {
+                return;
+            }
+            match link.link_target {
                 LinkTarget::Hard {
                     object_header_address,
                 } => entries.push(GroupEntry {
@@ -325,6 +600,7 @@ pub fn resolve_group_children(
                 }),
                 LinkTarget::Soft { target_path } => soft.push((link.name, target_path)),
                 LinkTarget::External { .. } => {}
+            }
         };
         let link_info = find_link_info(&header, os)?;
         if let Some(fh_addr) = link_info.fractal_heap_address {
@@ -341,7 +617,7 @@ pub fn resolve_group_children(
     }
 
     for (name, target) in soft {
-        match resolve_path_from(file_data, superblock, group_address, &target) {
+        match resolve_path_from_in(file_data, superblock, group_address, &target) {
             Ok(object_header_address) => entries.push(GroupEntry {
                 name,
                 object_header_address,
@@ -364,8 +640,8 @@ pub fn resolve_group_children(
 const MAX_SOFT_LINK_DEPTH: u8 = 16;
 
 /// Walk `path` from the group at `start`, following soft links.
-fn resolve_path_following_links(
+fn resolve_path_following_links<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     superblock: &Superblock,
     start: u64,
     path: &str,
@@ -383,24 +659,22 @@ fn resolve_path_following_links(
     let ls = superblock.length_size;
 
     let mut current_addr = start;
-    let mut current_header = ObjectHeader::parse(file_data, start as usize, os, ls)?;
+    let mut current_header = ObjectHeader::parse_in(file_data, checked_addr(start)?, os, ls)?;
 
     for (i, component) in components.iter().enumerate() {
-        let entries = resolve_group_entries(file_data, &current_header, os, ls)?;
+        match lookup_link(file_data, &current_header, component, os, ls)? {
-
+            Some(LinkTarget::Hard {
-        let found = entries
+                object_header_address,
-            .iter()
+            }) => {
-            .find(|e| e.name == *component && e.object_header_address != u64::MAX);
-        match found {
-            Some(entry) => {
                 if i == components.len() - 1 {
-                    return Ok(entry.object_header_address);
+                    return Ok(object_header_address);
                 }
-                current_addr = entry.object_header_address;
+                current_addr = object_header_address;
-                current_header = ObjectHeader::parse(file_data, current_addr as usize, os, ls)?;
+                current_header =
+                    ObjectHeader::parse_in(file_data, checked_addr(current_addr)?, os, ls)?;
             }
-            None => {
+            found => {
-                return match find_symbolic_link(file_data, &current_header, component, os, ls)? {
+                return match found {
                     Some(LinkTarget::Soft { target_path }) => {
                         if depth >= MAX_SOFT_LINK_DEPTH {
                             return Err(FormatError::NestingDepthExceeded);
@@ -436,8 +710,8 @@ fn resolve_path_following_links(
 }
 
 /// Resolve group entries from an object header, auto-detecting v1 vs v2.
-fn resolve_group_entries(
+fn resolve_group_entries<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     object_header: &ObjectHeader,
     offset_size: u8,
     length_size: u8,
@@ -450,9 +724,11 @@ fn resolve_group_entries(
             .find(|m| m.msg_type == MessageType::SymbolTable)
             .ok_or_else(|| FormatError::PathNotFound(String::from("no symbol table message")))?;
         let stm = SymbolTableMessage::parse(&sym_msg.data, offset_size)?;
-        group_v1::resolve_v1_group_entries(file_data, &stm, offset_size, length_size)
+        // A lookup: an entry with an empty name (which fails a listing) is
+        // skipped by the name comparison, as in libhdf5.
+        group_v1::v1_group_entries(file_data, &stm, offset_size, length_size)
     } else if is_v2_group(object_header) {
-        resolve_v2_group_entries(file_data, object_header, offset_size, length_size)
+        resolve_v2_group_entries_in(file_data, object_header, offset_size, length_size)
     } else {
         Err(FormatError::PathNotFound(String::from(
             "object header is not a group",

crates/clawhdf5-format/src/lane_partition.rs

@@ -112,7 +112,8 @@ pub fn partition(
 
     for idx in 0..num_items {
         let h = fxhash_combine(seed, idx as u64);
-        let lane = (h % num_lanes as u64) as usize;
+        // Below `num_lanes`, so it fits.
+        let lane = crate::addr::saturating_usize(h % num_lanes as u64);
         lanes[lane].push(idx);
     }
 

crates/clawhdf5-format/src/lib.rs

@@ -57,10 +57,12 @@
 #[cfg(not(feature = "std"))]
 extern crate alloc;
 
+pub mod addr;
 pub mod attribute;
 pub mod attribute_info;
 pub mod btree_v1;
 pub mod btree_v2;
+mod btree_v2_write;
 mod bulk_alloc;
 pub mod checksum;
 pub mod chunk_cache;
@@ -86,11 +88,15 @@ pub mod filters;
 mod filters_bitshuffle;
 #[cfg(feature = "blosc")]
 pub mod filters_blosc;
+#[cfg(feature = "blosc2")]
+pub mod filters_blosc2;
 #[cfg(feature = "bzip2")]
 mod filters_bzip2;
 #[cfg(feature = "lzf")]
 pub mod filters_lzf;
 mod filters_szip;
+#[cfg(feature = "zfp")]
+pub mod filters_zfp;
 pub mod fixed_array;
 pub mod float16;
 pub mod fractal_heap;
@@ -104,6 +110,7 @@ pub mod lane_partition;
 pub mod link_info;
 pub mod link_message;
 pub mod local_heap;
+pub mod lookup_stats;
 pub mod message_type;
 pub mod metadata_cache;
 pub mod metadata_index;
@@ -117,7 +124,9 @@ pub mod property_list;
 pub mod selection;
 pub mod shared_message;
 pub mod signature;
+pub mod storage;
 pub mod superblock;
+pub mod superblock_ext;
 pub mod symbol_table;
 #[cfg(all(
     test,

crates/clawhdf5-format/src/link_message.rs

@@ -3,6 +3,7 @@
 #[cfg(not(feature = "std"))]
 use alloc::{string::String, vec::Vec};
 
+use crate::addr::to_usize;
 use crate::datatype::CharacterSet;
 use crate::error::FormatError;
 
@@ -247,7 +248,7 @@ impl LinkMessage {
         };
 
         // Link name length
-        let name_len = read_offset(data, pos, name_size_field_width)? as usize;
+        let name_len = to_usize(read_offset(data, pos, name_size_field_width)?)?;
         pos += name_size_field_width as usize;
 
         // Link name

crates/clawhdf5-format/src/local_heap.rs

@@ -3,7 +3,9 @@
 #[cfg(not(feature = "std"))]
 use alloc::string::String;
 
+use crate::addr::to_usize;
 use crate::error::FormatError;
+use crate::storage::{Storage, len_usize, read_exact_at};
 
 /// Parsed HDF5 Local Heap header.
 #[derive(Debug, Clone)]
@@ -16,21 +18,6 @@ pub struct LocalHeap {
     pub data_segment_address: u64,
 }
 
-/// Checks that `[offset, offset + needed)` fits within `data`, guarding the
-/// addition against `usize` overflow from a crafted near-`usize::MAX` offset.
-fn ensure_len(data: &[u8], offset: usize, needed: usize) -> Result<(), FormatError> {
-    if offset
-        .checked_add(needed)
-        .is_none_or(|end| end > data.len())
-    {
-        return Err(FormatError::UnexpectedEof {
-            expected: offset.saturating_add(needed),
-            available: data.len(),
-        });
-    }
-    Ok(())
-}
-
 fn read_offset(data: &[u8], pos: usize, size: u8) -> Result<u64, FormatError> {
     let s = size as usize;
     if pos.checked_add(s).is_none_or(|end| end > data.len()) {
@@ -50,6 +37,10 @@ fn read_offset(data: &[u8], pos: usize, size: u8) -> Result<u64, FormatError> {
     })
 }
 
+/// First read of a name on a backend without the file in memory: most link
+/// names are shorter than this.
+const NAME_READ_START: usize = 64;
+
 impl LocalHeap {
     /// Parse a local heap header at the given offset in the file data.
     pub fn parse(
@@ -57,12 +48,24 @@ impl LocalHeap {
         offset: usize,
         offset_size: u8,
         length_size: u8,
+    ) -> Result<LocalHeap, FormatError> {
+        Self::parse_in(file_data, offset as u64, offset_size, length_size)
+    }
+
+    /// [`Self::parse`] over any [`Storage`]: one read of the header.
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
     ) -> Result<LocalHeap, FormatError> {
         // signature(4) + version(1) + reserved(3) = 8, then length_size*2 + offset_size
         let ls = length_size as usize;
         let os = offset_size as usize;
         let total = 8 + ls * 2 + os;
-        ensure_len(file_data, offset, total)?;
+        let header = read_exact_at(file, offset, total)?;
+        let file_data: &[u8] = &header;
+        let offset = 0usize;
 
         if &file_data[offset..offset + 4] != b"HEAP" {
             return Err(FormatError::InvalidLocalHeapSignature);
@@ -99,6 +102,16 @@ impl LocalHeap {
     /// The end of the list is `H5HL_FREE_NULL` (1); an all-ones value (the
     /// undefined address) is accepted as "no free list" too.
     pub fn validate_free_list(&self, file_data: &[u8], length_size: u8) -> Result<(), FormatError> {
+        self.validate_free_list_in(file_data, length_size)
+    }
+
+    /// [`Self::validate_free_list`] over any [`Storage`]: two small reads
+    /// per free block.
+    pub fn validate_free_list_in<S: Storage + ?Sized>(
+        &self,
+        file: &S,
+        length_size: u8,
+    ) -> Result<(), FormatError> {
         const FREE_NULL: u64 = 1;
         let ls = length_size as usize;
         let undefined = if ls >= 8 {
@@ -123,11 +136,12 @@ impl LocalHeap {
                 .and_then(|a| usize::try_from(a).ok())
                 .ok_or(FormatError::InvalidLocalHeapFreeList)?;
             let block_offset = next;
-            next = read_offset(file_data, at, length_size)?;
+            next = read_offset(&read_exact_at(file, at as u64, ls)?, 0, length_size)?;
             if next == 0 {
                 return Err(FormatError::InvalidLocalHeapFreeList);
             }
-            let block_size = read_offset(file_data, at + ls, length_size)?;
+            let block_size =
+                read_offset(&read_exact_at(file, (at + ls) as u64, ls)?, 0, length_size)?;
             if block_offset
                 .checked_add(block_size)
                 .is_none_or(|end| end > size)
@@ -140,43 +154,65 @@ impl LocalHeap {
 
     /// Read a null-terminated string from the heap's data segment at the given byte offset.
     pub fn read_string(&self, file_data: &[u8], string_offset: u64) -> Result<String, FormatError> {
-        let seg_addr = self.data_segment_address as usize;
+        self.read_string_in(file_data, string_offset)
+    }
+
+    /// [`Self::read_string`] over any [`Storage`]: one read of up to 64
+    /// bytes for a short name, more (each four times the last) up to the end
+    /// of the data segment for a longer one.
+    pub fn read_string_in<S: Storage + ?Sized>(
+        &self,
+        file: &S,
+        string_offset: u64,
+    ) -> Result<String, FormatError> {
+        let file_len = len_usize(file);
+        let seg_addr = to_usize(self.data_segment_address)?;
         let str_start =
             seg_addr
-                .checked_add(string_offset as usize)
+                .checked_add(to_usize(string_offset)?)
                 .ok_or(FormatError::Overflow(
                     "local heap seg_addr + string_offset overflow".into(),
                 ))?;
         let seg_end = seg_addr
-            .checked_add(self.data_segment_size as usize)
+            .checked_add(to_usize(self.data_segment_size)?)
             .ok_or(FormatError::Overflow(
                 "local heap seg_addr + data_segment_size overflow".into(),
             ))?;
 
-        if str_start >= file_data.len() || str_start >= seg_end {
+        if str_start >= file_len || str_start >= seg_end {
             return Err(FormatError::UnexpectedEof {
                 expected: str_start + 1,
-                available: file_data.len(),
+                available: file_len,
             });
         }
 
-        // Find null terminator
+        // Find the null terminator, which lies before the end of the data
-        let search_end = seg_end.min(file_data.len());
+        // segment (or of the file). In memory that is one borrowed slice;
-        let mut end = str_start;
+        // otherwise the bytes are read in growing pieces, so a name costs a
-        while end < search_end && file_data[end] != 0 {
+        // read of about its own length, not of the rest of the segment
-            end += 1;
+        // (whose size is an untrusted header field).
+        let search_end = seg_end.min(file_len);
+        let total = search_end - str_start;
+        let mut want = if file.as_contiguous().is_some() {
+            total
+        } else {
+            total.min(NAME_READ_START)
+        };
+        loop {
+            let rest = read_exact_at(file, str_start as u64, want)?;
+            if let Some(len) = rest.iter().position(|&b| b == 0) {
+                let s = core::str::from_utf8(&rest[..len])
+                    .map_err(|_| FormatError::InvalidLocalHeapSignature)?;
+                return Ok(String::from(s));
             }
-
+            if want == total {
-        if end >= search_end {
                 return Err(FormatError::UnexpectedEof {
-                expected: end + 1,
+                    expected: search_end + 1,
                     available: search_end,
                 });
             }
-
+            want = want.saturating_mul(4).min(total);
-        let s = core::str::from_utf8(&file_data[str_start..end])
+        }
-            .map_err(|_| FormatError::InvalidLocalHeapSignature)?;
-        Ok(String::from(s))
     }
 }
 
@@ -345,4 +381,77 @@ mod tests {
         let err = LocalHeap::parse(&file, 0, 8, 8).unwrap_err();
         assert_eq!(err, FormatError::InvalidLocalHeapVersion(1));
     }
+
+    /// Header, free list and strings read identically through a
+    /// `read_at`-only storage, for every truncation of the file.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::storage::CountingStorage;
+        let plain = build_heap_file(0, 64, &["", "alpha", "beta"], 8, 8);
+        // A free block of 16 bytes at segment offset 12, ending the list.
+        let mut free = build_heap_file(0, 64, &["", "alpha", "beta", &"x".repeat(20)], 8, 8);
+        free[16..24].copy_from_slice(&12u64.to_le_bytes());
+        free[64 + 12..64 + 20].copy_from_slice(&1u64.to_le_bytes());
+        free[64 + 20..64 + 28].copy_from_slice(&16u64.to_le_bytes());
+        let mut bad_free = free.clone();
+        bad_free[64 + 20..64 + 28].copy_from_slice(&99u64.to_le_bytes());
+        for full in [plain, free, bad_free] {
+            for cut in 0..=full.len() {
+                let f = &full[..cut];
+                let storage = CountingStorage::new(f.to_vec());
+                let want = LocalHeap::parse(f, 0, 8, 8);
+                let got = LocalHeap::parse_in(&storage, 0, 8, 8);
+                assert_eq!(format!("{got:?}"), format!("{want:?}"));
+                let Ok(heap) = want else { continue };
+                assert_eq!(
+                    heap.validate_free_list_in(&storage, 8),
+                    heap.validate_free_list(f, 8)
+                );
+                for off in [0u64, 1, 2, 6, 7, 11, 100] {
+                    assert_eq!(heap.read_string_in(&storage, off), heap.read_string(f, off));
+                }
+            }
+        }
+    }
+
+    /// Names of every length around the first read's size, and one with no
+    /// terminator, read identically through a `read_at`-only storage; a
+    /// short name in a heap whose header claims a huge data segment costs
+    /// one small read, not a read of the rest of the file.
+    #[test]
+    fn long_names_and_hostile_segment_sizes() {
+        use crate::storage::CountingStorage;
+        let names: Vec<String> = [0usize, 1, 63, 64, 65, 255, 256, 257, 1000, 5000]
+            .iter()
+            .map(|&n| "n".repeat(n))
+            .collect();
+        let refs: Vec<&str> = names.iter().map(String::as_str).collect();
+        let mut file = build_heap_file(0, 64, &refs, 8, 8);
+        let heap = LocalHeap::parse(&file, 0, 8, 8).unwrap();
+        let storage = CountingStorage::new(file.clone());
+        let mut off = 0u64;
+        for name in &names {
+            let got = heap.read_string_in(&storage, off);
+            assert_eq!(got, heap.read_string(&file, off));
+            assert_eq!(got.unwrap(), *name);
+            off += name.len() as u64 + 1;
+        }
+        // The last name loses its terminator: both report the same error.
+        let seg_end = 64 + heap.data_segment_size as usize;
+        file[seg_end - 1] = b'n';
+        let storage = CountingStorage::new(file.clone());
+        let last = off - names[names.len() - 1].len() as u64 - 1;
+        let want = heap.read_string(&file, last);
+        assert!(want.is_err());
+        assert_eq!(heap.read_string_in(&storage, last), want);
+
+        // A 64 MiB file whose heap claims a data segment reaching its end.
+        let mut big = build_heap_file(0, 64, &["short", "names"], 8, 8);
+        big.resize(64 << 20, 0);
+        big[8..16].copy_from_slice(&((64u64 << 20) - 64).to_le_bytes());
+        let heap = LocalHeap::parse(&big, 0, 8, 8).unwrap();
+        let storage = CountingStorage::new(big.clone());
+        assert_eq!(heap.read_string_in(&storage, 6).unwrap(), "names");
+        assert_eq!((storage.reads(), storage.bytes_read()), (1, 64));
+    }
 }

crates/clawhdf5-format/src/lookup_stats.rs

@@ -0,0 +1,31 @@
+//! Work counters for tests of lookup cost (feature `lookup-stats`).
+//!
+//! Counts fractal-heap objects read — each is one link or attribute message
+//! decoded out of a dense group or dense attribute storage — so a test can
+//! check that finding one name reads a handful of them, not the whole group.
+//! Per thread, so tests running in parallel do not see each other's reads.
+//! Without the feature the counting compiles to nothing.
+
+#[cfg(feature = "lookup-stats")]
+std::thread_local! {
+    static HEAP_OBJECTS: core::cell::Cell<u64> = const { core::cell::Cell::new(0) };
+}
+
+/// Record one heap object read.
+#[inline(always)]
+pub(crate) fn heap_object_read() {
+    #[cfg(feature = "lookup-stats")]
+    HEAP_OBJECTS.with(|c| c.set(c.get() + 1));
+}
+
+/// Heap objects read on this thread since the last [`reset`].
+#[cfg(feature = "lookup-stats")]
+pub fn heap_objects_read() -> u64 {
+    HEAP_OBJECTS.with(core::cell::Cell::get)
+}
+
+/// Zero this thread's counters.
+#[cfg(feature = "lookup-stats")]
+pub fn reset() {
+    HEAP_OBJECTS.with(|c| c.set(0));
+}

crates/clawhdf5-format/src/object_header.rs

@@ -1,12 +1,16 @@
 //! HDF5 Object Header parsing (v1 and v2).
 
 #[cfg(not(feature = "std"))]
-use alloc::vec::Vec;
+use alloc::{boxed::Box, collections::BTreeSet, vec::Vec};
+#[cfg(feature = "std")]
+use std::collections::BTreeSet;
 
 use byteorder::{ByteOrder, LittleEndian};
 
+use crate::addr::to_usize;
 use crate::error::FormatError;
 use crate::message_type::MessageType;
+use crate::storage::{Storage, Window, len_usize, read_exact_at};
 
 /// OHDR signature for v2 object headers.
 const OHDR_SIGNATURE: [u8; 4] = *b"OHDR";
@@ -75,42 +79,122 @@ fn read_offset(data: &[u8], pos: usize, size: u8) -> Result<u64, FormatError> {
     })
 }
 
+/// The kind of object an object header describes.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub enum ObjectClass {
+    /// A group: the header has a Symbol Table or a Link Info message.
+    Group,
+    /// A dataset: the header has a Datatype and a Dataspace message.
+    Dataset,
+    /// A committed (named) datatype: a Datatype message, no Dataspace.
+    NamedDatatype,
+}
+
 impl ObjectHeader {
+    /// The kind of object this header describes, decided as libhdf5 decides
+    /// it (`H5O__obj_class_real`): group first (a Symbol Table or Link Info
+    /// message), then dataset (a Datatype *and* a Dataspace message — not a
+    /// Data Layout message), then named datatype (a Datatype message).
+    /// `None` when none applies; libhdf5 then cannot open the object
+    /// ("unable to determine object type").
+    ///
+    /// A header with a Datatype and a Data Layout message but no Dataspace
+    /// is a named datatype to libhdf5, not a dataset.
+    pub fn object_class(&self) -> Option<ObjectClass> {
+        let has = |t: MessageType| self.messages.iter().any(|m| m.msg_type == t);
+        if has(MessageType::SymbolTable) || has(MessageType::LinkInfo) {
+            Some(ObjectClass::Group)
+        } else if has(MessageType::Datatype) && has(MessageType::Dataspace) {
+            Some(ObjectClass::Dataset)
+        } else if has(MessageType::Datatype) {
+            Some(ObjectClass::NamedDatatype)
+        } else {
+            None
+        }
+    }
+
     /// Parse an object header at the given offset in the data buffer.
     ///
     /// `offset_size` and `length_size` come from the superblock.
+    #[inline]
     pub fn parse(
         data: &[u8],
         offset: usize,
         offset_size: u8,
         length_size: u8,
     ) -> Result<ObjectHeader, FormatError> {
-        ensure_len(data, offset, 4)?;
+        Self::parse_slice(data, offset as u64, offset_size, length_size)
-        if data[offset..offset + 4] == OHDR_SIGNATURE {
+    }
-            Self::parse_v2(data, offset, offset_size, length_size)
+
-        } else {
+    /// [`Self::parse`] over any [`Storage`].
-            Self::parse_v1(data, offset, offset_size, length_size)
+    ///
+    /// Reads the prefix (at most [`V2_PREFIX_MAX`] bytes, signature
+    /// included), then each chunk as one bounded read, continuation chunks
+    /// included. A storage with the whole file in memory is parsed as its
+    /// slice, by code compiled in this crate (see
+    /// [`crate::storage`], "Slice entry points").
+    #[inline]
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
+    ) -> Result<ObjectHeader, FormatError> {
+        match file.as_contiguous() {
+            Some(all) => Self::parse_slice(all, offset, offset_size, length_size),
+            None => Self::parse_storage(file, offset, offset_size, length_size),
         }
     }
 
-    fn parse_v1(
+    /// [`Self::parse_storage`] for the slice, compiled in this crate: the
+    /// one copy [`Self::parse`] and [`Self::parse_in`] (in memory) call.
+    fn parse_slice(
         data: &[u8],
-        offset: usize,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
+    ) -> Result<ObjectHeader, FormatError> {
+        Self::parse_storage(data, offset, offset_size, length_size)
+    }
+
+    fn parse_storage<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+        length_size: u8,
+    ) -> Result<ObjectHeader, FormatError> {
+        // The longest prefix of either version, in one read. It holds the
+        // whole prefix or ends at the end of the file, so its bounds checks
+        // are the whole-file ones.
+        let prefix = Window::read(file, offset, V2_PREFIX_MAX)?;
+        prefix.ensure(0, 4)?;
+        if prefix.bytes[..4] == OHDR_SIGNATURE {
+            Self::parse_v2(file, offset, &prefix, offset_size, length_size)
+        } else {
+            Self::parse_v1(file, offset, &prefix, offset_size, length_size)
+        }
+    }
+
+    fn parse_v1<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        prefix: &Window<'_>,
         offset_size: u8,
         length_size: u8,
     ) -> Result<ObjectHeader, FormatError> {
         // version(1) + reserved(1) + num_messages(2) + ref_count(4) + header_size(4) = 12
         // then pad to 8-byte alignment from start of header
-        ensure_len(data, offset, 12)?;
+        prefix.ensure(0, 12)?;
+        let prefix = &prefix.bytes[..12];
 
-        let version = data[offset];
+        let version = prefix[0];
         if version != 1 {
             return Err(FormatError::InvalidObjectHeaderVersion(version));
         }
 
-        let num_messages = LittleEndian::read_u16(&data[offset + 2..offset + 4]) as usize;
+        let num_messages = LittleEndian::read_u16(&prefix[2..4]) as usize;
-        let reference_count = LittleEndian::read_u32(&data[offset + 4..offset + 8]);
+        let reference_count = LittleEndian::read_u32(&prefix[4..8]);
-        let header_data_size = LittleEndian::read_u32(&data[offset + 8..offset + 12]) as usize;
+        let header_data_size = LittleEndian::read_u32(&prefix[8..12]) as usize;
 
         // libhdf5 (H5O__prefix_deserialize): a header with messages needs room
         // for at least one message header, and one without has an empty chunk.
@@ -128,19 +212,19 @@ impl ObjectHeader {
             .checked_add(12 + padding)
             .ok_or(FormatError::UnexpectedEof {
                 expected: usize::MAX,
-                available: data.len(),
+                available: len_usize(file),
             })?;
 
-        ensure_len(data, msg_start, header_data_size)?;
+        // parse_v1_chunk reads the chunk, with the bounds check that was here.
-
+        // The prefix's count (NIL messages included, capped: it is untrusted)
-        let mut messages = Vec::new();
+        // sizes the list once instead of growing it message by message.
+        let mut messages = Vec::with_capacity(num_messages.min(64));
         let chunk0_count = Self::parse_v1_chunk(
-            data,
+            file,
             msg_start,
             header_data_size,
             offset_size,
             length_size,
-            MAX_V1_CONTINUATION_DEPTH,
             &mut messages,
         )?;
         // libhdf5 reads every message in the first chunk and refuses a header
@@ -174,24 +258,54 @@ impl ObjectHeader {
     /// 8; libhdf5 refuses a message that is not aligned, that runs past the
     /// end of the chunk, or leftover bytes too few for a message header (a
     /// "gap", which only version 2 allows).
-    #[allow(clippy::too_many_arguments)]
+    ///
-    fn parse_v1_chunk(
+    /// Continuation chunks are read in the order their messages are found,
-        data: &[u8],
+    /// as `H5O_protect` loads them (so the messages keep libhdf5's order):
-        offset: usize,
+    /// a queue of (address, length) pairs, each chunk read, parsed and
+    /// released before the next, so only one chunk buffer is alive at a
+    /// time whatever the storage. Every chunk must start at a new address
+    /// (else a cycle), and the chunks together may be no larger than the
+    /// file, so the bytes read stay within the file's size; a header of
+    /// more than [`MAX_V1_CHUNKS`] chunks is refused.
+    fn parse_v1_chunk<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
         length: usize,
         offset_size: u8,
         length_size: u8,
-        depth_remaining: u16,
         messages: &mut Vec<HeaderMessage>,
     ) -> Result<usize, FormatError> {
-        if depth_remaining == 0 {
+        // The chunks found so far are also the queue of chunks to read.
-            return Err(FormatError::NestingDepthExceeded);
+        let mut spans = ChunkSpans::new(file.len(), offset, length)?;
+        let mut chunk0_count = 0usize;
+        let mut next = 0usize;
+        while let Some((chunk_offset, chunk_length)) = spans.get(next) {
+            let chunk = read_exact_at(file, chunk_offset, chunk_length)?;
+            let count =
+                Self::parse_v1_messages(&chunk, offset_size, length_size, messages, &mut spans)?;
+            // Only the first chunk's messages are held to the prefix count.
+            if next == 0 {
+                chunk0_count = count;
+            }
+            next += 1;
+        }
+        Ok(chunk0_count)
     }
-        ensure_len(data, offset, length)?;
-        let end = offset + length;
-        let mut pos = offset;
-        let mut count = 0usize;
 
+    /// The messages of one version-1 chunk: each checked and appended to
+    /// `messages` (NIL ones dropped), each continuation added to `spans`.
+    /// Returns how many messages (NIL ones included) the chunk holds.
+    #[inline(never)]
+    fn parse_v1_messages(
+        data: &[u8],
+        offset_size: u8,
+        length_size: u8,
+        messages: &mut Vec<HeaderMessage>,
+        spans: &mut ChunkSpans,
+    ) -> Result<usize, FormatError> {
+        let end = data.len();
+        let mut pos = 0usize;
+        let mut count = 0usize;
         while pos < end {
             if end - pos < V1_MSG_HEADER_SIZE {
                 return Err(FormatError::InvalidObjectHeader(
@@ -215,7 +329,6 @@ impl ObjectHeader {
             let body = &data[pos..pos + msg_data_size];
             check_message(1, msg_type_raw, msg_flags, body, offset_size, length_size)?;
             count += 1;
-
             let msg_type = MessageType::from_u16(msg_type_raw);
             if msg_type != MessageType::Nil {
                 messages.push(HeaderMessage {
@@ -226,34 +339,34 @@ impl ObjectHeader {
                     data: body.to_vec(),
                 });
             }
-            pos += msg_data_size;
+            // Queue continuations (v1 continuation chunks are just raw
-
-            // Follow continuations (v1 continuation chunks are just raw
             // messages, no signature); check_message has checked the body.
             if msg_type == MessageType::ObjectHeaderContinuation {
-                let cont_offset = read_offset(body, 0, offset_size)? as usize;
+                let cont_offset = read_offset(body, 0, offset_size)?;
-                let cont_length = read_offset(body, offset_size as usize, length_size)? as usize;
+                let cont_length = to_usize(read_offset(body, offset_size as usize, length_size)?)?;
-                Self::parse_v1_chunk(
+                spans.add(cont_offset, cont_length)?;
-                    data,
-                    cont_offset,
-                    cont_length,
-                    offset_size,
-                    length_size,
-                    depth_remaining - 1,
-                    messages,
-                )?;
             }
+            pos += msg_data_size;
         }
-
         Ok(count)
     }
 
-    fn parse_v2(
+    fn parse_v2<S: Storage + ?Sized>(
-        data: &[u8],
+        file: &S,
-        offset: usize,
+        offset: u64,
+        prefix: &Window<'_>,
         offset_size: u8,
         length_size: u8,
     ) -> Result<ObjectHeader, FormatError> {
+        // `ensure_len` checks positions relative to the header against the
+        // prefix window and reports them as the whole-file check did, with
+        // absolute positions and the file's length.
+        let data: &[u8] = &prefix.bytes;
+        let file_len = len_usize(file);
+        let base = usize::try_from(offset).unwrap_or(usize::MAX);
+        let abs = |rel: usize| base.saturating_add(rel);
+        let ensure_len = |_: &[u8], rel: usize, needed: usize| prefix.ensure(rel, needed);
+        let offset = 0usize;
         // signature(4) + version(1) + flags(1) = 6
         ensure_len(data, offset, 6)?;
 
@@ -306,7 +419,7 @@ impl ObjectHeader {
             _ => unreachable!(),
         };
         ensure_len(data, pos, chunk_size_width as usize)?;
-        let chunk0_size = read_offset(data, pos, chunk_size_width)? as usize;
+        let chunk0_size = to_usize(read_offset(data, pos, chunk_size_width)?)?;
         pos += chunk_size_width as usize;
         // Bit 2: attribute creation order tracked → messages include creation order field
         let has_creation_order = flags & 0x04 != 0;
@@ -318,15 +431,20 @@ impl ObjectHeader {
         }
 
         let chunk0_msg_start = pos;
-        let chunk0_msg_end = pos
+        let Some(chunk0_abs_end) = abs(pos).checked_add(chunk0_size) else {
-            .checked_add(chunk0_size)
+            return Err(FormatError::UnexpectedEof {
-            .ok_or(FormatError::UnexpectedEof {
                 expected: usize::MAX,
-                available: data.len(),
+                available: file_len,
-            })?;
+            });
+        };
+        let chunk0_msg_end = chunk0_abs_end - base;
+
+        // The whole first chunk, prefix to checksum, in one read (its
+        // bounds check is the one on the checksum's 4 bytes).
+        let chunk0 = read_exact_at(file, base as u64, chunk0_msg_end.saturating_add(4))?;
+        let data: &[u8] = &chunk0;
 
         // Validate checksum: from OHDR signature through all messages (before checksum)
-        ensure_len(data, chunk0_msg_end, 4)?;
         #[cfg(feature = "checksum")]
         {
             let stored = LittleEndian::read_u32(&data[chunk0_msg_end..chunk0_msg_end + 4]);
@@ -353,16 +471,17 @@ impl ObjectHeader {
             &mut continuations,
         )?;
 
-        // Follow continuations (limit to prevent cycles in malformed data)
+        // Follow continuations, one chunk buffer at a time. A chunk address
-        let mut cont_remaining = 256u16;
+        // seen twice is a cycle in malformed data, and the chunks may add up
+        // to no more than the file; a valid header can have many chunks (libhdf5 adds one
+        // whenever a message no longer fits), up to the same bound as a
+        // version-1 header.
+        let mut spans = ChunkSpans::new(file.len(), base as u64, chunk0_msg_end.saturating_add(4))?;
         while let Some((cont_offset, cont_length)) = continuations.pop() {
-            if cont_remaining == 0 {
+            spans.add(cont_offset as u64, cont_length)?;
-                return Err(FormatError::NestingDepthExceeded);
-            }
-            cont_remaining -= 1;
             Self::parse_v2_continuation(
-                data,
+                file,
-                cont_offset,
+                cont_offset as u64,
                 cont_length,
                 has_creation_order,
                 offset_size,
@@ -439,8 +558,8 @@ impl ObjectHeader {
             let msg_type = MessageType::from_u16(msg_type_raw);
             if msg_type == MessageType::ObjectHeaderContinuation {
                 // check_message has checked the body holds both fields.
-                let cont_off = read_offset(body, 0, offset_size)? as usize;
+                let cont_off = to_usize(read_offset(body, 0, offset_size)?)?;
-                let cont_len = read_offset(body, offset_size as usize, length_size)? as usize;
+                let cont_len = to_usize(read_offset(body, offset_size as usize, length_size)?)?;
                 continuations.push((cont_off, cont_len));
             } else if msg_type == MessageType::Nil {
                 null_count += 1;
@@ -461,9 +580,9 @@ impl ObjectHeader {
     }
 
     #[allow(clippy::too_many_arguments)]
-    fn parse_v2_continuation(
+    fn parse_v2_continuation<S: Storage + ?Sized>(
-        data: &[u8],
+        file: &S,
-        offset: usize,
+        offset: u64,
         length: usize,
         has_creation_order: bool,
         offset_size: u8,
@@ -472,7 +591,9 @@ impl ObjectHeader {
         continuations: &mut Vec<(usize, usize)>,
     ) -> Result<(), FormatError> {
         // OCHK signature(4) + messages + checksum(4)
-        ensure_len(data, offset, length)?;
+        let chunk = read_exact_at(file, offset, length)?;
+        let data: &[u8] = &chunk;
+        let offset = 0usize;
         if length < 8 {
             return Err(FormatError::UnexpectedEof {
                 expected: 8,
@@ -513,11 +634,114 @@ impl ObjectHeader {
     }
 }
 
+/// Longest version-2 object header prefix: signature(4) + version(1) +
+/// flags(1) + times(16) + attribute phase change(4) + chunk-0 size(8).
+const V2_PREFIX_MAX: usize = 34;
+
 /// Size of a version-1 message header: type(2) + size(2) + flags(1) + reserved(3).
 const V1_MSG_HEADER_SIZE: usize = 8;
 
-/// How deep version-1 continuation chunks may chain (malformed-data guard).
+/// The chunks of one object header read so far, in the order they were
-const MAX_V1_CONTINUATION_DEPTH: u16 = 32;
+/// found (which is the order version-1 chunks are read in). A chunk starting
+/// where another did is a cycle. Chunks of a valid header do not overlap, so
+/// together they are no larger than the file; a header whose chunks add up
+/// to more is refused, which bounds what its chunks can make a reader read
+/// (a crafted chain of chunks each nested in the last would otherwise read
+/// the file over and over). Overlap itself is not refused: libhdf5 reads
+/// such headers (`cve-2025-7067.h5` has one).
+///
+/// Almost every header has at most a few chunks, and this runs once per
+/// header, so the first [`INLINE_CHUNKS`] live in an inline array and are
+/// checked for cycles by a scan; only a longer header allocates (the rest
+/// of the list, and a set of starts). Allocating a queue and a set for
+/// every header made parsing 401 small headers 1.8x slower.
+struct ChunkSpans {
+    inline: [(u64, usize); INLINE_CHUNKS],
+    /// Chunks after the first [`INLINE_CHUNKS`], and every chunk start.
+    spill: Option<Box<SpilledSpans>>,
+    /// How many chunks there are.
+    len: usize,
+    /// Bytes of the chunks so far, and the most they may add up to.
+    total: u64,
+    budget: u64,
+}
+
+/// The chunks of a [`ChunkSpans`] beyond its inline ones.
+struct SpilledSpans {
+    chunks: Vec<(u64, usize)>,
+    starts: BTreeSet<u64>,
+}
+
+/// How many chunks [`ChunkSpans`] holds without allocating.
+const INLINE_CHUNKS: usize = 8;
+
+impl ChunkSpans {
+    #[inline]
+    fn new(file_len: u64, start: u64, len: usize) -> Result<Self, FormatError> {
+        let mut s = Self {
+            inline: [(0, 0); INLINE_CHUNKS],
+            spill: None,
+            len: 0,
+            total: 0,
+            budget: file_len,
+        };
+        s.add(start, len)?;
+        Ok(s)
+    }
+
+    /// Record the chunk `len` bytes at `start`.
+    #[inline]
+    fn add(&mut self, start: u64, len: usize) -> Result<(), FormatError> {
+        self.total = self.total.saturating_add(len as u64);
+        if self.len < INLINE_CHUNKS {
+            if self.inline[..self.len].iter().any(|&(s, _)| s == start) {
+                return Err(FormatError::NestingDepthExceeded);
+            }
+            self.inline[self.len] = (start, len);
+        } else {
+            self.add_spilled(start, len)?;
+        }
+        self.len += 1;
+        if self.total > self.budget {
+            return Err(FormatError::InvalidObjectHeader(
+                "object header chunks larger than the file",
+            ));
+        }
+        Ok(())
+    }
+
+    #[cold]
+    #[inline(never)]
+    fn add_spilled(&mut self, start: u64, len: usize) -> Result<(), FormatError> {
+        let inline = &self.inline;
+        let spill = self.spill.get_or_insert_with(|| {
+            Box::new(SpilledSpans {
+                chunks: Vec::new(),
+                starts: inline.iter().map(|&(s, _)| s).collect(),
+            })
+        });
+        if !spill.starts.insert(start) || self.len >= MAX_V1_CHUNKS {
+            return Err(FormatError::NestingDepthExceeded);
+        }
+        spill.chunks.push((start, len));
+        Ok(())
+    }
+
+    /// The `i`th chunk recorded.
+    #[inline]
+    fn get(&self, i: usize) -> Option<(u64, usize)> {
+        if i < INLINE_CHUNKS {
+            (i < self.len).then(|| self.inline[i])
+        } else {
+            self.spill.as_ref()?.chunks.get(i - INLINE_CHUNKS).copied()
+        }
+    }
+}
+
+/// Most chunks a version-1 object header may have (malformed-data guard;
+/// libhdf5 has no limit, and a header that gains one continuation chunk per
+/// attribute added can have many).
+const MAX_V1_CHUNKS: usize = 1 << 16;
 
 /// Every defined version-2 object header status flag (libhdf5
 /// `H5O_HDR_ALL_FLAGS`): chunk-0 size width (bits 0-1), attribute creation
@@ -662,6 +886,54 @@ fn check_message(
 mod tests {
     use super::*;
 
+    fn header_with(types: &[MessageType]) -> ObjectHeader {
+        ObjectHeader {
+            version: 2,
+            messages: types
+                .iter()
+                .map(|&msg_type| HeaderMessage {
+                    msg_type,
+                    size: 0,
+                    flags: 0,
+                    creation_order: None,
+                    data: Vec::new(),
+                })
+                .collect(),
+            reference_count: None,
+            flags: 0,
+            access_time: None,
+            modification_time: None,
+            change_time: None,
+            birth_time: None,
+        }
+    }
+
+    #[test]
+    fn object_class_follows_libhdf5() {
+        use MessageType::*;
+        let class = |t: &[MessageType]| header_with(t).object_class();
+        assert_eq!(
+            class(&[Datatype, Dataspace, DataLayout]),
+            Some(ObjectClass::Dataset)
+        );
+        // A Data Layout message does not make a dataset without a dataspace
+        // (cve-2024-33874 `/Dset1`: h5py opens it as a named datatype).
+        assert_eq!(
+            class(&[Datatype, DataLayout]),
+            Some(ObjectClass::NamedDatatype)
+        );
+        assert_eq!(class(&[Datatype]), Some(ObjectClass::NamedDatatype));
+        // Group messages win over dataset messages.
+        assert_eq!(
+            class(&[Datatype, Dataspace, SymbolTable]),
+            Some(ObjectClass::Group)
+        );
+        assert_eq!(class(&[LinkInfo]), Some(ObjectClass::Group));
+        // Link messages alone are not a group; nothing is not an object.
+        assert_eq!(class(&[Link]), None);
+        assert_eq!(class(&[]), None);
+    }
+
     // Helper: build a v1 object header with given messages
     fn build_v1_header(
         messages: &[(u16, &[u8], u8)], // (type, data, flags)
@@ -673,13 +945,13 @@ mod tests {
         let mut msg_bytes = Vec::new();
         for (mtype, mdata, mflags) in messages {
             // v1 message sizes are multiples of 8 (the data is zero-padded).
-            let padded = mdata.len().div_ceil(8) * 8;
+            let padded = <[u8]>::len(mdata).div_ceil(8) * 8;
             msg_bytes.extend_from_slice(&mtype.to_le_bytes()); // type(2)
             msg_bytes.extend_from_slice(&(padded as u16).to_le_bytes()); // size(2)
             msg_bytes.push(*mflags); // flags(1)
             msg_bytes.extend_from_slice(&[0u8; 3]); // reserved(3)
             msg_bytes.extend_from_slice(mdata); // data
-            msg_bytes.resize(msg_bytes.len() + padded - mdata.len(), 0);
+            msg_bytes.resize(msg_bytes.len() + padded - <[u8]>::len(mdata), 0);
         }
 
         let mut buf = Vec::new();
@@ -775,6 +1047,176 @@ mod tests {
         assert_eq!(hdr.messages[1].data[..2], [5, 6]);
     }
 
+    /// A version-1 header whose continuation chunks form a chain: chunk k
+    /// holds a Dataspace message `[k]` and the continuation to chunk k + 1.
+    /// With `cycle`, the last chunk points back at the first continuation
+    /// chunk.
+    fn v1_chain(n: usize, cycle: bool) -> Vec<u8> {
+        // Each continuation chunk: dataspace (8 + 8) + continuation (8 + 16).
+        let chunk_len = 40u64;
+        let first = 64u64;
+        let cont = |addr: u64| {
+            let mut b = addr.to_le_bytes().to_vec();
+            b.extend_from_slice(&chunk_len.to_le_bytes());
+            b
+        };
+        let mut data = build_v1_header(&[(0x0010, &cont(first)[..], 0)], 8, 8);
+        data.resize(first as usize, 0);
+        for k in 0..n {
+            let mut c = Vec::new();
+            c.extend_from_slice(&1u16.to_le_bytes());
+            c.extend_from_slice(&8u16.to_le_bytes());
+            c.extend_from_slice(&[0; 4]);
+            c.extend_from_slice(&(k as u64).to_le_bytes());
+            let next = if k + 1 < n {
+                first + (k as u64 + 1) * chunk_len
+            } else if cycle {
+                first
+            } else {
+                // The last chunk ends in a NIL message instead.
+                c.extend_from_slice(&[0, 0, 16, 0, 0, 0, 0, 0]);
+                c.extend_from_slice(&[0; 16]);
+                data.extend_from_slice(&c);
+                continue;
+            };
+            c.extend_from_slice(&0x10u16.to_le_bytes());
+            c.extend_from_slice(&16u16.to_le_bytes());
+            c.extend_from_slice(&[0; 4]);
+            c.extend_from_slice(&cont(next));
+            data.extend_from_slice(&c);
+        }
+        data
+    }
+
+    /// libhdf5 reads any chain of continuation chunks (a header grows one
+    /// per attribute added when full); the reader used to stop at 32.
+    #[test]
+    fn long_v1_continuation_chains_are_read() {
+        let data = v1_chain(200, false);
+        let hdr = ObjectHeader::parse(&data, 0, 8, 8).unwrap();
+        let spaces: Vec<u8> = hdr
+            .messages
+            .iter()
+            .filter(|m| m.msg_type == MessageType::Dataspace)
+            .map(|m| m.data[0])
+            .collect();
+        assert_eq!(spaces, (0..200).map(|k| k as u8).collect::<Vec<_>>());
+    }
+
+    /// A crafted version-1 header whose continuation chunks nest: each
+    /// chunk's continuation message points at the rest of that chunk. Read
+    /// depth-first with every enclosing chunk kept alive, from storage that
+    /// hands out owned buffers, it read n^2 bytes and held them all at once
+    /// (a 192 KB file read 768 MB). Chunks adding up to more than the file
+    /// are refused, and the bytes read stay within the file's size.
+    #[test]
+    fn nested_v1_continuation_chunks_are_bounded() {
+        use crate::storage::CountingStorage;
+        let n = 2000u64;
+        let a = 64u64;
+        let cont = |addr: u64, len: u64| {
+            let mut m = vec![0x10, 0, 16, 0, 0, 0, 0, 0];
+            m.extend_from_slice(&addr.to_le_bytes());
+            m.extend_from_slice(&len.to_le_bytes());
+            m
+        };
+        // Prefix: version 1, one message, reference count 1, 24 bytes.
+        let mut buf = vec![1, 0, 1, 0, 1, 0, 0, 0, 24, 0, 0, 0, 0, 0, 0, 0];
+        buf.extend_from_slice(&cont(a, 24 * n));
+        buf.resize(a as usize, 0);
+        for k in 0..n {
+            if k + 1 < n {
+                buf.extend_from_slice(&cont(a + 24 * (k + 1), 24 * (n - k - 1)));
+            } else {
+                buf.extend_from_slice(&[0, 0, 16, 0, 0, 0, 0, 0]);
+                buf.extend_from_slice(&[0; 16]);
+            }
+        }
+        let len = buf.len() as u64;
+        let s = CountingStorage::new(buf);
+        assert!(matches!(
+            ObjectHeader::parse_in(&s, 0, 8, 8),
+            Err(FormatError::InvalidObjectHeader(
+                "object header chunks larger than the file"
+            ))
+        ));
+        assert!(
+            s.bytes_read() <= 2 * len,
+            "read {} of {len}",
+            s.bytes_read()
+        );
+    }
+
+    /// libhdf5 reads a continuation chunk that overlaps the chunk holding
+    /// its message (`cve-2025-7067.h5` has one), and so does this reader.
+    #[test]
+    fn overlapping_v1_continuation_chunk_is_read() {
+        // Chunk 0 (at 16): continuation (24 bytes), then a NIL message at
+        // 40; the continuation chunk is that NIL message's 8-byte header.
+        let mut cont = 40u64.to_le_bytes().to_vec();
+        cont.extend_from_slice(&8u64.to_le_bytes());
+        let data = build_v1_header(&[(0x0010, &cont[..], 0), (0x0000, &[][..], 0)], 8, 8);
+        let hdr = ObjectHeader::parse(&data, 0, 8, 8).unwrap();
+        assert_eq!(hdr.messages.len(), 1);
+    }
+
+    /// A valid chain over owned-buffer storage reads each chunk once.
+    #[test]
+    fn long_v1_chain_reads_each_chunk_once() {
+        use crate::storage::CountingStorage;
+        let data = v1_chain(3000, false);
+        let len = data.len() as u64;
+        let s = CountingStorage::new(data);
+        let hdr = ObjectHeader::parse_in(&s, 0, 8, 8).unwrap();
+        assert_eq!(
+            hdr.messages
+                .iter()
+                .filter(|m| m.msg_type == MessageType::Dataspace)
+                .count(),
+            3000
+        );
+        assert!(s.bytes_read() <= len, "read {} of {len}", s.bytes_read());
+    }
+
+    /// Continuation chunks are read in the order their messages are found
+    /// (libhdf5's `H5O_protect`), so a chunk's messages follow every
+    /// message of the chunk before, not the continuation message.
+    #[test]
+    fn v1_continuation_messages_keep_libhdf5_order() {
+        // Chunk 0: continuation to A, dataspace [1]; A: dataspace [2].
+        let a = 64u64;
+        let mut cont = a.to_le_bytes().to_vec();
+        cont.extend_from_slice(&16u64.to_le_bytes());
+        let mut data = build_v1_header(&[(0x0010, &cont[..], 0), (0x0001, &[1; 8][..], 0)], 8, 8);
+        data.resize(a as usize, 0);
+        data.extend_from_slice(&[1, 0, 8, 0, 0, 0, 0, 0]);
+        data.extend_from_slice(&[2; 8]);
+        let hdr = ObjectHeader::parse(&data, 0, 8, 8).unwrap();
+        let spaces: Vec<u8> = hdr
+            .messages
+            .iter()
+            .filter(|m| m.msg_type == MessageType::Dataspace)
+            .map(|m| m.data[0])
+            .collect();
+        assert_eq!(spaces, [1, 2]);
+    }
+
+    #[test]
+    fn v1_continuation_cycles_are_refused() {
+        // Within the inline chunk list, and past it (the cycle returns to
+        // an inline chunk once the list has spilled).
+        for n in [5, 7, 8, 9, 40] {
+            let data = v1_chain(n, true);
+            assert!(
+                matches!(
+                    ObjectHeader::parse(&data, 0, 8, 8),
+                    Err(FormatError::NestingDepthExceeded)
+                ),
+                "{n} chunks"
+            );
+        }
+    }
+
     #[test]
     fn parse_v1_unknown_message_ok() {
         let messages = [(0x00FFu16, &[0xAA, 0xBB][..], 0u8)];
@@ -1199,4 +1641,56 @@ mod tests {
         let err = ObjectHeader::parse(&data, 0, 8, 8).unwrap_err();
         assert!(matches!(err, FormatError::UnexpectedEof { .. }));
     }
+
+    /// Every header, and every truncation of it, parses to the same result
+    /// (or the same error) through a `read_at`-only storage as from a slice;
+    /// a header in one chunk takes two reads (prefix, chunk).
+    #[test]
+    fn parse_in_matches_slice_parse() {
+        use crate::storage::CountingStorage;
+        let mut headers = vec![
+            build_v1_header(&[], 8, 8),
+            build_v1_header(&[(0x0001, &[1, 2, 3], 0), (0x0003, &[9; 8], 0)], 8, 8),
+            build_v2_header(0x00, &[(0x01, &[42], 0)], None),
+            build_v2_header(0x03, &[(0x01, &[1, 2], 0), (0x03, &[3], 0)], None),
+            build_v2_header(0x24, &[(0x01, &[1], 0)], Some((1, 2, 3, 4))),
+            build_v2_header(0x35, &[(0x01, &[1], 0)], Some((5, 6, 7, 8))),
+        ];
+        // A v2 header with a continuation chunk at 256.
+        let mut ochk = OCHK_SIGNATURE.to_vec();
+        ochk.extend_from_slice(&[0x03, 2, 0, 0, 0xDE, 0xAD]);
+        let sum = crate::checksum::jenkins_lookup3(&ochk);
+        ochk.extend_from_slice(&sum.to_le_bytes());
+        let mut cont = 256u64.to_le_bytes().to_vec();
+        cont.extend_from_slice(&(ochk.len() as u64).to_le_bytes());
+        let main = build_v2_header(0x00, &[(0x01, &[42], 0), (0x10, &cont, 0)], None);
+        let mut with_cont = vec![0u8; 256 + ochk.len()];
+        with_cont[..main.len()].copy_from_slice(&main);
+        with_cont[256..].copy_from_slice(&ochk);
+        headers.push(with_cont);
+
+        for h in headers {
+            for at in [0usize, 3] {
+                for cut in 0..=h.len() {
+                    let mut f = vec![0u8; at];
+                    f.extend_from_slice(&h[..cut]);
+                    if at == 0 && cut == h.len() {
+                        f.resize(f.len() + 64, 0);
+                    }
+                    let want = ObjectHeader::parse(&f, at, 8, 8);
+                    let storage = CountingStorage::new(f.clone());
+                    let got = ObjectHeader::parse_in(&storage, at as u64, 8, 8);
+                    assert_eq!(
+                        format!("{got:?}"),
+                        format!("{want:?}"),
+                        "at {at}, cut {cut}"
+                    );
+                }
+            }
+        }
+        let one_chunk = build_v2_header(0x00, &[(0x01, &[42], 0)], None);
+        let storage = CountingStorage::new(one_chunk);
+        ObjectHeader::parse_in(&storage, 0, 8, 8).unwrap();
+        assert_eq!(storage.reads(), 2);
+    }
 }

crates/clawhdf5-format/src/object_header_writer.rs

@@ -12,9 +12,16 @@ use crate::message_type::MessageType;
 /// its size truncated to 16 bits produced files libhdf5 refuses.
 pub const MAX_MESSAGE_SIZE: usize = u16::MAX as usize;
 
+/// Object header flags: attribute creation order tracked (each message
+/// then carries a 2-byte creation order) and indexed.
+const OHDR_ATTR_CRT_ORDER_TRACKED: u8 = 0x04;
+const OHDR_ATTR_CRT_ORDER_INDEXED: u8 = 0x08;
+
 /// Writer for v2 object headers with proper checksums.
 pub struct ObjectHeaderWriter {
-    messages: Vec<(MessageType, Vec<u8>, u8)>, // (type, data, msg_flags)
+    messages: Vec<(MessageType, Vec<u8>, u8, u16)>, // (type, data, msg_flags, creation order)
+    /// Attribute creation order tracked and indexed.
+    attr_order: bool,
 }
 
 impl ObjectHeaderWriter {
@@ -22,17 +29,33 @@ impl ObjectHeaderWriter {
     pub fn new() -> Self {
         Self {
             messages: Vec::new(),
+            attr_order: false,
         }
     }
 
+    /// Track and index attribute creation order, as libhdf5 does for an
+    /// object created with `H5P_CRT_ORDER_TRACKED | H5P_CRT_ORDER_INDEXED`
+    /// (h5py's `track_order=True`): the header's flags say so, and every
+    /// message carries a creation order (an attribute's own; 0 for the
+    /// others). libhdf5 reads the setting back from these flags.
+    pub fn track_attr_order(&mut self) {
+        self.attr_order = true;
+    }
+
     /// Add a message to the header with default flags (0).
     pub fn add_message(&mut self, msg_type: MessageType, data: Vec<u8>) {
-        self.messages.push((msg_type, data, 0));
+        self.messages.push((msg_type, data, 0, 0));
     }
 
     /// Add a message with specific flags.
     pub fn add_message_with_flags(&mut self, msg_type: MessageType, data: Vec<u8>, flags: u8) {
-        self.messages.push((msg_type, data, flags));
+        self.messages.push((msg_type, data, flags, 0));
+    }
+
+    /// Add a message with its creation order, which is written only when
+    /// attribute creation order is tracked ([`Self::track_attr_order`]).
+    pub fn add_message_with_order(&mut self, msg_type: MessageType, data: Vec<u8>, order: u16) {
+        self.messages.push((msg_type, data, 0, order));
     }
 
     /// Serialize the complete v2 object header (OHDR + messages + checksum).
@@ -41,10 +64,10 @@ impl ObjectHeaderWriter {
     /// than [`MAX_MESSAGE_SIZE`] (e.g. an attribute over ~64 KiB, which would
     /// need dense attribute storage), rather than writing a corrupt header.
     pub fn serialize(&self) -> Result<Vec<u8>, FormatError> {
-        if let Some((msg_type, data, _)) = self
+        if let Some((msg_type, data, _, _)) = self
             .messages
             .iter()
-            .find(|(_, data, _)| data.len() > MAX_MESSAGE_SIZE)
+            .find(|(_, data, _, _)| data.len() > MAX_MESSAGE_SIZE)
         {
             return Err(FormatError::SerializationError(format!(
                 "{msg_type:?} message is {} bytes; an object header message holds at most \
@@ -52,11 +75,13 @@ impl ObjectHeaderWriter {
                 data.len()
             )));
         }
-        // Calculate total message bytes: each message has type(1) + size(2) + flags(1) + data
+        // Calculate total message bytes: each message has type(1) + size(2) +
+        // flags(1) [+ creation order(2)] + data
+        let msg_header = if self.attr_order { 6 } else { 4 };
         let msg_bytes_total: usize = self
             .messages
             .iter()
-            .map(|(_, data, _)| 4 + data.len())
+            .map(|(_, data, _, _)| msg_header + data.len())
             .sum();
 
         // Determine chunk size field width based on msg_bytes_total
@@ -68,6 +93,12 @@ impl ObjectHeaderWriter {
             (0x02u8, 4)
         };
 
+        let flags = if self.attr_order {
+            flags | OHDR_ATTR_CRT_ORDER_TRACKED | OHDR_ATTR_CRT_ORDER_INDEXED
+        } else {
+            flags
+        };
+
         let mut buf = Vec::new();
 
         // OHDR signature
@@ -85,7 +116,7 @@ impl ObjectHeaderWriter {
         }
 
         // Messages
-        for (msg_type, data, msg_flags) in &self.messages {
+        for (msg_type, data, msg_flags, order) in &self.messages {
             let type_id = msg_type.to_u16();
             assert!(
                 type_id <= 255,
@@ -94,6 +125,9 @@ impl ObjectHeaderWriter {
             buf.push(type_id as u8); // type (1 byte in v2)
             buf.extend_from_slice(&(data.len() as u16).to_le_bytes()); // size (2 bytes)
             buf.push(*msg_flags); // flags
+            if self.attr_order {
+                buf.extend_from_slice(&order.to_le_bytes()); // creation order
+            }
             buf.extend_from_slice(data);
         }
 
@@ -193,6 +227,21 @@ mod tests {
         assert_eq!(hdr.messages.len(), 0);
     }
 
+    #[test]
+    fn tracked_attribute_order_is_in_the_flags_and_every_message() {
+        let mut writer = ObjectHeaderWriter::new();
+        writer.track_attr_order();
+        writer.add_message(MessageType::Dataspace, vec![1, 2, 3, 4]);
+        writer.add_message_with_order(MessageType::Attribute, vec![5, 6], 7);
+        let bytes = writer.serialize().unwrap();
+        assert_eq!(bytes[5] & 0x0C, 0x0C);
+        let hdr = ObjectHeader::parse(&bytes, 0, 8, 8).unwrap();
+        assert_eq!(hdr.messages.len(), 2);
+        assert_eq!(hdr.messages[0].creation_order, Some(0));
+        assert_eq!(hdr.messages[1].creation_order, Some(7));
+        assert_eq!(hdr.messages[1].data, vec![5, 6]);
+    }
+
     #[test]
     fn two_messages_roundtrip() {
         let mut writer = ObjectHeaderWriter::new();

crates/clawhdf5-format/src/parallel_read.rs

@@ -12,6 +12,22 @@ use crate::error::FormatError;
 use crate::filter_pipeline::FilterPipeline;
 use crate::filters::decompress_chunk_exact;
 use crate::lane_partition::{self, LaneStats, PartitionStats};
+use crate::storage::{ExtentReq, Storage, for_each_extent_batch};
+
+/// The extents of `chunks`' stored bytes (see
+/// [`crate::chunked_read::chunk_req`]), fetched batch by batch with
+/// [`for_each_extent_batch`] when the file is not in memory (each chunk's
+/// bounds error is reported when that chunk is decoded, as before).
+fn chunk_reqs(
+    chunks: &[ChunkInfo],
+    pipeline: Option<&FilterPipeline>,
+    chunk_total_bytes: usize,
+) -> Vec<ExtentReq> {
+    chunks
+        .iter()
+        .map(|c| crate::chunked_read::chunk_req(c, pipeline, chunk_total_bytes, true))
+        .collect()
+}
 
 /// Threshold: only use parallel decompression when chunk count exceeds this.
 const PARALLEL_THRESHOLD: usize = 4;
@@ -41,6 +57,132 @@ pub fn pool_can_parallelise() -> bool {
     rayon::current_num_threads() > 1
 }
 
+/// How many rayon workers [`run_with_helpers`] should ask to help with
+/// `items` work items, given that the calling thread works too: the pool's
+/// other threads (all of them when the caller is not one), at most one per
+/// item beyond the caller's first.
+pub(crate) fn helper_count(items: usize) -> usize {
+    let pool = rayon::current_num_threads();
+    // A one-thread pool means "decode on the calling thread" (the setting
+    // benchmarks use to compare with h5py, where each call decodes on its
+    // caller): no helper, so one read never uses two cores.
+    if pool <= 1 {
+        return 0;
+    }
+    let others = if rayon::current_thread_index().is_some() {
+        pool.saturating_sub(1)
+    } else {
+        pool
+    };
+    others.min(items.saturating_sub(1))
+}
+
+/// Run `body` on the calling thread and on up to `helpers` rayon workers at
+/// once, returning when the caller's call has finished and every worker that
+/// started one has too. `body` shares its work out itself (typically by
+/// claiming items from an atomic counter until none are left).
+///
+/// The caller never waits for a worker to *become* free: helpers are queued
+/// on the pool, and one that only gets to run after the caller has finished
+/// returns without calling `body`. So a busy or small pool can only fail to
+/// speed a read up, never hold it back — with `par_iter`, the calling thread
+/// (not a pool worker) handed all the work to the pool and slept, and N
+/// threads reading through a 2-worker pool decoded on 2 cores.
+///
+/// A panic in `body`, on any thread, is resumed on the caller once every
+/// helper that started has stopped.
+pub(crate) fn run_with_helpers(helpers: usize, body: &(dyn Fn() + Sync)) {
+    use std::panic::{AssertUnwindSafe, catch_unwind, resume_unwind};
+    use std::sync::{Arc, Condvar, Mutex, PoisonError};
+
+    if helpers == 0 {
+        body();
+        return;
+    }
+
+    type Body = dyn Fn() + Sync + 'static;
+    struct Shared {
+        /// `body`, its lifetime erased. Only dereferenced by a helper that
+        /// registered in `state` while it was open (see below).
+        body: *const Body,
+        /// (closed, helpers inside `body`).
+        state: Mutex<(bool, usize)>,
+        idle: Condvar,
+        panic: Mutex<Option<Box<dyn core::any::Any + Send>>>,
+    }
+    // SAFETY: `body` points to a `Sync` closure, so calling it from other
+    // threads is allowed; the pointer is only used under the protocol below,
+    // which keeps it from outliving the closure.
+    unsafe impl Send for Shared {}
+    unsafe impl Sync for Shared {}
+
+    fn help(shared: &Shared) {
+        {
+            let mut state = shared.state.lock().unwrap_or_else(PoisonError::into_inner);
+            if state.0 {
+                return;
+            }
+            state.1 += 1;
+        }
+        // SAFETY: registered while open, so the caller of `run_with_helpers`
+        // is still inside it (it closes, then waits until no helper is
+        // registered, before returning), and `body` is alive.
+        let body = unsafe { &*shared.body };
+        if let Err(payload) = catch_unwind(AssertUnwindSafe(body)) {
+            shared
+                .panic
+                .lock()
+                .unwrap_or_else(PoisonError::into_inner)
+                .get_or_insert(payload);
+        }
+        let mut state = shared.state.lock().unwrap_or_else(PoisonError::into_inner);
+        state.1 -= 1;
+        if state.1 == 0 {
+            shared.idle.notify_all();
+        }
+    }
+
+    let body_ptr: *const (dyn Fn() + Sync + '_) = body;
+    // SAFETY: only the lifetime changes (same fat-pointer layout). The
+    // pointer is dereferenced only while this function is running: see
+    // `help` and the wait below.
+    let body_ptr: *const Body = unsafe { core::mem::transmute(body_ptr) };
+    let shared = Arc::new(Shared {
+        body: body_ptr,
+        state: Mutex::new((false, 0)),
+        idle: Condvar::new(),
+        panic: Mutex::new(None),
+    });
+    for _ in 0..helpers {
+        let shared = Arc::clone(&shared);
+        rayon::spawn(move || help(&shared));
+    }
+    let caller = catch_unwind(AssertUnwindSafe(body));
+    {
+        // Close, then wait for the helpers inside `body`; later ones return
+        // at once. This must happen even if `body` panicked on this thread.
+        let mut state = shared.state.lock().unwrap_or_else(PoisonError::into_inner);
+        state.0 = true;
+        while state.1 > 0 {
+            state = shared
+                .idle
+                .wait(state)
+                .unwrap_or_else(PoisonError::into_inner);
+        }
+    }
+    if let Err(payload) = caller {
+        resume_unwind(payload);
+    }
+    let helper_panic = shared
+        .panic
+        .lock()
+        .unwrap_or_else(PoisonError::into_inner)
+        .take();
+    if let Some(payload) = helper_panic {
+        resume_unwind(payload);
+    }
+}
+
 /// Decompress chunks in parallel using lane-partitioned assignment.
 ///
 /// Instead of naive `par_iter`, chunks are deterministically assigned to lanes
@@ -63,6 +205,27 @@ pub fn decompress_chunks_lane_partitioned(
     element_size: u32,
     seed: u64,
     num_lanes: Option<usize>,
+) -> Result<(Vec<Vec<u8>>, PartitionStats), FormatError> {
+    decompress_chunks_lane_partitioned_in(
+        file_data,
+        chunks,
+        pipeline,
+        chunk_total_bytes,
+        element_size,
+        seed,
+        num_lanes,
+    )
+}
+
+/// [`decompress_chunks_lane_partitioned`] over any [`Storage`].
+pub fn decompress_chunks_lane_partitioned_in<S: Storage + ?Sized>(
+    file_data: &S,
+    chunks: &[ChunkInfo],
+    pipeline: &FilterPipeline,
+    chunk_total_bytes: usize,
+    element_size: u32,
+    seed: u64,
+    num_lanes: Option<usize>,
 ) -> Result<(Vec<Vec<u8>>, PartitionStats), FormatError> {
     use rayon::prelude::*;
 
@@ -72,31 +235,28 @@ pub fn decompress_chunks_lane_partitioned(
             .unwrap_or(1)
     });
 
-    let assignments = lane_partition::partition_chunks(chunks.len(), lanes, seed);
+    let reqs = chunk_reqs(chunks, Some(pipeline), chunk_total_bytes);
-    let num_lanes = assignments.len();
+    let mut ordered: Vec<Vec<u8>> = Vec::with_capacity(chunks.len());
-
+    let mut partition_stats = PartitionStats::new(0);
+    partition_stats.total_chunks = chunks.len();
+    // Each batch of fetched chunks is partitioned into lanes and decoded
+    // before the next batch is fetched (with the file in memory there is
+    // one batch: all the chunks).
+    for_each_extent_batch(file_data, &reqs, |batch, raw_bytes| {
+        let assignments = lane_partition::partition_chunks(batch.len(), lanes, seed);
         // Each lane processes its assigned chunks and returns results + stats.
-    let lane_results: Result<Vec<(Vec<DecompressedChunk>, LaneStats)>, FormatError> = assignments
+        let lane_results: Result<Vec<(Vec<DecompressedChunk>, LaneStats)>, FormatError> =
+            assignments
                 .into_par_iter()
                 .map(|indices| {
                     let mut results = Vec::with_capacity(indices.len());
                     let mut stats = LaneStats::default();
 
-            for &index in &indices {
+                    for &local in &indices {
+                        let index = batch.start + local;
                         let chunk_info = &chunks[index];
-                let c_addr = chunk_info.address as usize;
                         let size = chunk_info.chunk_size as usize;
-
+                        let raw_chunk = raw_bytes.get(index, &reqs[index])?;
-                if c_addr
-                    .checked_add(size)
-                    .is_none_or(|end| end > file_data.len())
-                {
-                    return Err(FormatError::UnexpectedEof {
-                        expected: c_addr.saturating_add(size),
-                        available: file_data.len(),
-                    });
-                }
-                let raw_chunk = &file_data[c_addr..c_addr + size];
 
                         let decompressed = decompress_chunk_exact(
                             raw_chunk,
@@ -120,14 +280,19 @@ pub fn decompress_chunks_lane_partitioned(
                     Ok((results, stats))
                 })
                 .collect();
-
         let lane_results = lane_results?;
 
         // Aggregate stats
-    let mut partition_stats = PartitionStats::new(num_lanes);
+        if partition_stats.per_lane.len() < lane_results.len() {
-    partition_stats.total_chunks = chunks.len();
+            partition_stats
-    for (lane_idx, (_, stats)) in lane_results.iter().enumerate() {
+                .per_lane
-        partition_stats.per_lane[lane_idx] = stats.clone();
+                .resize_with(lane_results.len(), LaneStats::default);
+            partition_stats.num_lanes = lane_results.len();
+        }
+        for (lane, (_, stats)) in partition_stats.per_lane.iter_mut().zip(&lane_results) {
+            lane.chunks_processed += stats.chunks_processed;
+            lane.compressed_bytes += stats.compressed_bytes;
+            lane.decompressed_bytes += stats.decompressed_bytes;
         }
 
         // Flatten and sort by original index to restore order
@@ -136,8 +301,9 @@ pub fn decompress_chunks_lane_partitioned(
             .flat_map(|(chunks, _)| chunks)
             .collect();
         all_chunks.sort_by_key(|dc| dc.index);
-
+        ordered.extend(all_chunks.into_iter().map(|dc| dc.data));
-    let ordered = all_chunks.into_iter().map(|dc| dc.data).collect();
+        Ok(())
+    })?;
     Ok((ordered, partition_stats))
 }
 
@@ -155,25 +321,29 @@ pub fn decompress_chunks_parallel(
     pipeline: &FilterPipeline,
     chunk_total_bytes: usize,
     element_size: u32,
+) -> Result<Vec<Vec<u8>>, FormatError> {
+    decompress_chunks_parallel_in(file_data, chunks, pipeline, chunk_total_bytes, element_size)
+}
+
+/// [`decompress_chunks_parallel`] over any [`Storage`].
+pub fn decompress_chunks_parallel_in<S: Storage + ?Sized>(
+    file_data: &S,
+    chunks: &[ChunkInfo],
+    pipeline: &FilterPipeline,
+    chunk_total_bytes: usize,
+    element_size: u32,
 ) -> Result<Vec<Vec<u8>>, FormatError> {
     use rayon::prelude::*;
 
-    let results: Result<Vec<DecompressedChunk>, FormatError> = chunks
+    let reqs = chunk_reqs(chunks, Some(pipeline), chunk_total_bytes);
-        .par_iter()
+    let mut ordered: Vec<Vec<u8>> = Vec::with_capacity(chunks.len());
-        .enumerate()
+    for_each_extent_batch(file_data, &reqs, |batch, raw_bytes| {
-        .map(|(index, chunk_info)| {
+        let results: Result<Vec<DecompressedChunk>, FormatError> = batch
-            let c_addr = chunk_info.address as usize;
+            .clone()
-            let size = chunk_info.chunk_size as usize;
+            .into_par_iter()
-            if c_addr
+            .map(|index| {
-                .checked_add(size)
+                let chunk_info = &chunks[index];
-                .is_none_or(|end| end > file_data.len())
+                let raw_chunk = raw_bytes.get(index, &reqs[index])?;
-            {
-                return Err(FormatError::UnexpectedEof {
-                    expected: c_addr.saturating_add(size),
-                    available: file_data.len(),
-                });
-            }
-            let raw_chunk = &file_data[c_addr..c_addr + size];
 
                 let decompressed = decompress_chunk_exact(
                     raw_chunk,
@@ -193,7 +363,10 @@ pub fn decompress_chunks_parallel(
 
         let mut result_vec = results?;
         result_vec.sort_by_key(|dc| dc.index);
-    Ok(result_vec.into_iter().map(|dc| dc.data).collect())
+        ordered.extend(result_vec.into_iter().map(|dc| dc.data));
+        Ok(())
+    })?;
+    Ok(ordered)
 }
 
 /// Decompress chunks sequentially (fallback when parallel is not warranted).
@@ -204,20 +377,23 @@ pub fn decompress_chunks_sequential(
     chunk_total_bytes: usize,
     element_size: u32,
 ) -> Result<Vec<Vec<u8>>, FormatError> {
+    decompress_chunks_sequential_in(file_data, chunks, pipeline, chunk_total_bytes, element_size)
+}
+
+/// [`decompress_chunks_sequential`] over any [`Storage`].
+pub fn decompress_chunks_sequential_in<S: Storage + ?Sized>(
+    file_data: &S,
+    chunks: &[ChunkInfo],
+    pipeline: Option<&FilterPipeline>,
+    chunk_total_bytes: usize,
+    element_size: u32,
+) -> Result<Vec<Vec<u8>>, FormatError> {
+    let reqs = chunk_reqs(chunks, pipeline, chunk_total_bytes);
     let mut result = Vec::with_capacity(chunks.len());
-    for chunk_info in chunks {
+    for_each_extent_batch(file_data, &reqs, |batch, raw_bytes| {
-        let c_addr = chunk_info.address as usize;
+        for i in batch {
-        let size = chunk_info.chunk_size as usize;
+            let chunk_info = &chunks[i];
-        if c_addr
+            let raw_chunk = raw_bytes.get(i, &reqs[i])?;
-            .checked_add(size)
-            .is_none_or(|end| end > file_data.len())
-        {
-            return Err(FormatError::UnexpectedEof {
-                expected: c_addr.saturating_add(size),
-                available: file_data.len(),
-            });
-        }
-        let raw_chunk = &file_data[c_addr..c_addr + size];
 
             let decompressed = if let Some(pl) = pipeline {
                 decompress_chunk_exact(
@@ -233,6 +409,8 @@ pub fn decompress_chunks_sequential(
             };
             result.push(decompressed);
         }
+        Ok(())
+    })?;
     Ok(result)
 }
 
@@ -258,6 +436,60 @@ mod tests {
         (file, infos)
     }
 
+    /// Every item is processed exactly once, whatever mix of caller and
+    /// helpers ends up doing it.
+    #[test]
+    fn run_with_helpers_shares_all_work() {
+        use core::sync::atomic::{AtomicUsize, Ordering};
+        for helpers in [0, 1, 3, 16] {
+            let n = 1000;
+            let next = AtomicUsize::new(0);
+            let done: Vec<AtomicUsize> = (0..n).map(|_| AtomicUsize::new(0)).collect();
+            run_with_helpers(helpers, &|| {
+                loop {
+                    let i = next.fetch_add(1, Ordering::Relaxed);
+                    if i >= n {
+                        break;
+                    }
+                    done[i].fetch_add(1, Ordering::Relaxed);
+                }
+            });
+            assert!(done.iter().all(|d| d.load(Ordering::Relaxed) == 1));
+        }
+    }
+
+    /// A panic in the shared body reaches the caller whichever thread it
+    /// happened on, and only after the helpers inside the body have left it
+    /// (they borrow the caller's stack).
+    #[test]
+    fn run_with_helpers_propagates_panics() {
+        use core::sync::atomic::{AtomicUsize, Ordering};
+        use std::panic::{AssertUnwindSafe, catch_unwind};
+        let caller = std::thread::current().id();
+        for panic_on_caller in [true, false] {
+            let inside = AtomicUsize::new(0);
+            let calls = AtomicUsize::new(0);
+            let result = catch_unwind(AssertUnwindSafe(|| {
+                run_with_helpers(4, &|| {
+                    inside.fetch_add(1, Ordering::SeqCst);
+                    calls.fetch_add(1, Ordering::SeqCst);
+                    let on_caller = std::thread::current().id() == caller;
+                    std::thread::sleep(std::time::Duration::from_millis(20));
+                    inside.fetch_sub(1, Ordering::SeqCst);
+                    if on_caller == panic_on_caller {
+                        panic!("boom");
+                    }
+                });
+            }));
+            // A helper may never have run (the pool was slow to start it),
+            // in which case nothing panicked when `panic_on_caller` is false.
+            if panic_on_caller || calls.load(Ordering::SeqCst) > 1 {
+                assert!(result.is_err());
+            }
+            assert_eq!(inside.load(Ordering::SeqCst), 0);
+        }
+    }
+
     /// Every parallel decoder refuses a chunk that decodes short, naming it.
     #[test]
     fn short_decoded_chunk_is_an_error() {

crates/clawhdf5-format/src/partial_read.rs

@@ -18,14 +18,15 @@ use alloc::{format, vec, vec::Vec};
 #[cfg(feature = "std")]
 use std::string as alloc_or_std;
 
-use crate::chunked_read::{alloc_output, checked_byte_len, list_chunks};
+use crate::chunked_read::{alloc_output, checked_byte_len, list_chunks_for_read_in};
 use crate::data_layout::DataLayout;
 use crate::data_read::extract_selection_from_buffer;
 use crate::dataspace::Dataspace;
 use crate::error::FormatError;
 use crate::filter_pipeline::FilterPipeline;
-use crate::filters::{all_filters_skipped, decompress_chunk_exact};
+use crate::filters::{all_filters_skipped, decompress_chunk_exact_with};
 use crate::selection::Selection;
+use crate::storage::{ExtentReq, Storage, for_each_extent_batch};
 
 /// The smallest axis-aligned box containing every selected element, as
 /// `(start, extent)` per dimension. `None` when there is nothing to gain or
@@ -203,7 +204,12 @@ fn copy_overlap(
     };
     let (src_strides, out_strides) = (strides(src_shape), strides(box_extent));
     let last = rank - 1;
-    let run = ((hi[last] - lo[last]) as usize) * elem_size;
+    // Byte offsets into the in-memory buffers; one that does not fit `usize`
+    // (a 32-bit target) is out of both buffers, like one past their ends.
+    let bytes = |elements: u64| usize::try_from(elements).ok()?.checked_mul(elem_size);
+    let Some(run) = bytes(hi[last] - lo[last]) else {
+        return;
+    };
 
     let mut idx = lo.clone();
     loop {
@@ -213,8 +219,12 @@ fn copy_overlap(
         let out_at: u64 = (0..rank)
             .map(|d| (idx[d] - box_start[d]) * out_strides[d])
             .sum();
-        let (s, o) = (src_at as usize * elem_size, out_at as usize * elem_size);
+        if let (Some(s), Some(o)) = (bytes(src_at), bytes(out_at))
-        if let (Some(from), Some(to)) = (src.get(s..s + run), out.get_mut(o..o + run)) {
+            && let (Some(from), Some(to)) = (
+                src.get(s..s.saturating_add(run)),
+                out.get_mut(o..o.saturating_add(run)),
+            )
+        {
             to.copy_from_slice(from);
         }
         // Advance over every dimension but the last.
@@ -247,6 +257,30 @@ pub fn read_selection(
     offset_size: u8,
     length_size: u8,
     selection: &Selection,
+) -> Result<Option<Vec<u8>>, FormatError> {
+    read_selection_in(
+        file_data,
+        layout,
+        dataspace,
+        elem_size,
+        pipeline,
+        offset_size,
+        length_size,
+        selection,
+    )
+}
+
+/// [`read_selection`] over any [`Storage`].
+#[allow(clippy::too_many_arguments)]
+pub fn read_selection_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    elem_size: usize,
+    pipeline: Option<&FilterPipeline>,
+    offset_size: u8,
+    length_size: u8,
+    selection: &Selection,
 ) -> Result<Option<Vec<u8>>, FormatError> {
     let dims = &dataspace.dimensions;
     if dims.is_empty() || elem_size == 0 {
@@ -267,15 +301,34 @@ pub fn read_selection(
         validate(selection, dims)?;
         let base = usize::try_from(*address)
             .map_err(|_| FormatError::Overflow("data address exceeds usize".into()))?;
-        let data = file_data
+        let file_len = crate::storage::len_usize(file_data);
+        let eof = FormatError::UnexpectedEof {
+            expected: base,
+            available: file_len,
+        };
+        if let Some(all) = file_data.as_contiguous() {
+            let data = all
                 .get(base..)
                 .and_then(|d| d.get(..checked_byte_len(total, elem_size).ok()?))
-            .ok_or(FormatError::UnexpectedEof {
+                .ok_or(eof)?;
-                expected: base,
-                available: file_data.len(),
-            })?;
             return crate::gather::gather::<u8>(data, dims, elem_size, selection).map(Some);
         }
+        // Not in memory: the same bounds check, then only the selected runs
+        // are read.
+        let len = checked_byte_len(total, elem_size)
+            .ok()
+            .filter(|&len| base <= file_len && len <= file_len - base)
+            .ok_or(eof)?;
+        return crate::gather::gather_storage(
+            file_data,
+            base as u64,
+            len,
+            dims,
+            elem_size,
+            selection,
+        )
+        .map(Some);
+    }
     let Some((box_start, box_extent)) = bounding_box(selection, dims) else {
         return Ok(None);
     };
@@ -294,52 +347,61 @@ pub fn read_selection(
             btree_address: Some(_),
             ..
         } => {
-            let (chunks, chunk_dims) = list_chunks(
+            let (chunks, chunk_dims) = list_chunks_for_read_in(
                 file_data,
                 layout,
                 dataspace,
                 elem_size,
+                pipeline,
                 offset_size,
                 length_size,
             )?;
             let rank = dims.len();
             let chunk_shape: Vec<u64> = chunk_dims.iter().map(|&d| d as u64).collect();
             let chunk_bytes = crate::chunked_read::checked_chunk_byte_len(&chunk_dims, elem_size)?;
-            for chunk in &chunks {
+            // The chunks overlapping the box, in index order.
+            let wanted: Vec<&crate::chunked_read::ChunkInfo> = chunks
+                .iter()
+                .filter(|chunk| {
                     if chunk.offsets.len() < rank || chunk.address == u64::MAX {
-                    continue;
+                        return false;
                     }
                     let origin = &chunk.offsets[..rank];
-                let overlaps = (0..rank).all(|d| {
+                    (0..rank).all(|d| {
                         origin[d] < box_start[d] + box_extent[d]
                             && origin[d].saturating_add(chunk_shape[d]) > box_start[d]
-                });
+                    })
-                if !overlaps {
+                })
-                    continue;
+                .collect();
-                }
+            // Their stored bytes, batch by batch when the file is not in
-                let at = usize::try_from(chunk.address)
+            // memory; each batch's chunks are decoded into this thread's
-                    .map_err(|_| FormatError::Overflow("chunk address exceeds usize".into()))?;
+            // reusable buffers before the next batch is fetched.
-                let raw = at
+            let reqs: Vec<ExtentReq> = wanted
-                    .checked_add(chunk.chunk_size as usize)
+                .iter()
-                    .and_then(|end| file_data.get(at..end))
+                .map(|c| crate::chunked_read::chunk_req(c, pipeline, chunk_bytes, true))
-                    .ok_or(FormatError::UnexpectedEof {
+                .collect();
-                        expected: at.saturating_add(chunk.chunk_size as usize),
+            for_each_extent_batch(file_data, &reqs, |batch, raw_bytes| {
-                        available: file_data.len(),
+                crate::chunked_read::with_scratch(|scratch| -> Result<(), FormatError> {
+                    for i in batch {
+                        let chunk = wanted[i];
+                        let origin = &chunk.offsets[..rank];
+                        usize::try_from(chunk.address).map_err(|_| {
+                            FormatError::Overflow("chunk address exceeds usize".into())
                         })?;
-                // Mirrors the full-read path: filter-mask bit i set means
+                        let raw = raw_bytes.get(i, &reqs[i])?;
-                // filter i was not applied to this chunk.
+                        // Mirrors the full-read path: filter-mask bit i set
-                let decoded;
+                        // means filter i was not applied to this chunk.
                         let data: &[u8] = match pipeline {
                             Some(pl) if !all_filters_skipped(pl, chunk.filter_mask) => {
-                        decoded = decompress_chunk_exact(
+                                decompress_chunk_exact_with(
                                     raw,
                                     pl,
                                     chunk_bytes,
                                     elem_size as u32,
                                     chunk.filter_mask,
                                     &chunk.offsets[..rank],
-                        )?;
+                                    scratch,
-                        &decoded
+                                )?
                             }
                             _ => raw,
                         };
@@ -353,6 +415,9 @@ pub fn read_selection(
                             elem_size,
                         );
                     }
+                    Ok(())
+                })
+            })?;
         }
         _ => return Ok(None),
     }

crates/clawhdf5-format/src/provenance.rs

@@ -13,7 +13,6 @@ use sha2::{Digest, Sha256};
 
 use crate::attribute::AttributeMessage;
 use crate::data_layout::DataLayout;
-use crate::data_read::read_raw_data;
 use crate::dataspace::Dataspace;
 use crate::datatype::Datatype;
 use crate::error::FormatError;
@@ -128,10 +127,20 @@ pub fn verify_dataset(
     header: &ObjectHeader,
     offset_size: u8,
     length_size: u8,
+) -> Result<VerifyResult, FormatError> {
+    verify_dataset_in(file_data, header, offset_size, length_size)
+}
+
+/// [`verify_dataset`] over any [`Storage`](crate::storage::Storage).
+pub fn verify_dataset_in<S: crate::storage::Storage + ?Sized>(
+    file_data: &S,
+    header: &ObjectHeader,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<VerifyResult, FormatError> {
     // 1. Extract all attributes (compact + dense).
     let attrs =
-        crate::attribute::extract_attributes_full(file_data, header, offset_size, length_size)?;
+        crate::attribute::extract_attributes_full_in(file_data, header, offset_size, length_size)?;
 
     // 2. Find the stored hash.
     let stored_hash = attrs
@@ -174,7 +183,7 @@ pub fn verify_dataset(
         .transpose()?;
 
     let raw = match &dl {
-        DataLayout::Chunked { .. } => crate::chunked_read::read_chunked_data(
+        DataLayout::Chunked { .. } => crate::chunked_read::read_chunked_data_in(
             file_data,
             &dl,
             &ds,
@@ -183,7 +192,7 @@ pub fn verify_dataset(
             offset_size,
             length_size,
         )?,
-        _ => read_raw_data(file_data, &dl, &ds, &dt)?,
+        _ => crate::data_read::read_raw_data_in(file_data, &dl, &ds, &dt)?,
     };
 
     // 4. Compare.

crates/clawhdf5-format/src/selection.rs

@@ -19,6 +19,7 @@ use alloc::{vec, vec::Vec};
 
 use core::ops::Range;
 
+use crate::addr::to_usize;
 use crate::error::FormatError;
 
 /// A selection describing which elements of a dataset to access.
@@ -562,7 +563,7 @@ fn decode_hyperslab(r: &mut SelReader, version: u64) -> Result<SerializedSelecti
     if !matches!(enc_size, 2 | 4 | 8) {
         return Err(sel_err("unsupported hyperslab coordinate encoding size"));
     }
-    let rank = r.uint(4)? as usize;
+    let rank = to_usize(r.uint(4)?)?;
     // HDF5 caps dataspace rank at 32 (H5S_MAX_RANK). Reject anything else so a
     // corrupt rank can't drive a huge allocation or read loop.
     if rank == 0 || rank > 32 {
@@ -625,11 +626,11 @@ fn decode_hyperslab(r: &mut SelReader, version: u64) -> Result<SerializedSelecti
         return Err(FormatError::UnexpectedEof {
             expected: r
                 .pos
-                .saturating_add(nblocks.saturating_mul(per_block) as usize),
+                .saturating_add(to_usize(nblocks.saturating_mul(per_block))?),
             available: r.data.len(),
         });
     }
-    let n = nblocks as usize * rank;
+    let n = to_usize(nblocks)? * rank;
     let (mut starts, mut ends) = (Vec::with_capacity(n), Vec::with_capacity(n));
     for _ in 0..nblocks {
         for _ in 0..rank {
@@ -662,7 +663,7 @@ fn blocks_union_coords(
             .filter(|&t| t <= MAX_EXPANDED_POINTS)
             .ok_or_else(|| sel_err("irregular hyperslab selection is too large to expand"))?;
     }
-    let mut out = Vec::with_capacity(total as usize);
+    let mut out = Vec::with_capacity(to_usize(total)?);
     for (s, e) in starts.chunks_exact(rank).zip(ends.chunks_exact(rank)) {
         let mut cur = s.to_vec();
         'block: loop {

crates/clawhdf5-format/src/shared_message.rs

@@ -23,11 +23,12 @@ use alloc::vec::Vec;
 #[cfg(feature = "std")]
 use std::borrow::Cow;
 
-use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records};
+use crate::btree_v2::{BTreeV2Header, collect_btree_v2_records_in};
 use crate::error::FormatError;
 use crate::fractal_heap::FractalHeapHeader;
 use crate::message_type::MessageType;
 use crate::object_header::ObjectHeader;
+use crate::storage::{Storage, Window, read_exact_at};
 
 /// Fractal heap ID length for SOHM entries (fixed at 8 bytes).
 const FHEAP_ID_LEN: usize = 8;
@@ -253,17 +254,31 @@ pub fn parse_sohm_table(
     nindexes: u8,
     offset_size: u8,
 ) -> Result<SohmTable, FormatError> {
-    ensure_len(file_data, table_addr, 4)?;
+    parse_sohm_table_in(file_data, table_addr as u64, nindexes, offset_size)
-    if &file_data[table_addr..table_addr + 4] != b"SMTB" {
+}
+
+/// [`parse_sohm_table`] over any [`Storage`]: one read of the signature,
+/// one of every index entry.
+pub fn parse_sohm_table_in<S: Storage + ?Sized>(
+    file: &S,
+    table_addr: u64,
+    nindexes: u8,
+    offset_size: u8,
+) -> Result<SohmTable, FormatError> {
+    let sig = read_exact_at(file, table_addr, 4)?;
+    if *sig != *b"SMTB" {
         return Err(FormatError::InvalidSohmTableSignature);
     }
-    let mut pos = table_addr + 4;
     let os = offset_size as usize;
     let entry_size = 1 + 1 + 2 + 4 + 2 + 2 + 2 + os + os; // 14 + 2*offset_size
+    // Positions below are relative to the table.
+    let w = Window::read(file, table_addr, 4 + nindexes as usize * entry_size)?;
+    let file_data: &[u8] = &w.bytes;
+    let mut pos = 4;
 
     let mut indexes = Vec::with_capacity(nindexes as usize);
     for _ in 0..nindexes {
-        ensure_len(file_data, pos, entry_size)?;
+        w.ensure(pos, entry_size)?;
         let version = file_data[pos];
         if version != 0 {
             return Err(FormatError::InvalidSohmTableVersion(version));
@@ -369,16 +384,29 @@ pub fn parse_sohm_list(
     num_messages: u16,
     offset_size: u8,
 ) -> Result<Vec<SohmEntry>, FormatError> {
-    ensure_len(file_data, list_addr, 4)?;
+    parse_sohm_list_in(file_data, list_addr as u64, num_messages, offset_size)
-    if &file_data[list_addr..list_addr + 4] != b"SMLI" {
+}
+
+/// [`parse_sohm_list`] over any [`Storage`]: one read of the signature, one
+/// of every entry.
+pub fn parse_sohm_list_in<S: Storage + ?Sized>(
+    file: &S,
+    list_addr: u64,
+    num_messages: u16,
+    offset_size: u8,
+) -> Result<Vec<SohmEntry>, FormatError> {
+    let sig = read_exact_at(file, list_addr, 4)?;
+    if *sig != *b"SMLI" {
         return Err(FormatError::InvalidSohmListSignature);
     }
     let entry_sz = sohm_entry_size(offset_size);
-    let mut pos = list_addr + 4;
+    // Positions below are relative to the list.
+    let w = Window::read(file, list_addr, 4 + num_messages as usize * entry_sz)?;
+    let mut pos = 4;
     let mut entries = Vec::with_capacity(num_messages as usize);
     for _ in 0..num_messages {
-        ensure_len(file_data, pos, entry_sz)?;
+        w.ensure(pos, entry_sz)?;
-        let entry = parse_sohm_entry(&file_data[pos..], offset_size)?;
+        let entry = parse_sohm_entry(&w.bytes[pos..], offset_size)?;
         entries.push(entry);
         pos += entry_sz;
     }
@@ -392,8 +420,18 @@ pub fn parse_sohm_btree_entries(
     offset_size: u8,
     length_size: u8,
 ) -> Result<Vec<SohmEntry>, FormatError> {
-    let header = BTreeV2Header::parse(file_data, btree_addr, offset_size, length_size)?;
+    parse_sohm_btree_entries_in(file_data, btree_addr as u64, offset_size, length_size)
-    let records = collect_btree_v2_records(file_data, &header, offset_size, length_size)?;
+}
+
+/// [`parse_sohm_btree_entries`] over any [`Storage`].
+pub fn parse_sohm_btree_entries_in<S: Storage + ?Sized>(
+    file: &S,
+    btree_addr: u64,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Vec<SohmEntry>, FormatError> {
+    let header = BTreeV2Header::parse_in(file, btree_addr, offset_size, length_size)?;
+    let records = collect_btree_v2_records_in(file, &header, offset_size, length_size)?;
     let mut entries = Vec::with_capacity(records.len());
     for rec in &records {
         let entry = parse_sohm_entry(&rec.data, offset_size)?;
@@ -413,15 +451,24 @@ pub fn load_sohm_table(
     offset_size: u8,
     length_size: u8,
 ) -> Result<Option<SohmTable>, FormatError> {
-    let sig = crate::signature::find_signature(file_data)?;
+    load_sohm_table_in(file_data, offset_size, length_size)
-    let sb = crate::superblock::Superblock::parse(file_data, sig)?;
+}
+
+/// [`load_sohm_table`] over any [`Storage`].
+pub fn load_sohm_table_in<S: Storage + ?Sized>(
+    file_data: &S,
+    offset_size: u8,
+    length_size: u8,
+) -> Result<Option<SohmTable>, FormatError> {
+    let sig = crate::signature::find_signature_in(file_data)?;
+    let sb = crate::superblock::Superblock::parse_in(file_data, sig)?;
     let Some(ext_addr) = sb
         .superblock_extension_address
         .filter(|&a| !is_undefined(a, offset_size))
     else {
         return Ok(None);
     };
-    let ext = ObjectHeader::parse(file_data, ext_addr as usize, offset_size, length_size)?;
+    let ext = ObjectHeader::parse_in(file_data, ext_addr, offset_size, length_size)?;
     let Some(msg) = ext
         .messages
         .iter()
@@ -430,9 +477,9 @@ pub fn load_sohm_table(
         return Ok(None);
     };
     let table_msg = parse_sohm_table_message(&msg.data, offset_size)?;
-    parse_sohm_table(
+    parse_sohm_table_in(
         file_data,
-        table_msg.table_address as usize,
+        table_msg.table_address,
         table_msg.nindexes,
         offset_size,
     )
@@ -446,17 +493,27 @@ pub fn message_data_with_sohm<'a>(
     msg: &'a crate::object_header::HeaderMessage,
     offset_size: u8,
     length_size: u8,
+) -> Result<Cow<'a, [u8]>, FormatError> {
+    message_data_with_sohm_in(file_data, msg, offset_size, length_size)
+}
+
+/// [`message_data_with_sohm`] over any [`Storage`].
+pub fn message_data_with_sohm_in<'a, S: Storage + ?Sized>(
+    file_data: &S,
+    msg: &'a crate::object_header::HeaderMessage,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Cow<'a, [u8]>, FormatError> {
     if !is_shared(msg.flags) {
         return Ok(Cow::Borrowed(&msg.data));
     }
     let shared_ref = parse_shared_ref_sized(&msg.data, offset_size, length_size)?;
     let table = if shared_ref.heap_id.is_some() {
-        load_sohm_table(file_data, offset_size, length_size)?
+        load_sohm_table_in(file_data, offset_size, length_size)?
     } else {
         None
     };
-    resolve_shared_message_with_sohm(
+    resolve_shared_message_with_sohm_in(
         file_data,
         &shared_ref,
         msg.msg_type,
@@ -495,6 +552,25 @@ pub fn resolve_sohm_message(
     target_msg_type: MessageType,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<u8>, FormatError> {
+    resolve_sohm_message_in(
+        &file_data,
+        heap_id,
+        sohm_table,
+        target_msg_type,
+        offset_size,
+        length_size,
+    )
+}
+
+/// [`resolve_sohm_message`] over any [`Storage`].
+pub fn resolve_sohm_message_in<S: Storage + ?Sized>(
+    file_data: &S,
+    heap_id: &[u8; FHEAP_ID_LEN],
+    sohm_table: &SohmTable,
+    target_msg_type: MessageType,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<u8>, FormatError> {
     let index = find_index_for_msg_type(sohm_table, target_msg_type)
         .ok_or(FormatError::InvalidSharedMessageVersion(2))?;
@@ -503,13 +579,9 @@ pub fn resolve_sohm_message(
         return Err(FormatError::InvalidSharedMessageVersion(2));
     }
 
-    let fh_header = FractalHeapHeader::parse(
+    let fh_header =
-        file_data,
+        FractalHeapHeader::parse_in(file_data, index.heap_addr, offset_size, length_size)?;
-        index.heap_addr as usize,
+    fh_header.read_managed_object_in(file_data, heap_id, offset_size)
-        offset_size,
-        length_size,
-    )?;
-    fh_header.read_managed_object(file_data, heap_id, offset_size)
 }
 
 /// The payload of an object-header message, following the indirection if the
@@ -526,12 +598,22 @@ pub fn message_data<'a>(
     msg: &'a crate::object_header::HeaderMessage,
     offset_size: u8,
     length_size: u8,
+) -> Result<Cow<'a, [u8]>, FormatError> {
+    message_data_in(file_data, msg, offset_size, length_size)
+}
+
+/// [`message_data`] over any [`Storage`].
+pub fn message_data_in<'a, S: Storage + ?Sized>(
+    file_data: &S,
+    msg: &'a crate::object_header::HeaderMessage,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Cow<'a, [u8]>, FormatError> {
     if !is_shared(msg.flags) {
         return Ok(Cow::Borrowed(&msg.data));
     }
     let shared_ref = parse_shared_ref_sized(&msg.data, offset_size, length_size)?;
-    resolve_shared_message(
+    resolve_shared_message_in(
         file_data,
         &shared_ref,
         msg.msg_type,
@@ -553,13 +635,30 @@ pub fn resolve_shared_message(
     target_msg_type: MessageType,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<u8>, FormatError> {
+    resolve_shared_message_in(
+        &file_data,
+        shared_ref,
+        target_msg_type,
+        offset_size,
+        length_size,
+    )
+}
+
+/// [`resolve_shared_message`] over any [`Storage`].
+pub fn resolve_shared_message_in<S: Storage + ?Sized>(
+    file_data: &S,
+    shared_ref: &SharedMessageRef,
+    target_msg_type: MessageType,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<u8>, FormatError> {
     let table = if shared_ref.heap_id.is_some() {
-        load_sohm_table(file_data, offset_size, length_size)?
+        load_sohm_table_in(file_data, offset_size, length_size)?
     } else {
         None
     };
-    resolve_shared_message_with_sohm(
+    resolve_shared_message_with_sohm_in(
         file_data,
         shared_ref,
         target_msg_type,
@@ -577,6 +676,25 @@ pub fn resolve_shared_message_with_sohm(
     offset_size: u8,
     length_size: u8,
     sohm_table: Option<&SohmTable>,
+) -> Result<Vec<u8>, FormatError> {
+    resolve_shared_message_with_sohm_in(
+        &file_data,
+        shared_ref,
+        target_msg_type,
+        offset_size,
+        length_size,
+        sohm_table,
+    )
+}
+
+/// [`resolve_shared_message_with_sohm`] over any [`Storage`].
+pub fn resolve_shared_message_with_sohm_in<S: Storage + ?Sized>(
+    file_data: &S,
+    shared_ref: &SharedMessageRef,
+    target_msg_type: MessageType,
+    offset_size: u8,
+    length_size: u8,
+    sohm_table: Option<&SohmTable>,
 ) -> Result<Vec<u8>, FormatError> {
     // Dispatch on what the reference carries rather than on `ref_type`: v1/v2
     // references are always an object-header address whatever their type
@@ -586,8 +704,7 @@ pub fn resolve_shared_message_with_sohm(
         shared_ref.heap_id.as_ref(),
     ) {
         (Some(addr), _) => {
-            let target_header =
+            let target_header = ObjectHeader::parse_in(file_data, addr, offset_size, length_size)?;
-                ObjectHeader::parse(file_data, addr as usize, offset_size, length_size)?;
             for msg in &target_header.messages {
                 if msg.msg_type == target_msg_type && !is_shared(msg.flags) {
                     return Ok(msg.data.clone());
@@ -614,7 +731,7 @@ pub fn resolve_shared_message_with_sohm(
         }
         (None, Some(heap_id)) => {
             let table = sohm_table.ok_or(FormatError::InvalidSharedMessageVersion(2))?;
-            resolve_sohm_message(
+            resolve_sohm_message_in(
                 file_data,
                 heap_id,
                 table,
@@ -1052,4 +1169,82 @@ mod tests {
         // With 2-byte offsets: OH=2+2=4, heap=12, entry=1+4+12=17
         assert_eq!(sohm_entry_size(2), 17);
     }
+
+    /// SOHM tables and lists parse identically through a read_at-only
+    /// CountingStorage: at two offsets, with 4- and 8-byte offsets, cut at
+    /// every length and with a bad signature.
+    #[test]
+    fn storage_reads_match_slice_reads() {
+        use crate::storage::CountingStorage;
+        let idx = |t: u8, n: u16| SohmIndex {
+            index_type: t,
+            mesg_types: 0x0008,
+            min_mesg_size: 50,
+            list_max: 50,
+            btree_min: 40,
+            num_messages: n,
+            index_addr: 0x3000,
+            heap_addr: 0x4000,
+        };
+        let heap_entry = |h: u32| SohmEntry {
+            location: 0,
+            hash: h,
+            heap_id: Some([1, 2, 3, 4, 5, 6, 7, h as u8]),
+            ref_count: Some(h),
+            mesg_index: None,
+            oh_addr: None,
+        };
+        let oh_entry = SohmEntry {
+            location: 1,
+            hash: 9,
+            heap_id: None,
+            ref_count: None,
+            mesg_index: Some(3),
+            oh_addr: Some(0x7000),
+        };
+        let mut compared = 0;
+        for os in [4u8, 8] {
+            let smtb = build_smtb(&[idx(0, 2), idx(1, 7)], os);
+            let smli = build_smli(&[heap_entry(1), oh_entry.clone(), heap_entry(2)], os);
+            for (body, n) in [(smtb, 2u16), (smli, 3)] {
+                let is_table = &body[..4] == b"SMTB";
+                for at in [0usize, 0x40] {
+                    let mut full = vec![0u8; at];
+                    full.extend_from_slice(&body);
+                    let mut files = Vec::new();
+                    for cut in at..=full.len() {
+                        files.push(full[..cut].to_vec());
+                    }
+                    let mut bad = full.clone();
+                    bad[at] = b'X';
+                    files.push(bad);
+                    for f in files {
+                        let st = CountingStorage::new(f.clone());
+                        let (want, got) = if is_table {
+                            (
+                                format!("{:?}", parse_sohm_table(&f, at, n as u8, os)),
+                                format!("{:?}", parse_sohm_table_in(&st, at as u64, n as u8, os)),
+                            )
+                        } else {
+                            (
+                                format!("{:?}", parse_sohm_list(&f, at, n, os)),
+                                format!("{:?}", parse_sohm_list_in(&st, at as u64, n, os)),
+                            )
+                        };
+                        assert_eq!(got, want, "{} bytes", f.len());
+                        assert!(st.reads() <= 2);
+                        compared += 1;
+                    }
+                }
+            }
+        }
+        assert!(compared > 200);
+        // The B-tree index reads through Storage too, errors included.
+        let junk = vec![0u8; 64];
+        let st = CountingStorage::new(junk.clone());
+        assert_eq!(
+            format!("{:?}", parse_sohm_btree_entries_in(&st, 0, 8, 8)),
+            format!("{:?}", parse_sohm_btree_entries(&junk, 0, 8, 8))
+        );
+    }
 }

crates/clawhdf5-format/src/signature.rs

@@ -1,6 +1,7 @@
 //! HDF5 file signature (magic bytes) detection.
 
 use crate::error::FormatError;
+use crate::storage::{Storage, read_exact_at};
 
 /// The 8-byte HDF5 magic signature.
 pub const HDF5_SIGNATURE: [u8; 8] = [0x89, b'H', b'D', b'F', b'\r', b'\n', 0x1A, b'\n'];
@@ -39,6 +40,20 @@ pub fn find_signature(data: &[u8]) -> Result<usize, FormatError> {
     Err(FormatError::SignatureNotFound)
 }
 
+/// [`find_signature`] over any [`Storage`]: one 8-byte read per candidate
+/// offset.
+pub fn find_signature_in<S: Storage + ?Sized>(file: &S) -> Result<u64, FormatError> {
+    let len = file.len();
+    let mut offset = 0u64;
+    while offset.checked_add(8).is_some_and(|end| end <= len) {
+        if *read_exact_at(file, offset, 8)? == HDF5_SIGNATURE {
+            return Ok(offset);
+        }
+        offset = if offset == 0 { 512 } else { offset * 2 };
+    }
+    Err(FormatError::SignatureNotFound)
+}
+
 /// Split a file into its user block and its HDF5 bytes.
 ///
 /// Returns `(user_block, hdf5)`: `user_block` is everything before the
@@ -132,4 +147,26 @@ mod tests {
         data[512..520].copy_from_slice(&HDF5_SIGNATURE);
         assert_eq!(find_signature(&data), Ok(0));
     }
+
+    #[test]
+    fn find_signature_in_matches_slice_search() {
+        use crate::storage::CountingStorage;
+        for (len, at) in [
+            (0, None),
+            (7, None),
+            (8, Some(0)),
+            (600, Some(512)),
+            (5000, Some(4096)),
+            (3000, Some(2048)),
+            (3000, None),
+        ] {
+            let mut data = vec![0u8; len];
+            if let Some(at) = at {
+                data[at..at + 8].copy_from_slice(&HDF5_SIGNATURE);
+            }
+            let want = find_signature(&data).map(|o| o as u64);
+            let got = find_signature_in(&CountingStorage::new(data));
+            assert_eq!(got, want, "{len} {at:?}");
+        }
+    }
 }

crates/clawhdf5-format/src/storage.rs

@@ -0,0 +1,723 @@
+//! Where the parsers read the file from: the [`Storage`] trait.
+//!
+//! Every parser used to take the whole file as one `&[u8]`. [`Storage`] is
+//! the abstraction that replaces it (see `docs/design/range-reads.md`,
+//! option (a)): a parser asks for the bytes it needs, `[offset, offset +
+//! len)`, with 64-bit offsets, and gets them back as a [`Cow`] — borrowed
+//! when the backend holds the file in memory (a `Vec`, an mmap), owned when
+//! it had to fetch them (a range request, a block cache).
+//!
+//! `impl Storage for [u8]` serves the in-memory case with no copy, and
+//! [`Storage::as_contiguous`] lets a hot loop borrow the whole file at once
+//! when the backend has it. Modules are converted one at a time: a converted
+//! parser has an `*_in<S: Storage + ?Sized>(file: &S, ..)` core and keeps
+//! its old `&[u8]` signature as a thin wrapper, so callers do not change.
+//!
+//! The cores are generic rather than taking `&dyn Storage` so that the
+//! wrappers monomorphise for `[u8]`: the bounds check of each structure read
+//! inlines to what the slice code did, with no indirect call and no copy,
+//! which keeps local files as fast as before the migration. A `&dyn Storage`
+//! still works (`S = dyn Storage`), and a remote backend pays one indirect
+//! call per structure read.
+//!
+//! # Slice entry points
+//!
+//! A generic core is instantiated in the crate that calls it, so a
+//! downstream crate calling `parse_in::<[u8]>` gets its own copy of the
+//! parser, compiled without this crate's private helpers inlined (there is
+//! no cross-crate inlining of non-`#[inline]` functions without LTO): a
+//! metadata walk through the facade ran about 6% slower that way than
+//! through the `&[u8]` wrappers. The `*_in` entry points on the facade's hot
+//! paths (object headers, group listing and lookup, attributes) therefore
+//! check [`Storage::as_contiguous`] first and hand an in-memory file to
+//! their non-generic `&[u8]` wrapper, compiled here; both run the one core.
+//!
+//! The trait is synchronous and `no_std`: parsing is CPU work, and a remote
+//! backend bridges to its own I/O.
+
+#[cfg(not(feature = "std"))]
+use alloc::{borrow::Cow, boxed::Box, vec::Vec};
+#[cfg(feature = "std")]
+use std::{borrow::Cow, boxed::Box, vec::Vec};
+
+use core::ops::Range;
+
+use crate::error::FormatError;
+
+/// A random-access source of file bytes.
+///
+/// Offsets are relative to the start of the HDF5 data (the superblock), like
+/// every address in the file.
+pub trait Storage {
+    /// Bytes `[offset, offset + len)`.
+    ///
+    /// The result is shorter than `len` only when the range runs past the
+    /// end of the storage (and empty when `offset` is at or past the end);
+    /// a backend that cannot serve a range returns an error instead of a
+    /// short read.
+    /// It is never longer than `len`; the parsers cut a longer result to
+    /// `len` (see [`exact_len`]) rather than read bytes from outside the
+    /// range.
+    fn read_at(&self, offset: u64, len: usize) -> Result<Cow<'_, [u8]>, FormatError>;
+
+    /// Current length of the storage in bytes.
+    fn len(&self) -> u64;
+
+    /// Whether the storage holds no bytes.
+    fn is_empty(&self) -> bool {
+        self.len() == 0
+    }
+
+    /// Several reads at once, in the order given. Backends that talk to a
+    /// remote store coalesce and parallelise these; the default reads them
+    /// one by one with [`Storage::read_at`].
+    fn read_ranges(&self, ranges: &[Range<u64>]) -> Result<Vec<Cow<'_, [u8]>>, FormatError> {
+        ranges
+            .iter()
+            .map(|r| {
+                let len = usize::try_from(r.end.saturating_sub(r.start)).map_err(|_| {
+                    FormatError::Overflow("read range longer than the address space".into())
+                })?;
+                self.read_at(r.start, len)
+            })
+            .collect()
+    }
+
+    /// The whole storage as one slice, when the backend has it in memory
+    /// (a `Vec`, an mmap). Hot loops use this to keep their zero-copy path;
+    /// `None` means every byte has to go through [`Storage::read_at`].
+    fn as_contiguous(&self) -> Option<&[u8]> {
+        None
+    }
+}
+
+impl Storage for [u8] {
+    #[inline]
+    fn read_at(&self, offset: u64, len: usize) -> Result<Cow<'_, [u8]>, FormatError> {
+        let n = self.len();
+        let start = usize::try_from(offset).map_or(n, |o| o.min(n));
+        let end = start.saturating_add(len).min(n);
+        Ok(Cow::Borrowed(&self[start..end]))
+    }
+
+    #[inline]
+    fn len(&self) -> u64 {
+        <[u8]>::len(self) as u64
+    }
+
+    #[inline]
+    fn as_contiguous(&self) -> Option<&[u8]> {
+        Some(self)
+    }
+}
+
+impl Storage for Vec<u8> {
+    #[inline]
+    fn read_at(&self, offset: u64, len: usize) -> Result<Cow<'_, [u8]>, FormatError> {
+        self.as_slice().read_at(offset, len)
+    }
+
+    #[inline]
+    fn len(&self) -> u64 {
+        Vec::len(self) as u64
+    }
+
+    #[inline]
+    fn as_contiguous(&self) -> Option<&[u8]> {
+        Some(self.as_slice())
+    }
+}
+
+impl<T: Storage + ?Sized> Storage for &T {
+    #[inline]
+    fn read_at(&self, offset: u64, len: usize) -> Result<Cow<'_, [u8]>, FormatError> {
+        (**self).read_at(offset, len)
+    }
+
+    #[inline]
+    fn len(&self) -> u64 {
+        (**self).len()
+    }
+
+    #[inline]
+    fn read_ranges(&self, ranges: &[Range<u64>]) -> Result<Vec<Cow<'_, [u8]>>, FormatError> {
+        (**self).read_ranges(ranges)
+    }
+
+    #[inline]
+    fn as_contiguous(&self) -> Option<&[u8]> {
+        (**self).as_contiguous()
+    }
+}
+
+impl<T: Storage + ?Sized> Storage for Box<T> {
+    #[inline]
+    fn read_at(&self, offset: u64, len: usize) -> Result<Cow<'_, [u8]>, FormatError> {
+        (**self).read_at(offset, len)
+    }
+
+    #[inline]
+    fn len(&self) -> u64 {
+        (**self).len()
+    }
+
+    #[inline]
+    fn read_ranges(&self, ranges: &[Range<u64>]) -> Result<Vec<Cow<'_, [u8]>>, FormatError> {
+        (**self).read_ranges(ranges)
+    }
+
+    #[inline]
+    fn as_contiguous(&self) -> Option<&[u8]> {
+        (**self).as_contiguous()
+    }
+}
+
+#[cfg(feature = "std")]
+impl<T: Storage + ?Sized> Storage for std::sync::Arc<T> {
+    #[inline]
+    fn read_at(&self, offset: u64, len: usize) -> Result<Cow<'_, [u8]>, FormatError> {
+        (**self).read_at(offset, len)
+    }
+
+    #[inline]
+    fn len(&self) -> u64 {
+        (**self).len()
+    }
+
+    #[inline]
+    fn read_ranges(&self, ranges: &[Range<u64>]) -> Result<Vec<Cow<'_, [u8]>>, FormatError> {
+        (**self).read_ranges(ranges)
+    }
+
+    #[inline]
+    fn as_contiguous(&self) -> Option<&[u8]> {
+        (**self).as_contiguous()
+    }
+}
+
+/// `storage.len()` as the `usize` the parsers' end-of-file errors report
+/// (saturating on targets where the file is larger than the address space).
+#[inline]
+pub(crate) fn len_usize<S: Storage + ?Sized>(file: &S) -> usize {
+    usize::try_from(file.len()).unwrap_or(usize::MAX)
+}
+
+/// Bytes `[offset, offset + len)`, all of them.
+///
+/// A range that runs past the end of the storage is
+/// [`FormatError::UnexpectedEof`] with `expected = offset + len` and
+/// `available = storage length` — the error the `&[u8]` parsers give for
+/// the same bounds check (`offset + len > file_data.len()`).
+#[inline]
+pub fn read_exact_at<S: Storage + ?Sized>(
+    file: &S,
+    offset: u64,
+    len: usize,
+) -> Result<Cow<'_, [u8]>, FormatError> {
+    let eof = || FormatError::UnexpectedEof {
+        expected: usize::try_from(offset)
+            .unwrap_or(usize::MAX)
+            .saturating_add(len),
+        available: len_usize(file),
+    };
+    // In-memory fast path: plain slicing (for `S = [u8]` this inlines to
+    // the slice code's bounds check).
+    if let Some(all) = file.as_contiguous() {
+        return usize::try_from(offset)
+            .ok()
+            .and_then(|start| all.get(start..start.checked_add(len)?))
+            .map(Cow::Borrowed)
+            .ok_or_else(eof);
+    }
+    match offset.checked_add(len as u64) {
+        Some(end) if end <= file.len() => {}
+        _ => return Err(eof()),
+    }
+    // A short read (the storage shrank, or the backend served less inside
+    // the file) is an error: never parse a partial structure.
+    exact_len(file.read_at(offset, len)?, len)
+}
+
+/// `bytes`, the result of asking a [`Storage`] for `len` bytes, as exactly
+/// `len` bytes: a longer result (a backend that broke
+/// [`Storage::read_at`]'s contract) is cut to `len`, so bytes from outside
+/// the range asked for are never parsed or returned; a shorter one is an
+/// error (the storage shrank, or the backend failed), never a partial
+/// structure.
+#[inline]
+pub fn exact_len(bytes: Cow<'_, [u8]>, len: usize) -> Result<Cow<'_, [u8]>, FormatError> {
+    match bytes.len().cmp(&len) {
+        core::cmp::Ordering::Equal => Ok(bytes),
+        core::cmp::Ordering::Less => Err(short_read()),
+        core::cmp::Ordering::Greater => Ok(match bytes {
+            Cow::Borrowed(b) => Cow::Borrowed(&b[..len]),
+            Cow::Owned(mut v) => {
+                v.truncate(len);
+                Cow::Owned(v)
+            }
+        }),
+    }
+}
+
+#[cold]
+#[inline(never)]
+fn short_read() -> FormatError {
+    FormatError::Storage(
+        "short read inside the file (the storage shrank or the backend failed)".into(),
+    )
+}
+
+/// Largest paged data block (fixed or extensible array) read in one piece.
+/// A bigger one is read as its prefix and then page by page, only the pages
+/// in use, so a block whose size fields claim more than the file holds
+/// costs no more than the pages it really has.
+pub(crate) const PAGED_BLOCK_ONE_READ_MAX: usize = 1 << 20;
+
+/// A window of the file: up to `max` bytes read at `base`, fewer only at
+/// the end of the file. Its [`Window::ensure`] reports a bounds failure
+/// exactly as the whole-file check `ensure_len(file_data, base + rel, n)`
+/// did — with the absolute position and the file's length — as long as
+/// every position checked lies within the `max` bytes the window was asked
+/// for: then a position past the window is past the end of the file.
+pub(crate) struct Window<'a> {
+    /// The bytes, from `base` on.
+    pub bytes: Cow<'a, [u8]>,
+    base: usize,
+    file_len: usize,
+}
+
+impl<'a> Window<'a> {
+    /// Read up to `max` bytes at `base`.
+    pub fn read<S: Storage + ?Sized>(
+        file: &'a S,
+        base: u64,
+        max: usize,
+    ) -> Result<Self, FormatError> {
+        Ok(Window {
+            bytes: read_upto(file, base, max)?,
+            base: usize::try_from(base).unwrap_or(usize::MAX),
+            file_len: len_usize(file),
+        })
+    }
+
+    /// A whole in-memory file as one window (base 0).
+    #[cfg(test)]
+    pub fn whole(bytes: &'a [u8]) -> Self {
+        Window {
+            bytes: Cow::Borrowed(bytes),
+            base: 0,
+            file_len: bytes.len(),
+        }
+    }
+
+    /// [`Window::ensure`] for a window at `base` that has not been read:
+    /// whether `[rel, rel + needed)` lies in the file, with the same error.
+    /// Lets a parser whose first step is to check a structure's whole extent
+    /// (a checksum at its end) fail before reading a structure that a
+    /// hostile size field has stretched past the end of the file.
+    pub fn check_extent<S: Storage + ?Sized>(
+        file: &S,
+        base: u64,
+        rel: usize,
+        needed: usize,
+    ) -> Result<(), FormatError> {
+        let base = usize::try_from(base).unwrap_or(usize::MAX);
+        let file_len = len_usize(file);
+        match base.checked_add(rel).and_then(|p| p.checked_add(needed)) {
+            Some(end) if end <= file_len => Ok(()),
+            _ => Err(FormatError::UnexpectedEof {
+                expected: base.saturating_add(rel).saturating_add(needed),
+                available: file_len,
+            }),
+        }
+    }
+
+    /// Check that `[rel, rel + needed)` (relative to `base`) is in the file.
+    #[inline]
+    pub fn ensure(&self, rel: usize, needed: usize) -> Result<(), FormatError> {
+        match rel.checked_add(needed) {
+            Some(end) if end <= self.bytes.len() => Ok(()),
+            _ => Err(FormatError::UnexpectedEof {
+                expected: self.base.saturating_add(rel).saturating_add(needed),
+                available: self.file_len,
+            }),
+        }
+    }
+}
+
+/// Up to `max` bytes from `offset` on: fewer only at the end of the
+/// storage. For structures whose size is only known once their prefix has
+/// been parsed and whose parsers bound-check what they are given.
+#[inline]
+pub fn read_upto<S: Storage + ?Sized>(
+    file: &S,
+    offset: u64,
+    max: usize,
+) -> Result<Cow<'_, [u8]>, FormatError> {
+    if let Some(all) = file.as_contiguous() {
+        let start = usize::try_from(offset).map_or(all.len(), |o| o.min(all.len()));
+        let end = start.saturating_add(max).min(all.len());
+        return Ok(Cow::Borrowed(&all[start..end]));
+    }
+    let avail = file.len().saturating_sub(offset);
+    let len = usize::try_from(avail).map_or(max, |a| a.min(max));
+    exact_len(file.read_at(offset, len)?, len)
+}
+
+/// Most stored bytes fetched by one [`Storage::read_ranges`] call when a
+/// read gathers many extents (a chunked dataset's chunks, a selection's
+/// runs): a larger read is fetched and decoded batch by batch, so a backend
+/// without the file in memory never holds more than this much undecoded
+/// data per read (or one extent, when a single one is larger — and every
+/// chunk's extent is bounded by what the chunk can need, see
+/// [`crate::filters::stored_chunk_limit`]).
+pub const RAW_BATCH_BYTES: usize = 64 << 20;
+
+/// One extent of a raw-data read: `len` bytes stored at `addr`, whose
+/// bounds are checked against the file, of which the first `fetch` bytes
+/// are read (`None`: only checked, not read — its bytes are not needed).
+///
+/// `fetch` below `len` bounds what a crafted size field can make a read
+/// fetch: a chunk never needs more of its stored bytes than its decoded
+/// size allows, however large its index entry says it is.
+#[derive(Debug, Clone, Copy)]
+pub(crate) struct ExtentReq {
+    pub addr: u64,
+    pub len: usize,
+    pub fetch: Option<usize>,
+}
+
+impl ExtentReq {
+    /// How many bytes are read for this extent.
+    #[inline]
+    fn fetch_len(&self) -> usize {
+        self.fetch.map_or(0, |f| f.min(self.len))
+    }
+}
+
+/// One extent's bytes on their own (see [`ExtentReq`]): the whole extent's
+/// bounds checked as [`read_exact_at`] checks them, and its first
+/// `req.fetch` bytes read (none when `fetch` is `None`).
+pub(crate) fn read_extent<'a, S: Storage + ?Sized>(
+    file: &'a S,
+    req: &ExtentReq,
+) -> Result<Cow<'a, [u8]>, FormatError> {
+    let start = usize::try_from(req.addr).unwrap_or(usize::MAX);
+    match start.checked_add(req.len) {
+        Some(end) if end <= len_usize(file) => read_exact_at(file, req.addr, req.fetch_len()),
+        _ => Err(FormatError::UnexpectedEof {
+            expected: start.saturating_add(req.len),
+            available: len_usize(file),
+        }),
+    }
+}
+
+/// The stored bytes of one batch of extents (chunks, contiguous runs),
+/// fetched together: [`Storage::read_ranges`] is called once for the batch,
+/// so a remote backend can coalesce and parallelise the requests. See
+/// [`for_each_extent_batch`], which is how every raw-data read gets them.
+///
+/// With the whole file in memory nothing is fetched: [`Self::get`] slices
+/// it, as the slice readers did. Either way an extent that does not lie in
+/// the file is the error the slice readers gave for it
+/// ([`FormatError::UnexpectedEof`] with its end and the file length, or
+/// [`FormatError::Overflow`] for an address past this platform's `usize`),
+/// reported when that extent is asked for — so a read reports the first
+/// failing extent in its own order, whatever fails after it.
+pub(crate) enum ExtentBytes<'a> {
+    /// The whole file.
+    Contiguous(&'a [u8]),
+    /// Each extent's bytes, or its bounds error; the first is extent
+    /// `base` of the read.
+    Fetched {
+        base: usize,
+        extents: Vec<Extent<'a>>,
+    },
+}
+
+/// One extent of [`ExtentBytes::Fetched`].
+pub(crate) enum Extent<'a> {
+    /// Its bytes.
+    Bytes(Cow<'a, [u8]>),
+    /// In the file, but not fetched (the caller did not want its bytes).
+    NotFetched,
+    /// The error reading it gives.
+    Err(FormatError),
+}
+
+impl<'a> ExtentBytes<'a> {
+    /// Fetch `reqs`, extents `base..base + reqs.len()` of the read: the
+    /// bytes of those wanted, and the bounds check of all of them.
+    pub(crate) fn fetch<S: Storage + ?Sized>(
+        file: &'a S,
+        reqs: &[ExtentReq],
+        base: usize,
+    ) -> Result<Self, FormatError> {
+        if let Some(all) = file.as_contiguous() {
+            return Ok(ExtentBytes::Contiguous(all));
+        }
+        let file_len = len_usize(file);
+        let mut ranges = Vec::new();
+        let mut out = Vec::with_capacity(reqs.len());
+        // Positions in `out` of the extents being read, in `ranges` order.
+        let mut slots = Vec::new();
+        for req in reqs {
+            let checked = crate::addr::to_usize(req.addr).and_then(|start| {
+                match start.checked_add(req.len) {
+                    Some(end) if end <= file_len => Ok(()),
+                    _ => Err(FormatError::UnexpectedEof {
+                        expected: start.saturating_add(req.len),
+                        available: file_len,
+                    }),
+                }
+            });
+            match checked {
+                Ok(()) if req.fetch.is_some() => {
+                    slots.push(out.len());
+                    ranges.push(req.addr..req.addr + req.fetch_len() as u64);
+                    out.push(Extent::NotFetched);
+                }
+                Ok(()) => out.push(Extent::NotFetched),
+                Err(e) => out.push(Extent::Err(e)),
+            }
+        }
+        if !ranges.is_empty() {
+            let got = file.read_ranges(&ranges)?;
+            if got.len() != ranges.len() {
+                return Err(FormatError::Storage(
+                    "read_ranges returned the wrong number of ranges".into(),
+                ));
+            }
+            for ((slot, bytes), r) in slots.into_iter().zip(got).zip(&ranges) {
+                let len = crate::addr::saturating_usize(r.end - r.start);
+                out[slot] = Extent::Bytes(exact_len(bytes, len)?);
+            }
+        }
+        Ok(ExtentBytes::Fetched { base, extents: out })
+    }
+
+    /// Whether extent `i` of the read (`req`) lies in the file: its bounds
+    /// error if not.
+    pub(crate) fn check(&self, i: usize, req: &ExtentReq) -> Result<(), FormatError> {
+        match self {
+            ExtentBytes::Contiguous(_) => self.get(i, req).map(|_| ()),
+            ExtentBytes::Fetched { base, extents } => {
+                match i.checked_sub(*base).and_then(|j| extents.get(j)) {
+                    Some(Extent::Err(e)) => Err(e.clone()),
+                    Some(_) => Ok(()),
+                    None => Err(not_fetched()),
+                }
+            }
+        }
+    }
+
+    /// Extent `i` of the read (`req`): its first `req.fetch` bytes, the
+    /// same whether the file is in memory or not.
+    pub(crate) fn get(&self, i: usize, req: &ExtentReq) -> Result<&[u8], FormatError> {
+        match self {
+            ExtentBytes::Contiguous(all) => {
+                let start = crate::addr::to_usize(req.addr)?;
+                start
+                    .checked_add(req.len)
+                    .and_then(|end| all.get(start..end))
+                    .map(|b| &b[..req.fetch_len()])
+                    .ok_or(FormatError::UnexpectedEof {
+                        expected: start.saturating_add(req.len),
+                        available: <[u8]>::len(all),
+                    })
+            }
+            ExtentBytes::Fetched { base, extents } => {
+                match i.checked_sub(*base).and_then(|j| extents.get(j)) {
+                    Some(Extent::Bytes(b)) => Ok(b),
+                    Some(Extent::Err(e)) => Err(e.clone()),
+                    _ => Err(not_fetched()),
+                }
+            }
+        }
+    }
+}
+
+#[cold]
+fn not_fetched() -> FormatError {
+    FormatError::Storage("an extent that was not fetched was asked for".into())
+}
+
+/// The one way raw-data reads fetch stored bytes: `reqs` are split into
+/// consecutive batches of at most [`RAW_BATCH_BYTES`] of fetched bytes (at
+/// least one extent each — and no extent fetches more than its
+/// [`ExtentReq::fetch`]), and for each batch in turn its bytes are fetched
+/// with one [`Storage::read_ranges`] call and `f(batch, &bytes)` is called,
+/// with `bytes` indexed by the extent's position in `reqs`. A batch's bytes
+/// are dropped before the next batch is fetched, and an error from `f`
+/// stops the read before anything more is fetched.
+///
+/// With the whole file in memory there is nothing to fetch: one call, over
+/// all of `reqs`, that slices the file.
+pub(crate) fn for_each_extent_batch<'a, S: Storage + ?Sized>(
+    file: &'a S,
+    reqs: &[ExtentReq],
+    mut f: impl FnMut(Range<usize>, &ExtentBytes<'a>) -> Result<(), FormatError>,
+) -> Result<(), FormatError> {
+    let contiguous = file.as_contiguous().is_some();
+    for batch in raw_batches(reqs.len(), contiguous, |i| reqs[i].fetch_len()) {
+        let bytes = ExtentBytes::fetch(file, &reqs[batch.clone()], batch.start)?;
+        f(batch, &bytes)?;
+    }
+    Ok(())
+}
+
+/// Split `n` extents, whose sizes `size(i)` gives, into consecutive batches
+/// of at most [`RAW_BATCH_BYTES`] (at least one extent each): the ranges of
+/// `0..n` to fetch together. With the whole file in memory (`contiguous`)
+/// there is nothing to fetch, and one batch.
+pub(crate) fn raw_batches(
+    n: usize,
+    contiguous: bool,
+    size: impl Fn(usize) -> usize,
+) -> Vec<Range<usize>> {
+    if contiguous || n == 0 {
+        return core::iter::once(0..n).collect();
+    }
+    let mut out = Vec::new();
+    let (mut start, mut bytes) = (0, 0usize);
+    for i in 0..n {
+        let s = size(i);
+        if i > start && bytes.saturating_add(s) > RAW_BATCH_BYTES {
+            out.push(start..i);
+            start = i;
+            bytes = 0;
+        }
+        bytes = bytes.saturating_add(s);
+    }
+    out.push(start..n);
+    out
+}
+
+/// Borrow the whole file for a code path that has not been converted to
+/// [`Storage`] yet. On a backend without a contiguous view this is the
+/// clean [`FormatError::ContiguousStorageRequired`] error, never a guess.
+#[inline]
+pub fn require_contiguous<'a, S: Storage + ?Sized>(
+    file: &'a S,
+    what: &'static str,
+) -> Result<&'a [u8], FormatError> {
+    file.as_contiguous()
+        .ok_or(FormatError::ContiguousStorageRequired(what))
+}
+
+/// A [`Storage`] over an in-memory buffer that serves every byte through
+/// [`Storage::read_at`] (its [`Storage::as_contiguous`] is `None`, so no
+/// parser can take the whole-slice shortcut), copies what it serves (as a
+/// remote backend would), and counts the reads and bytes.
+///
+/// It is the equivalence harness of the range-read migration: parsing a
+/// file through it must give exactly what parsing the `&[u8]` gives, and
+/// the counters are the request counts a cacheless range reader would make.
+#[derive(Debug)]
+pub struct CountingStorage {
+    data: Vec<u8>,
+    reads: portable_atomic::AtomicU64,
+    bytes: portable_atomic::AtomicU64,
+}
+
+impl CountingStorage {
+    /// Serve `data` (the file from the superblock on).
+    pub fn new(data: Vec<u8>) -> Self {
+        CountingStorage {
+            data,
+            reads: portable_atomic::AtomicU64::new(0),
+            bytes: portable_atomic::AtomicU64::new(0),
+        }
+    }
+
+    /// Number of `read_at` calls served so far.
+    pub fn reads(&self) -> u64 {
+        self.reads.load(portable_atomic::Ordering::Relaxed)
+    }
+
+    /// Number of bytes served so far.
+    pub fn bytes_read(&self) -> u64 {
+        self.bytes.load(portable_atomic::Ordering::Relaxed)
+    }
+
+    /// Reset both counters.
+    pub fn reset(&self) {
+        self.reads.store(0, portable_atomic::Ordering::Relaxed);
+        self.bytes.store(0, portable_atomic::Ordering::Relaxed);
+    }
+}
+
+impl Storage for CountingStorage {
+    fn read_at(&self, offset: u64, len: usize) -> Result<Cow<'_, [u8]>, FormatError> {
+        let got = self.data.as_slice().read_at(offset, len)?;
+        self.reads.fetch_add(1, portable_atomic::Ordering::Relaxed);
+        self.bytes
+            .fetch_add(got.len() as u64, portable_atomic::Ordering::Relaxed);
+        Ok(Cow::Owned(got.into_owned()))
+    }
+
+    fn len(&self) -> u64 {
+        self.data.len() as u64
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn slice_reads_are_borrowed_and_clamped() {
+        let data: Vec<u8> = (0u8..10).collect();
+        let s: &[u8] = &data;
+        let dynamic: &dyn Storage = &s;
+        assert_eq!(dynamic.len(), 10);
+        let r = dynamic.read_at(2, 3).unwrap();
+        assert!(matches!(r, Cow::Borrowed(_)));
+        assert_eq!(&*r, &[2, 3, 4]);
+        assert_eq!(&*dynamic.read_at(8, 5).unwrap(), &[8, 9]);
+        assert!(dynamic.read_at(10, 5).unwrap().is_empty());
+        assert!(dynamic.read_at(u64::MAX, 5).unwrap().is_empty());
+        assert_eq!(dynamic.as_contiguous(), Some(&data[..]));
+        let v: &dyn Storage = &data;
+        assert_eq!(v.as_contiguous(), Some(&data[..]));
+    }
+
+    #[test]
+    fn read_exact_matches_slice_bounds_errors() {
+        let data = [0u8; 10];
+        let s: &[u8] = &data;
+        assert_eq!(&*read_exact_at(&s, 4, 6).unwrap(), &[0; 6]);
+        assert_eq!(
+            read_exact_at(&s, 4, 7).unwrap_err(),
+            FormatError::UnexpectedEof {
+                expected: 11,
+                available: 10
+            }
+        );
+        assert!(read_exact_at(&s, u64::MAX, 1).is_err());
+        assert_eq!(read_upto(&s, 7, 100).unwrap().len(), 3);
+        assert_eq!(read_upto(&s, 70, 100).unwrap().len(), 0);
+    }
+
+    #[test]
+    fn counting_storage_counts_and_hides_the_slice() {
+        let c = CountingStorage::new((0u8..10).collect());
+        assert!(c.as_contiguous().is_none());
+        let r = c.read_at(3, 4).unwrap();
+        assert!(matches!(r, Cow::Owned(_)));
+        assert_eq!(&*r, &[3, 4, 5, 6]);
+        c.read_at(8, 4).unwrap();
+        assert_eq!((c.reads(), c.bytes_read()), (2, 6));
+        c.reset();
+        assert_eq!((c.reads(), c.bytes_read()), (0, 0));
+    }
+
+    #[test]
+    fn read_ranges_default_loops() {
+        let data: Vec<u8> = (0u8..10).collect();
+        let s: &[u8] = &data;
+        let got = s.read_ranges(&[1..3, 5..9]).unwrap();
+        assert_eq!(&*got[0], &[1, 2]);
+        assert_eq!(&*got[1], &[5, 6, 7, 8]);
+    }
+}

crates/clawhdf5-format/src/superblock.rs

@@ -7,6 +7,11 @@ use byteorder::{ByteOrder, LittleEndian};
 
 use crate::error::FormatError;
 use crate::signature::HDF5_SIGNATURE;
+use crate::storage::{Storage, read_upto};
+
+/// Bytes read to parse a superblock: more than the largest one (version 1
+/// with 8-byte offsets and lengths, 100 bytes).
+const SUPERBLOCK_READ_LEN: usize = 128;
 
 /// Parsed HDF5 superblock (all versions).
 #[derive(Debug, Clone, PartialEq, Eq)]
@@ -161,7 +166,16 @@ impl Superblock {
         file_data: &[u8],
         signature_offset: usize,
     ) -> Result<u64, FormatError> {
-        let refreshed = Superblock::parse(file_data, signature_offset)?;
+        self.refresh_eof_in(file_data, signature_offset as u64)
+    }
+
+    /// [`Self::refresh_eof`] over any [`Storage`].
+    pub fn refresh_eof_in<S: Storage + ?Sized>(
+        &mut self,
+        file: &S,
+        signature_offset: u64,
+    ) -> Result<u64, FormatError> {
+        let refreshed = Superblock::parse_in(file, signature_offset)?;
         self.eof_address = refreshed.eof_address;
         self.consistency_flags = refreshed.consistency_flags;
         Ok(self.eof_address)
@@ -219,15 +233,23 @@ impl Superblock {
     /// [`FormatError::UserBlockNotStripped`] because the addresses in the
     /// returned superblock would otherwise be applied to the wrong bytes.
     pub fn parse(data: &[u8], signature_offset: usize) -> Result<Superblock, FormatError> {
-        if signature_offset != 0 {
+        Self::parse_in(data, signature_offset as u64)
-            return Err(FormatError::UserBlockNotStripped(signature_offset as u64));
     }
-        let d = data
+
-            .get(signature_offset..)
+    /// [`Self::parse`] over any [`Storage`]: one read of the first
-            .ok_or(FormatError::UnexpectedEof {
+    /// [`SUPERBLOCK_READ_LEN`] bytes (fewer when the file is shorter, which
-                expected: signature_offset + 1,
+    /// is then refused with the same end-of-file errors as a short slice).
-                available: data.len(),
+    pub fn parse_in<S: Storage + ?Sized>(
-            })?;
+        file: &S,
+        signature_offset: u64,
+    ) -> Result<Superblock, FormatError> {
+        if signature_offset != 0 {
+            return Err(FormatError::UserBlockNotStripped(signature_offset));
+        }
+        // Every bounds check below needs at most 100 bytes, so on a longer
+        // file none of them can fail and the window's length does not show.
+        let window = read_upto(file, 0, SUPERBLOCK_READ_LEN)?;
+        let d: &[u8] = &window;
         ensure_len(d, 9)?; // signature(8) + version(1)
 
         // Verify signature
@@ -894,4 +916,34 @@ mod tests {
         assert_eq!(parsed.version, 3);
         assert_eq!(parsed.page_size, None);
     }
+
+    /// Through a storage that serves only `read_at`, every version parses
+    /// to the same superblock, and every truncation to the same error, as
+    /// from a slice — in one read.
+    #[test]
+    fn parse_in_matches_slice_parse() {
+        use crate::storage::CountingStorage;
+        let mut files = vec![
+            build_v0_bytes(8),
+            build_v0_bytes(4),
+            build_v1_bytes(8),
+            build_v1_bytes(4),
+            build_v2_bytes(8, 2),
+            build_v2_bytes(4, 3),
+        ];
+        for f in files.clone() {
+            let mut long = f.clone();
+            long.resize(4096, 0xAB);
+            files.push(long);
+            for cut in [0, 5, 9, 13, 20, 30, f.len() - 1] {
+                files.push(f[..cut.min(f.len())].to_vec());
+            }
+        }
+        for f in files {
+            let want = Superblock::parse(&f, 0);
+            let storage = CountingStorage::new(f.clone());
+            assert_eq!(Superblock::parse_in(&storage, 0), want, "{} bytes", f.len());
+            assert_eq!(storage.reads(), 1);
+        }
+    }
 }

crates/clawhdf5-format/src/superblock_ext.rs

@@ -0,0 +1,922 @@
+//! The superblock extension of a version 2 or 3 superblock, and the
+//! metadata cache image it can point to.
+//!
+//! libhdf5 reads the extension when it opens a file (`H5F__super_read`) and
+//! decodes the messages that configure the file: v1 B-tree "K" values, File
+//! Space Info, and the Metadata Cache Image. A message that does not decode
+//! makes the file fail to open, so [`read_superblock_extension`] decodes and
+//! checks them the way libhdf5 does.
+//!
+//! A metadata cache image (written with `H5Pset_mdc_image_config`) is a
+//! block holding serialized metadata cache entries — object headers, B-tree
+//! nodes, heaps — each with its file address. libhdf5 loads it into its
+//! cache before it reads any other metadata (`H5C__load_cache_image`,
+//! `H5C__reconstruct_cache_contents`), and the entries take the place of
+//! the file's bytes at their addresses: the file itself may hold stale or
+//! no metadata there (in `h5clear_mdc_image.h5` the root group's header is
+//! only in the image). [`CacheImage::apply`] does the same with bytes: it
+//! writes every entry at its address, so every parser reads what libhdf5
+//! reads. It writes into whatever the opener gives it — a private
+//! copy-on-write mapping of the file, or a buffer the opener owns — so the
+//! file is never copied whole.
+
+#[cfg(not(feature = "std"))]
+use alloc::{borrow::Cow, collections::BTreeSet, vec::Vec};
+#[cfg(feature = "std")]
+use std::{borrow::Cow, collections::BTreeSet};
+
+use crate::error::FormatError;
+use crate::message_type::MessageType;
+use crate::object_header::ObjectHeader;
+use crate::storage::{Storage, read_exact_at};
+use crate::superblock::Superblock;
+
+/// Message type of the File Space Info message.
+const MSG_FSINFO: u16 = 0x0017;
+/// Message type of the Metadata Cache Image message.
+const MSG_MDCI: u16 = 0x0018;
+/// Header message flag: the library did not know the message when it wrote
+/// it back (`H5O_MSG_FLAG_WAS_UNKNOWN`); libhdf5 then ignores its contents.
+const MSG_FLAG_WAS_UNKNOWN: u8 = 0x20;
+
+/// `H5F_FILE_SPACE_PAGE_SIZE_MIN` / `_MAX`.
+const PAGE_SIZE_MIN: u64 = 512;
+const PAGE_SIZE_MAX: u64 = 1024 * 1024 * 1024;
+/// libhdf5's default file space page size, used for a version 0 message.
+const PAGE_SIZE_DEFAULT: u64 = 4096;
+/// Free-space managers whose addresses a persisting version 1 File Space
+/// Info message lists (`H5F_MEM_PAGE_SUPER` .. `H5F_MEM_PAGE_NTYPES`), and
+/// a version 0 one (`H5FD_MEM_SUPER` .. `H5FD_MEM_NTYPES`).
+const FSM_ADDRS_V1: usize = 12;
+const FSM_ADDRS_V0: usize = 6;
+
+/// Metadata cache image block limits (`H5Cimage.c`, `H5ACprivate.h`).
+const MDCI_SIGNATURE: &[u8; 4] = b"MDCI";
+const MDCI_HAVE_RESIZE_STATUS: u8 = 0x01;
+const MDCI_ENTRY_IS_FD_PARENT: u8 = 0x04;
+const MDCI_ENTRY_IS_FD_CHILD: u8 = 0x08;
+/// `H5AC_NTYPES`: entry type ids are below this.
+const MDCI_NTYPES: u8 = 30;
+/// `H5C_RING_NTYPES`.
+const MDCI_RING_NTYPES: u8 = 6;
+/// `H5AC__CACHE_IMAGE__ENTRY_AGEOUT__MAX`.
+const MDCI_AGE_MAX: u8 = 100;
+
+/// A decoded File Space Info message (0x0017), mapped to version 1 as
+/// libhdf5 maps a version 0 one.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct FileSpaceInfo {
+    /// Message version as stored (0 or 1).
+    pub version: u8,
+    /// File space strategy (`H5F_fspace_strategy_t`).
+    pub strategy: u8,
+    /// Whether free space is persisted.
+    pub persist: bool,
+    /// Free-space section threshold.
+    pub threshold: u64,
+    /// File space page size.
+    pub page_size: u64,
+}
+
+/// Where a metadata cache image block is (Metadata Cache Image message,
+/// 0x0018).
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub struct CacheImageLocation {
+    /// Address of the image block.
+    pub address: u64,
+    /// Length of the image block in bytes.
+    pub length: u64,
+}
+
+/// The messages of a superblock extension that libhdf5 decodes at open.
+#[derive(Debug, Clone, Default, PartialEq, Eq)]
+pub struct SuperblockExtension {
+    /// v1 B-tree "K" values (chunk index, symbol table node, symbol table
+    /// leaf), when the extension overrides the defaults.
+    pub btree_k: Option<(u16, u16, u16)>,
+    /// The File Space Info message.
+    pub file_space_info: Option<FileSpaceInfo>,
+    /// The metadata cache image, when the file has one.
+    pub cache_image: Option<CacheImageLocation>,
+}
+
+fn ext_err(why: &'static str) -> FormatError {
+    FormatError::InvalidSuperblockExtension(why)
+}
+
+const RAN_OFF: &str = "ran off end of input buffer while decoding";
+
+/// A little-endian cursor over one message or block, failing with
+/// `overrun` when it runs off the end.
+struct Cursor<'a> {
+    data: &'a [u8],
+    pos: usize,
+    overrun: FormatError,
+}
+
+impl<'a> Cursor<'a> {
+    fn new(data: &'a [u8], overrun: FormatError) -> Self {
+        Cursor {
+            data,
+            pos: 0,
+            overrun,
+        }
+    }
+
+    fn take(&mut self, n: usize) -> Result<&'a [u8], FormatError> {
+        let end = self
+            .pos
+            .checked_add(n)
+            .filter(|&e| e <= self.data.len())
+            .ok_or_else(|| self.overrun.clone())?;
+        let s = &self.data[self.pos..end];
+        self.pos = end;
+        Ok(s)
+    }
+
+    fn u8(&mut self) -> Result<u8, FormatError> {
+        Ok(self.take(1)?[0])
+    }
+
+    fn uint(&mut self, width: u8) -> Result<u64, FormatError> {
+        let b = self.take(width as usize)?;
+        Ok(b.iter()
+            .rev()
+            .fold(0u64, |acc, &x| (acc << 8) | u64::from(x)))
+    }
+
+    /// An address of `width` bytes; `None` when undefined (all ones).
+    fn addr(&mut self, width: u8) -> Result<Option<u64>, FormatError> {
+        let v = self.uint(width)?;
+        let undef = if width >= 8 {
+            u64::MAX
+        } else {
+            (1u64 << (8 * u32::from(width))) - 1
+        };
+        Ok((v != undef).then_some(v))
+    }
+}
+
+/// Decode and check the superblock extension of `sb`, as libhdf5 does when
+/// it opens the file. `data` is the file from the superblock on, up to the
+/// end of file the superblock records (its end is libhdf5's "eoa").
+///
+/// Returns `Ok(None)` for a superblock without an extension (versions 0
+/// and 1 have none). A message libhdf5 fails to decode, or a cache image
+/// that does not lie inside the file, is an error: libhdf5 refuses to open
+/// such a file (`cve-2020-10810`: a File Space Info message too short for
+/// the free-space manager addresses it announces; `cve-2020-10812`: a cache
+/// image past the end of the file).
+pub fn read_superblock_extension(
+    data: &[u8],
+    sb: &Superblock,
+) -> Result<Option<SuperblockExtension>, FormatError> {
+    read_superblock_extension_in(data, sb)
+}
+
+/// [`read_superblock_extension`] over any [`Storage`]; its length is the
+/// end of file.
+pub fn read_superblock_extension_in<S: Storage + ?Sized>(
+    file: &S,
+    sb: &Superblock,
+) -> Result<Option<SuperblockExtension>, FormatError> {
+    let os = sb.offset_size;
+    let ls = sb.length_size;
+    let undef = if os >= 8 {
+        u64::MAX
+    } else {
+        (1u64 << (8 * u32::from(os))) - 1
+    };
+    let Some(addr) = sb.superblock_extension_address.filter(|&a| a != undef) else {
+        return Ok(None);
+    };
+    let addr = usize::try_from(addr).map_err(|_| ext_err("address out of range"))?;
+    let header = ObjectHeader::parse_in(file, addr as u64, os, ls)?;
+    let eoa = file.len();
+
+    let mut ext = SuperblockExtension::default();
+    for msg in &header.messages {
+        match msg.msg_type {
+            MessageType::BTreeKValues => {
+                let mut c = Cursor::new(&msg.data, ext_err(RAN_OFF));
+                if c.u8()? != 0 {
+                    return Err(ext_err("bad version number for v1 B-tree 'K' message"));
+                }
+                let chunk = c.uint(2)? as u16;
+                let snode = c.uint(2)? as u16;
+                let leaf = c.uint(2)? as u16;
+                ext.btree_k = Some((chunk, snode, leaf));
+            }
+            MessageType::Unknown(MSG_FSINFO) if msg.flags & MSG_FLAG_WAS_UNKNOWN == 0 => {
+                ext.file_space_info = Some(decode_fsinfo(&msg.data, os, ls)?);
+            }
+            MessageType::Unknown(MSG_MDCI) => {
+                let mut c = Cursor::new(&msg.data, ext_err(RAN_OFF));
+                if c.u8()? != 0 {
+                    return Err(ext_err(
+                        "bad version number for metadata cache image message",
+                    ));
+                }
+                let address = c.addr(os)?;
+                let length = c.uint(ls)?;
+                let Some(address) = address else {
+                    return Err(ext_err("metadata cache image address is undefined"));
+                };
+                if address.checked_add(length).is_none_or(|end| end > eoa) {
+                    return Err(ext_err(
+                        "metadata cache image: address plus size exceeds file eoa",
+                    ));
+                }
+                ext.cache_image = Some(CacheImageLocation { address, length });
+            }
+            _ => {}
+        }
+    }
+    Ok(Some(ext))
+}
+
+/// `H5O__fsinfo_decode` plus the checks `H5F__super_read` makes on it.
+fn decode_fsinfo(data: &[u8], os: u8, ls: u8) -> Result<FileSpaceInfo, FormatError> {
+    let mut c = Cursor::new(data, ext_err(RAN_OFF));
+    let version = c.u8()?;
+    let info = if version == 0 {
+        let old_strategy = c.u8()?;
+        let threshold = c.uint(ls)?;
+        // H5F_file_space_type_t: 1 ALL_PERSIST, 2 ALL, 3 AGGR_VFD, 4 VFD.
+        let (strategy, persist) = match old_strategy {
+            1 => {
+                for _ in 0..FSM_ADDRS_V0 {
+                    c.addr(os)?;
+                }
+                (0, true)
+            }
+            2 => (0, false),
+            3 => (2, false),
+            4 => (3, false),
+            _ => return Err(ext_err("invalid file space strategy")),
+        };
+        FileSpaceInfo {
+            version,
+            strategy,
+            persist,
+            threshold,
+            page_size: PAGE_SIZE_DEFAULT,
+        }
+    } else {
+        if version > 1 {
+            return Err(ext_err("File space info message's version out of bounds"));
+        }
+        let strategy = c.u8()?;
+        let persist = c.u8()? != 0;
+        let threshold = c.uint(ls)?;
+        let page_size = c.uint(ls)?;
+        if page_size == 0 || page_size > PAGE_SIZE_MAX {
+            return Err(ext_err("invalid page size in file space info"));
+        }
+        c.uint(2)?; // page end metadata threshold
+        c.addr(os)?; // EOA before the free-space managers
+        if persist {
+            for _ in 0..FSM_ADDRS_V1 {
+                c.addr(os)?;
+            }
+        }
+        FileSpaceInfo {
+            version,
+            strategy,
+            persist,
+            threshold,
+            page_size,
+        }
+    };
+    if info.page_size < PAGE_SIZE_MIN {
+        return Err(ext_err("file space page size too small"));
+    }
+    Ok(info)
+}
+
+/// One entry of a metadata cache image: `len` bytes at `image_offset` in
+/// the image block, belonging at file address `address`.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+struct ImageEntry {
+    address: u64,
+    image_offset: usize,
+    len: usize,
+}
+
+/// A decoded metadata cache image: where its block is, and the entries it
+/// holds. [`CacheImage::apply`] writes the entries over a file's bytes.
+///
+/// Only the entry list is kept, never a copy of the file: an opener that
+/// maps the file applies the image to a private copy-on-write mapping, so
+/// only the pages the entries land on are copied.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct CacheImage {
+    location: CacheImageLocation,
+    entries: Vec<ImageEntry>,
+}
+
+/// What an opener must do about a file's metadata cache image.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub enum CacheImageState {
+    /// The file has no image: its bytes are its metadata.
+    Absent,
+    /// The file has an image that loads: apply it with [`CacheImage::apply`].
+    Loaded(CacheImage),
+    /// The file has an image libhdf5 fails to load. libhdf5 still opens the
+    /// file (the image loads at the first metadata read), and that read
+    /// fails with this error.
+    Unloadable(FormatError),
+}
+
+impl CacheImage {
+    /// Decode the metadata cache image at `location` in `data` (the file
+    /// from the superblock on, up to its recorded end of file). The image is
+    /// checked as libhdf5 checks it (`H5C__decode_cache_image_header`,
+    /// `H5C__reconstruct_cache_entry`): signature and version, the image
+    /// length it records, entry types, rings and ages in range, entry
+    /// addresses inside the file and not repeated, flush-dependency parents
+    /// already in the cache.
+    ///
+    /// One check is stricter than libhdf5's: an entry must end inside the
+    /// file. libhdf5 checks only that it starts there, and serves the rest
+    /// from the image; the images libhdf5 writes never do this (every entry
+    /// lies below the image block, which is written last), and the bytes an
+    /// entry would put past the end of file have nowhere to go in a view of
+    /// the file.
+    ///
+    /// libhdf5 does not verify the block's trailing checksum when it loads
+    /// an image, so neither does this.
+    pub fn decode(
+        data: &[u8],
+        location: CacheImageLocation,
+        sb: &Superblock,
+    ) -> Result<Self, FormatError> {
+        Self::decode_in(data, location, sb)
+    }
+
+    /// [`Self::decode`] over any [`Storage`]: one read of the image block.
+    pub fn decode_in<S: Storage + ?Sized>(
+        file: &S,
+        location: CacheImageLocation,
+        sb: &Superblock,
+    ) -> Result<Self, FormatError> {
+        let (offset_size, length_size) = (sb.offset_size, sb.length_size);
+        let bad = FormatError::InvalidCacheImage;
+        let block = image_block_in(file, location)?;
+        let eoa = file.len();
+        let mut c = Cursor::new(&block, bad(RAN_OFF));
+
+        // Header: signature, version, flags, image data length, entry count.
+        if c.take(4)? != MDCI_SIGNATURE {
+            return Err(bad("bad metadata cache image header signature"));
+        }
+        if c.u8()? != 0 {
+            return Err(bad("bad metadata cache image version"));
+        }
+        if c.u8()? & MDCI_HAVE_RESIZE_STATUS != 0 {
+            return Err(bad("MDC resize status not yet supported"));
+        }
+        if c.uint(length_size)? != location.length {
+            return Err(bad("bad metadata cache image data length"));
+        }
+        let n_entries = c.uint(4)?;
+        if n_entries == 0 {
+            return Err(bad("bad metadata cache entry count"));
+        }
+
+        let mut entries = Vec::new();
+        // What is in libhdf5's cache when it loads the image: the superblock
+        // and the superblock extension's object header (read to find the
+        // image). Each entry's flush-dependency parents are looked up in the
+        // cache as the entry is inserted (`H5C__reconstruct_cache_contents`
+        // searches the index inside the loop that inserts the entries, in
+        // HDF5 1.14.6 and 2.0.0 alike), so a parent must be one of those or
+        // an earlier entry.
+        let mut cached = BTreeSet::new();
+        cached.insert(0);
+        if let Some(ext) = sb.superblock_extension_address {
+            cached.insert(ext);
+        }
+        let mut seen = BTreeSet::new();
+        for _ in 0..n_entries {
+            let type_id = c.u8()?;
+            if type_id >= MDCI_NTYPES {
+                return Err(bad("type id is out of valid range"));
+            }
+            let flags = c.u8()?;
+            if c.u8()? >= MDCI_RING_NTYPES {
+                return Err(bad("ring is out of valid range"));
+            }
+            if c.u8()? > MDCI_AGE_MAX {
+                return Err(bad("entry age is out of policy range"));
+            }
+            let children = c.uint(2)?;
+            // libhdf5 checks the parent flag against the child count only in
+            // debug builds (release builds refuse any entry with children);
+            // the image format's own rule is checked here.
+            if (flags & MDCI_ENTRY_IS_FD_PARENT != 0) != (children > 0) {
+                return Err(bad("flush dependency parent flag and child count disagree"));
+            }
+            c.uint(2)?; // dirty dependency children: reset for a read-only open
+            let parents = c.uint(2)?;
+            if (flags & MDCI_ENTRY_IS_FD_CHILD != 0) != (parents > 0) {
+                return Err(bad("flush dependency child flag and parent count disagree"));
+            }
+            c.uint(4)?; // LRU rank
+            let address = c
+                .addr(offset_size)?
+                .filter(|&a| a < eoa)
+                .ok_or(bad("invalid entry address range"))?;
+            let size = c.uint(length_size)?;
+            if size == 0 {
+                return Err(bad("invalid entry size"));
+            }
+            for _ in 0..parents {
+                let parent = c
+                    .addr(offset_size)?
+                    .ok_or(bad("invalid flush dependency parent offset"))?;
+                if !seen.contains(&parent) && !cached.contains(&parent) {
+                    return Err(bad("fd parent not in cache"));
+                }
+            }
+            let len = usize::try_from(size).map_err(|_| bad(RAN_OFF))?;
+            let image_offset = c.pos;
+            c.take(len)?;
+            if address.checked_add(size).is_none_or(|end| end > eoa) {
+                return Err(bad("entry extends past the end of file"));
+            }
+            if !seen.insert(address) {
+                return Err(bad("duplicate addresses in cache"));
+            }
+            entries.push(ImageEntry {
+                address,
+                image_offset,
+                len,
+            });
+        }
+        Ok(CacheImage { location, entries })
+    }
+
+    /// Where the image block is.
+    pub fn location(&self) -> CacheImageLocation {
+        self.location
+    }
+
+    /// The number of entries in the image.
+    pub fn len(&self) -> usize {
+        self.entries.len()
+    }
+
+    /// Whether the image has no entries (a decoded image always has some).
+    pub fn is_empty(&self) -> bool {
+        self.entries.is_empty()
+    }
+
+    /// The file ranges (address, length) the image's entries replace.
+    pub fn entry_ranges(&self) -> impl Iterator<Item = (u64, usize)> + '_ {
+        self.entries.iter().map(|e| (e.address, e.len))
+    }
+
+    /// The image block in `data`, the bytes [`Self::decode`] read it from.
+    pub fn block<'a>(&self, data: &'a [u8]) -> Result<&'a [u8], FormatError> {
+        image_block(data, self.location)
+    }
+
+    /// [`Self::block`] over any [`Storage`].
+    pub fn block_in<'a, S: Storage + ?Sized>(
+        &self,
+        file: &'a S,
+    ) -> Result<Cow<'a, [u8]>, FormatError> {
+        image_block_in(file, self.location)
+    }
+
+    /// Every entry as `(file address, its bytes)`, taken from `block` (the
+    /// image block, see [`Self::block_in`]), in the order [`Self::apply`]
+    /// writes them: for a reader that cannot write the image over the
+    /// file's bytes and lays the entries over each read instead.
+    pub fn entries<'b>(&self, block: &'b [u8]) -> Result<Vec<(u64, &'b [u8])>, FormatError> {
+        let short = || FormatError::InvalidCacheImage("image applied to the wrong file");
+        self.entries
+            .iter()
+            .map(|e| {
+                let src = block
+                    .get(e.image_offset..e.image_offset + e.len)
+                    .ok_or_else(short)?;
+                Ok((e.address, src))
+            })
+            .collect()
+    }
+
+    /// Write every entry over `dst`, the file's bytes from the superblock
+    /// on (as long as the `data` the image was decoded from), taking the
+    /// entries from `block` (the image block, see [`Self::block`]). `block`
+    /// must not alias `dst`: an entry may land on the block itself.
+    pub fn apply(&self, block: &[u8], dst: &mut [u8]) -> Result<(), FormatError> {
+        let short = || FormatError::InvalidCacheImage("image applied to the wrong file");
+        for e in &self.entries {
+            let src = block
+                .get(e.image_offset..e.image_offset + e.len)
+                .ok_or_else(short)?;
+            let at = usize::try_from(e.address).map_err(|_| short())?;
+            dst.get_mut(at..at + e.len)
+                .ok_or_else(short)?
+                .copy_from_slice(src);
+        }
+        Ok(())
+    }
+}
+
+fn image_block(data: &[u8], location: CacheImageLocation) -> Result<&[u8], FormatError> {
+    let (start, len) = image_block_range(data.len() as u64, location)?;
+    let start = crate::addr::to_usize(start)?;
+    Ok(&data[start..start + len])
+}
+
+fn image_block_in<S: Storage + ?Sized>(
+    file: &S,
+    location: CacheImageLocation,
+) -> Result<Cow<'_, [u8]>, FormatError> {
+    let (start, len) = image_block_range(file.len(), location)?;
+    read_exact_at(file, start, len)
+}
+
+/// Where the image block is, checked against a file of `file_len` bytes.
+fn image_block_range(
+    file_len: u64,
+    location: CacheImageLocation,
+) -> Result<(u64, usize), FormatError> {
+    let bad = FormatError::InvalidCacheImage;
+    let start = usize::try_from(location.address).map_err(|_| bad("address out of range"))?;
+    let len = usize::try_from(location.length).map_err(|_| bad("length out of range"))?;
+    start
+        .checked_add(len)
+        .filter(|&end| end as u64 <= file_len)
+        .ok_or(bad("image block extends past the end of the file"))?;
+    Ok((start as u64, len))
+}
+
+/// What an opener must do before reading a file's metadata: check the
+/// superblock extension ([`read_superblock_extension`]; an error means
+/// libhdf5 refuses to open the file) and decode any metadata cache image
+/// ([`CacheImage::decode`]). `data` is the file from the superblock on, up
+/// to its recorded end of file.
+pub fn cache_image_state(data: &[u8], sb: &Superblock) -> Result<CacheImageState, FormatError> {
+    cache_image_state_in(data, sb)
+}
+
+/// [`cache_image_state`] over any [`Storage`].
+pub fn cache_image_state_in<S: Storage + ?Sized>(
+    file: &S,
+    sb: &Superblock,
+) -> Result<CacheImageState, FormatError> {
+    match read_superblock_extension_in(file, sb)? {
+        Some(SuperblockExtension {
+            cache_image: Some(location),
+            ..
+        }) => Ok(match CacheImage::decode_in(file, location, sb) {
+            Ok(image) => CacheImageState::Loaded(image),
+            Err(e) => CacheImageState::Unloadable(e),
+        }),
+        _ => Ok(CacheImageState::Absent),
+    }
+}
+
+/// [`cache_image_state`] for a reader that holds the file's bytes in a
+/// buffer of its own: check the superblock extension and write any cache
+/// image over `data` in place (only the image block is copied). An image
+/// libhdf5 cannot load is an error here: such a reader has no way to open
+/// the file and fail each object instead.
+pub fn apply_cache_image_in_place(data: &mut [u8], sb: &Superblock) -> Result<(), FormatError> {
+    match cache_image_state(data, sb)? {
+        CacheImageState::Absent => Ok(()),
+        CacheImageState::Unloadable(e) => Err(e),
+        CacheImageState::Loaded(image) => {
+            let block = image.block(data)?.to_vec();
+            image.apply(&block, data)
+        }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    /// The file's bytes with the image at `loc` applied.
+    fn apply_cache_image(
+        data: &[u8],
+        loc: CacheImageLocation,
+        sb: &Superblock,
+    ) -> Result<Vec<u8>, FormatError> {
+        let image = CacheImage::decode(data, loc, sb)?;
+        let mut out = data.to_vec();
+        image.apply(image.block(data)?, &mut out)?;
+        Ok(out)
+    }
+
+    fn sb_v2(ext: u64) -> Superblock {
+        Superblock {
+            version: 2,
+            offset_size: 8,
+            length_size: 8,
+            base_address: 0,
+            eof_address: 0,
+            root_group_address: 0,
+            group_leaf_node_k: None,
+            group_internal_node_k: None,
+            indexed_storage_internal_node_k: None,
+            free_space_address: None,
+            driver_info_address: None,
+            consistency_flags: 0,
+            superblock_extension_address: Some(ext),
+            checksum: None,
+            page_size: None,
+        }
+    }
+
+    /// A file whose superblock extension (a version 1 object header at 48)
+    /// holds the given messages, padded to `len` bytes.
+    fn file_with_ext(messages: &[(u16, &[u8])], len: usize) -> Vec<u8> {
+        let mut body = Vec::new();
+        for &(t, d) in messages {
+            let padded = d.len().div_ceil(8) * 8;
+            body.extend_from_slice(&t.to_le_bytes());
+            body.extend_from_slice(&(padded as u16).to_le_bytes());
+            body.extend_from_slice(&[0x14, 0, 0, 0]);
+            body.extend_from_slice(d);
+            body.resize(body.len() + padded - d.len(), 0);
+        }
+        let mut f = vec![0u8; 48];
+        f.push(1);
+        f.push(0);
+        f.extend_from_slice(&(messages.len() as u16).to_le_bytes());
+        f.extend_from_slice(&1u32.to_le_bytes());
+        f.extend_from_slice(&(body.len() as u32).to_le_bytes());
+        f.extend_from_slice(&[0; 4]);
+        f.extend_from_slice(&body);
+        f.resize(len, 0);
+        f
+    }
+
+    fn fsinfo_v1(page_size: u64, persist: bool, n_addrs: usize) -> Vec<u8> {
+        let mut m = vec![1, 1, u8::from(persist)];
+        m.extend_from_slice(&1u64.to_le_bytes());
+        m.extend_from_slice(&page_size.to_le_bytes());
+        m.extend_from_slice(&0u16.to_le_bytes());
+        m.extend_from_slice(&u64::MAX.to_le_bytes());
+        for _ in 0..n_addrs {
+            m.extend_from_slice(&u64::MAX.to_le_bytes());
+        }
+        m
+    }
+
+    fn mdci(address: u64, length: u64) -> Vec<u8> {
+        let mut m = vec![0];
+        m.extend_from_slice(&address.to_le_bytes());
+        m.extend_from_slice(&length.to_le_bytes());
+        m
+    }
+
+    #[test]
+    fn no_extension() {
+        assert_eq!(
+            read_superblock_extension(&[0; 64], &sb_v2(u64::MAX)).unwrap(),
+            None
+        );
+    }
+
+    #[test]
+    fn file_space_info_as_libhdf5_decodes_it() {
+        // What FileWriter::with_page_size writes.
+        let f = file_with_ext(&[(MSG_FSINFO, &fsinfo_v1(4096, false, 0))], 256);
+        let ext = read_superblock_extension(&f, &sb_v2(48)).unwrap().unwrap();
+        assert_eq!(ext.file_space_info.unwrap().page_size, 4096);
+        let f = file_with_ext(&[(MSG_FSINFO, &fsinfo_v1(4096, true, 12))], 512);
+        assert!(read_superblock_extension(&f, &sb_v2(48)).is_ok());
+
+        let refused = |m: Vec<u8>| {
+            let f = file_with_ext(&[(MSG_FSINFO, &m)], 512);
+            read_superblock_extension(&f, &sb_v2(48)).unwrap_err()
+        };
+        // Persisting, but too short for the manager addresses.
+        let mut short = fsinfo_v1(4096, true, 12);
+        short.truncate(short.len() - 8);
+        assert_eq!(refused(short), ext_err(RAN_OFF));
+        assert!(matches!(
+            refused(fsinfo_v1(256, false, 0)),
+            FormatError::InvalidSuperblockExtension(_)
+        ));
+        assert!(matches!(
+            refused(fsinfo_v1(0, false, 0)),
+            FormatError::InvalidSuperblockExtension(_)
+        ));
+        let mut v2 = fsinfo_v1(4096, false, 0);
+        v2[0] = 2;
+        assert!(matches!(
+            refused(v2),
+            FormatError::InvalidSuperblockExtension(_)
+        ));
+        // cve-2020-10810: version 0, strategy ALL_PERSIST, and a message of
+        // 32 bytes that cannot hold the six addresses that follow.
+        let mut v0 = vec![0u8, 1];
+        v0.extend_from_slice(&[0, 1, 0, 0, 0, 0, 0, 0]);
+        v0.resize(32, 0xff);
+        assert_eq!(refused(v0), ext_err(RAN_OFF));
+        // A version 0 message without persistence is fine.
+        let mut v0 = vec![0u8, 2];
+        v0.extend_from_slice(&[0; 8]);
+        let f = file_with_ext(&[(MSG_FSINFO, &v0)], 256);
+        assert!(read_superblock_extension(&f, &sb_v2(48)).is_ok());
+    }
+
+    #[test]
+    fn cache_image_location_must_be_inside_the_file() {
+        let f = file_with_ext(&[(MSG_MDCI, &mdci(128, 64))], 192);
+        let ext = read_superblock_extension(&f, &sb_v2(48)).unwrap().unwrap();
+        assert_eq!(
+            ext.cache_image,
+            Some(CacheImageLocation {
+                address: 128,
+                length: 64
+            })
+        );
+        // cve-2020-10812: 256 MiB at 0x10100 in a 2565-byte file.
+        let f = file_with_ext(&[(MSG_MDCI, &mdci(0x10100, 0x1000_0000))], 2565);
+        assert!(matches!(
+            read_superblock_extension(&f, &sb_v2(48)),
+            Err(FormatError::InvalidSuperblockExtension(_))
+        ));
+        let f = file_with_ext(&[(MSG_MDCI, &mdci(u64::MAX, 8))], 256);
+        assert!(read_superblock_extension(&f, &sb_v2(48)).is_err());
+    }
+
+    /// A cache image block with `entries` of (address, bytes).
+    fn image(entries: &[(u64, &[u8])]) -> Vec<u8> {
+        let with_deps: Vec<_> = entries.iter().map(|&(a, b)| (a, b, 0, None)).collect();
+        image_with_deps(&with_deps)
+    }
+
+    /// A cache image block with `entries` of (address, bytes, flush
+    /// dependency children, flush dependency parent).
+    fn image_with_deps(entries: &[(u64, &[u8], u16, Option<u64>)]) -> Vec<u8> {
+        let mut b = Vec::new();
+        b.extend_from_slice(MDCI_SIGNATURE);
+        b.push(0);
+        b.push(0);
+        b.extend_from_slice(&0u64.to_le_bytes()); // length, patched below
+        b.extend_from_slice(&(entries.len() as u32).to_le_bytes());
+        for &(addr, bytes, children, parent) in entries {
+            let mut flags = 0x02; // in LRU
+            if children > 0 {
+                flags |= MDCI_ENTRY_IS_FD_PARENT;
+            }
+            if parent.is_some() {
+                flags |= MDCI_ENTRY_IS_FD_CHILD;
+            }
+            b.extend_from_slice(&[5, flags, 1, 0]); // type, flags, ring, age
+            b.extend_from_slice(&children.to_le_bytes());
+            b.extend_from_slice(&0u16.to_le_bytes()); // dirty children
+            b.extend_from_slice(&u16::from(parent.is_some()).to_le_bytes());
+            b.extend_from_slice(&0i32.to_le_bytes());
+            b.extend_from_slice(&addr.to_le_bytes());
+            b.extend_from_slice(&(bytes.len() as u64).to_le_bytes());
+            if let Some(p) = parent {
+                b.extend_from_slice(&p.to_le_bytes());
+            }
+            b.extend_from_slice(bytes);
+        }
+        b.extend_from_slice(&[0; 4]); // checksum (not verified, as in libhdf5)
+        let n = b.len() as u64;
+        b[6..14].copy_from_slice(&n.to_le_bytes());
+        b
+    }
+
+    #[test]
+    fn cache_image_entries_replace_the_file_bytes() {
+        let img = image(&[(16, b"HEADER"), (40, b"NODE")]);
+        let mut f = vec![0u8; 64];
+        let at = f.len() as u64;
+        f.extend_from_slice(&img);
+        let loc = CacheImageLocation {
+            address: at,
+            length: img.len() as u64,
+        };
+        let out = apply_cache_image(&f, loc, &sb_v2(u64::MAX)).unwrap();
+        assert_eq!(out.len(), f.len());
+        assert_eq!(&out[16..22], b"HEADER");
+        assert_eq!(&out[40..44], b"NODE");
+        assert_eq!(&out[..16], &f[..16]);
+
+        let bad = |img: Vec<u8>| {
+            let mut f = vec![0u8; 64];
+            f.extend_from_slice(&img);
+            let loc = CacheImageLocation {
+                address: 64,
+                length: img.len() as u64,
+            };
+            apply_cache_image(&f, loc, &sb_v2(u64::MAX)).unwrap_err()
+        };
+        let mut sig = image(&[(16, b"x")]);
+        sig[0] = b'X';
+        assert!(matches!(bad(sig), FormatError::InvalidCacheImage(_)));
+        assert!(matches!(
+            bad(image(&[(16, b"a"), (16, b"b")])),
+            FormatError::InvalidCacheImage("duplicate addresses in cache")
+        ));
+        assert!(matches!(
+            bad(image(&[(1 << 20, b"far")])),
+            FormatError::InvalidCacheImage("invalid entry address range")
+        ));
+        let mut len = image(&[(16, b"x")]);
+        len[6] ^= 1;
+        assert!(matches!(bad(len), FormatError::InvalidCacheImage(_)));
+        // An entry that starts inside the file (64 bytes, then a 60-byte
+        // image) but runs past its end.
+        assert!(matches!(
+            bad(image(&[(123, b"8 bytes!")])),
+            FormatError::InvalidCacheImage("entry extends past the end of file")
+        ));
+        let mut cut = image(&[(16, b"abcdef")]);
+        let n = cut.len() as u64 - 8;
+        cut.truncate(cut.len() - 8);
+        cut[6..14].copy_from_slice(&n.to_le_bytes());
+        assert!(matches!(bad(cut), FormatError::InvalidCacheImage(_)));
+    }
+
+    /// libhdf5 resolves an entry's flush-dependency parents as it inserts
+    /// the entry (`H5C__reconstruct_cache_contents`): a parent must be an
+    /// earlier entry, or the superblock or its extension's object header,
+    /// which are cached before the image loads. A parent listed after its
+    /// child fails ("fd parent not in cache?!?").
+    #[test]
+    fn flush_dependency_parents_must_already_be_cached() {
+        let load = |img: Vec<u8>| {
+            let mut f = vec![0u8; 64];
+            f.extend_from_slice(&img);
+            let loc = CacheImageLocation {
+                address: 64,
+                length: img.len() as u64,
+            };
+            apply_cache_image(&f, loc, &sb_v2(48))
+        };
+        // Parent first, as libhdf5 writes images.
+        assert!(
+            load(image_with_deps(&[
+                (16, b"P", 1, None),
+                (40, b"C", 0, Some(16))
+            ]))
+            .is_ok()
+        );
+        // Child first: libhdf5 does not find the parent.
+        assert_eq!(
+            load(image_with_deps(&[
+                (40, b"C", 0, Some(16)),
+                (16, b"P", 1, None)
+            ]))
+            .unwrap_err(),
+            FormatError::InvalidCacheImage("fd parent not in cache")
+        );
+        // The superblock extension's header (at 48 here) is in the cache.
+        assert!(load(image_with_deps(&[(40, b"C", 0, Some(48))])).is_ok());
+        // An entry cannot be its own parent.
+        assert!(load(image_with_deps(&[(40, b"C", 1, Some(40))])).is_err());
+    }
+
+    /// The extension and cache image decode identically through a
+    /// `read_at`-only storage, errors included.
+    #[test]
+    fn storage_parse_matches_slice_parse() {
+        use crate::storage::CountingStorage;
+        let img = image(&[(16, b"HEADER"), (40, b"NODE")]);
+        let mut with_image = file_with_ext(&[(MSG_MDCI, &mdci(256, img.len() as u64))], 256);
+        with_image.extend_from_slice(&img);
+        let mut bad_image = with_image.clone();
+        bad_image[256] = b'X';
+        let files = [
+            file_with_ext(&[(MSG_FSINFO, &fsinfo_v1(4096, false, 0))], 256),
+            file_with_ext(&[(MSG_FSINFO, &fsinfo_v1(256, false, 0))], 256),
+            file_with_ext(&[(MSG_MDCI, &mdci(128, 64))], 192),
+            file_with_ext(&[(MSG_MDCI, &mdci(0x10100, 0x1000_0000))], 2565),
+            file_with_ext(&[(MSG_FSINFO, &fsinfo_v1(4096, true, 12))], 60),
+            with_image,
+            bad_image,
+        ];
+        for f in files {
+            let storage = CountingStorage::new(f.clone());
+            let sb = sb_v2(48);
+            assert_eq!(
+                read_superblock_extension_in(&storage, &sb),
+                read_superblock_extension(&f, &sb)
+            );
+            assert_eq!(
+                cache_image_state_in(&storage, &sb),
+                cache_image_state(&f, &sb)
+            );
+            if let Ok(CacheImageState::Loaded(image)) = cache_image_state(&f, &sb) {
+                assert_eq!(
+                    &*image.block_in(&storage).unwrap(),
+                    image.block(&f).unwrap()
+                );
+            }
+        }
+    }
+}

crates/clawhdf5-format/src/symbol_table.rs

@@ -4,6 +4,7 @@
 use alloc::vec::Vec;
 
 use crate::error::FormatError;
+use crate::storage::{Storage, read_exact_at};
 
 /// Symbol Table message (type 0x0011) found in v1 group object headers.
 #[derive(Debug, Clone, PartialEq)]
@@ -79,65 +80,49 @@ impl SymbolTableNode {
         offset: usize,
         offset_size: u8,
     ) -> Result<SymbolTableNode, FormatError> {
-        // signature(4) + version(1) + reserved(1) + number_of_symbols(2) = 8
+        Self::parse_in(file_data, offset as u64, offset_size)
-        if offset
-            .checked_add(8)
-            .is_none_or(|end| end > file_data.len())
-        {
-            return Err(FormatError::UnexpectedEof {
-                expected: offset.saturating_add(8),
-                available: file_data.len(),
-            });
     }
 
-        if &file_data[offset..offset + 4] != b"SNOD" {
+    /// [`Self::parse`] over any [`Storage`]: one read of the node's header,
+    /// one of its entries.
+    pub fn parse_in<S: Storage + ?Sized>(
+        file: &S,
+        offset: u64,
+        offset_size: u8,
+    ) -> Result<SymbolTableNode, FormatError> {
+        // signature(4) + version(1) + reserved(1) + number_of_symbols(2) = 8
+        let header = read_exact_at(file, offset, 8)?;
+
+        if &header[..4] != b"SNOD" {
             return Err(FormatError::InvalidSymbolTableNodeSignature);
         }
 
-        let version = file_data[offset + 4];
+        let version = header[4];
         if version != 1 {
             return Err(FormatError::InvalidSymbolTableNodeVersion(version));
         }
 
-        let num_symbols =
+        let num_symbols = u16::from_le_bytes([header[6], header[7]]) as usize;
-            u16::from_le_bytes([file_data[offset + 6], file_data[offset + 7]]) as usize;
 
         let os = offset_size as usize;
         // Each entry: link_name_offset(os) + obj_hdr_addr(os) + cache_type(4) + reserved(4) + scratch(16)
         let entry_size = os + os + 4 + 4 + 16;
-        let entries_start = offset + 8;
+        // `offset + 8` fits: the header's read checked it. The entries'
-        let needed = entries_start.checked_add(num_symbols * entry_size).ok_or(
+        // read is the bounds check (`offset + 8 + entries > file length`,
-            FormatError::UnexpectedEof {
+        // which cannot overflow: at most 65535 entries of 40 bytes).
-                expected: usize::MAX,
+        let body = read_exact_at(file, offset + 8, num_symbols * entry_size)?;
-                available: file_data.len(),
+        let file_data: &[u8] = &body;
-            },
-        )?;
-        if needed > file_data.len() {
-            return Err(FormatError::UnexpectedEof {
-                expected: needed,
-                available: file_data.len(),
-            });
-        }
 
         let mut entries = Vec::with_capacity(num_symbols);
-        let mut pos = entries_start;
+        for entry in file_data.chunks_exact(entry_size) {
-        for _ in 0..num_symbols {
+            let link_name_offset = read_offset(entry, 0, offset_size)?;
-            let link_name_offset = read_offset(file_data, pos, offset_size)?;
+            let object_header_address = read_offset(entry, os, offset_size)?;
-            pos += os;
+            let pos = 2 * os;
-            let object_header_address = read_offset(file_data, pos, offset_size)?;
+            let cache_type =
-            pos += os;
+                u32::from_le_bytes([entry[pos], entry[pos + 1], entry[pos + 2], entry[pos + 3]]);
-            let cache_type = u32::from_le_bytes([
-                file_data[pos],
-                file_data[pos + 1],
-                file_data[pos + 2],
-                file_data[pos + 3],
-            ]);
-            pos += 4;
             // reserved 4 bytes
-            pos += 4;
             let mut scratch_pad = [0u8; 16];
-            scratch_pad.copy_from_slice(&file_data[pos..pos + 16]);
+            scratch_pad.copy_from_slice(&entry[pos + 8..pos + 24]);
-            pos += 16;
 
             entries.push(SymbolTableEntry {
                 link_name_offset,
@@ -256,4 +241,28 @@ mod tests {
         let result = SymbolTableNode::parse(&data, usize::MAX / 2, 8);
         assert!(result.is_err());
     }
+
+    /// Nodes, cut at every length and at an offset, parse identically
+    /// through a `read_at`-only storage.
+    #[test]
+    fn storage_parse_matches_slice_parse() {
+        use crate::storage::CountingStorage;
+        for os in [4u8, 8] {
+            let node = build_snod(&[(0, 0x100, 0), (8, 0x200, 1), (16, 0x300, 2)], os);
+            let mut bad = node.clone();
+            bad[4] = 2;
+            for full in [node, bad] {
+                for at in [0usize, 7] {
+                    for cut in 0..=full.len() {
+                        let mut f = vec![0u8; at];
+                        f.extend_from_slice(&full[..cut]);
+                        let storage = CountingStorage::new(f.clone());
+                        let want = SymbolTableNode::parse(&f, at, os);
+                        let got = SymbolTableNode::parse_in(&storage, at as u64, os);
+                        assert_eq!(format!("{got:?}"), format!("{want:?}"));
+                    }
+                }
+            }
+        }
+    }
 }

crates/clawhdf5-format/src/type_builders.rs

@@ -296,7 +296,8 @@ impl EnumTypeBuilder {
 
 // ---- Attribute helper ----
 
-pub(crate) fn build_attr_message(name: &str, value: &AttrValue) -> AttributeMessage {
+/// The attribute message the writers store for `value` under `name`.
+pub fn build_attr_message(name: &str, value: &AttrValue) -> AttributeMessage {
     match value {
         AttrValue::F64(v) => AttributeMessage {
             name: name.to_string(),
@@ -503,6 +504,9 @@ pub struct DatasetBuilder {
     /// `data` field is ignored; instead the global heap blob is built from
     /// these mappings and a VDS layout message is emitted.
     pub(crate) virtual_sources: Option<Vec<VdsMapping>>,
+    /// Track (and index) attribute creation order; `None` follows the
+    /// file's default (`FileWriter::track_order`).
+    pub(crate) track_order: Option<bool>,
     #[cfg(feature = "provenance")]
     pub(crate) provenance: Option<ProvenanceConfig>,
 }
@@ -522,11 +526,22 @@ impl DatasetBuilder {
             compact: false,
             alignment: 0,
             virtual_sources: None,
+            track_order: None,
             #[cfg(feature = "provenance")]
             provenance: None,
         }
     }
 
+    /// Track the creation order of this dataset's attributes, and index it,
+    /// as h5py's `create_dataset(..., track_order=True)` does: libhdf5 (and
+    /// h5py) then list the attributes in the order they were set rather
+    /// than by name. libhdf5 numbers at most 65 535 attributes on an object
+    /// that tracks their order; more is an error when the file is written.
+    pub fn track_order(&mut self, track: bool) -> &mut Self {
+        self.track_order = Some(track);
+        self
+    }
+
     pub fn with_f64_data(&mut self, data: &[f64]) -> &mut Self {
         self.datatype = Some(make_f64_type());
         let mut b = Vec::with_capacity(data.len() * 8);
@@ -986,10 +1001,11 @@ impl GroupBuilder {
         self.attrs.push((name.to_string(), value));
     }
 
-    /// Track the creation order of this group's links, and index it, as
+    /// Track the creation order of this group's links and attributes, and
-    /// h5py's `track_order=True` does: libhdf5 (and h5py) then list the
+    /// index it, as h5py's `track_order=True` does: libhdf5 (and h5py) then
-    /// group's members in the order they were added rather than by name.
+    /// list the group's members, and its attributes, in the order they were
-    /// Applies to links only, not to attributes.
+    /// added rather than by name. libhdf5 numbers at most 65 535 attributes
+    /// on an object that tracks their order.
     pub fn track_order(&mut self, track: bool) -> &mut Self {
         self.track_order = Some(track);
         self

crates/clawhdf5-format/src/vds.rs

@@ -15,11 +15,13 @@
 #[cfg(not(feature = "std"))]
 use alloc::{format, string::String, vec, vec::Vec};
 
+use crate::addr::{checked_addr, to_usize};
 use crate::data_layout::{DataLayout, VdsMapping, parse_vds_mappings};
 use crate::dataspace::Dataspace;
 use crate::datatype::Datatype;
 use crate::error::FormatError;
 use crate::selection::{SerializedSelection, UNLIMITED};
+use crate::storage::Storage;
 
 /// Resolves the name of an external VDS source file, as stored in the
 /// mapping, to that file's bytes.
@@ -191,8 +193,8 @@ fn non_unlimited_elements(sel: &SerializedSelection, skip: usize) -> Option<u64>
 }
 
 /// Load and decode the mapping list of a virtual layout.
-fn load_mappings(
+fn load_mappings<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     layout: &DataLayout,
     length_size: u8,
 ) -> Result<Vec<Mapping>, FormatError> {
@@ -207,8 +209,11 @@ fn load_mappings(
     let Some(addr) = *global_heap_address else {
         return Ok(Vec::new());
     };
-    let coll =
+    let coll = crate::global_heap::GlobalHeapCollection::parse_in(
-        crate::global_heap::GlobalHeapCollection::parse(file_data, addr as usize, length_size)?;
+        file_data,
+        checked_addr(addr)?,
+        length_size,
+    )?;
     let index = u16::try_from(*global_heap_index)
         .map_err(|_| vds_err("VDS mapping heap index out of range"))?;
     let obj = coll
@@ -329,7 +334,11 @@ enum Step {
 /// Work out the extent libhdf5 gives the virtual dataset
 /// (`H5D__virtual_set_extent_unlim`, default view `H5D_VDS_LAST_AVAILABLE`
 /// with a printf gap of 0) and how much of each unlimited mapping is read.
-fn plan(mappings: &[Mapping], stored: &[u64], sources: &mut Sources) -> Result<Plan, FormatError> {
+fn plan<S: Storage + ?Sized>(
+    mappings: &[Mapping],
+    stored: &[u64],
+    sources: &mut Sources<'_, '_, S>,
+) -> Result<Plan, FormatError> {
     let overflow = || FormatError::Overflow("VDS extent overflow".into());
     let rank = stored.len();
     let mut new_dims: Vec<Option<u64>> = vec![None; rank];
@@ -471,6 +480,25 @@ pub fn virtual_dataset_extent(
     _offset_size: u8,
     length_size: u8,
     resolver: Option<&VdsFileResolver>,
+) -> Result<Vec<u64>, FormatError> {
+    virtual_dataset_extent_in(
+        file_data,
+        layout,
+        dataspace,
+        _offset_size,
+        length_size,
+        resolver,
+    )
+}
+
+/// [`virtual_dataset_extent`] over any [`Storage`].
+pub fn virtual_dataset_extent_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    _offset_size: u8,
+    length_size: u8,
+    resolver: Option<&VdsFileResolver>,
 ) -> Result<Vec<u64>, FormatError> {
     let mappings = load_mappings(file_data, layout, length_size)?;
     if mappings.iter().all(|m| m.kind == Kind::Fixed) {
@@ -497,6 +525,30 @@ pub fn read_virtual_dataset(
     _offset_size: u8,
     length_size: u8,
     resolver: Option<&VdsFileResolver>,
+) -> Result<VirtualData, FormatError> {
+    read_virtual_dataset_in(
+        file_data,
+        layout,
+        dataspace,
+        datatype,
+        fill,
+        _offset_size,
+        length_size,
+        resolver,
+    )
+}
+
+/// [`read_virtual_dataset`] over any [`Storage`].
+#[allow(clippy::too_many_arguments)]
+pub fn read_virtual_dataset_in<S: Storage + ?Sized>(
+    file_data: &S,
+    layout: &DataLayout,
+    dataspace: &Dataspace,
+    datatype: &Datatype,
+    fill: Option<&[u8]>,
+    _offset_size: u8,
+    length_size: u8,
+    resolver: Option<&VdsFileResolver>,
 ) -> Result<VirtualData, FormatError> {
     let mappings = load_mappings(file_data, layout, length_size)?;
     let mut sources = Sources::new(file_data, resolver);
@@ -510,7 +562,7 @@ pub fn read_virtual_dataset(
     let mut data = crate::chunked_read::alloc_output(crate::chunked_read::checked_byte_len(
         total, elem_size,
     )?)?;
-    if let Some(fill) = fill.filter(|f| f.len() == elem_size && f.iter().any(|&b| b != 0)) {
+    if let Some(fill) = fill.filter(|f| <[u8]>::len(f) == elem_size && f.iter().any(|&b| b != 0)) {
         for element in data.chunks_exact_mut(elem_size) {
             element.copy_from_slice(fill);
         }
@@ -611,12 +663,12 @@ fn scatter(
         return Err(vds_err("virtual/source selection element counts differ"));
     }
     for (&v, &s) in vidx.iter().zip(sidx) {
-        let (vo, so) = (v as usize * elem_size, s as usize * elem_size);
+        let (vo, so) = (to_usize(v)? * elem_size, to_usize(s)? * elem_size);
         if vo + elem_size > out.len() || so + elem_size > src.len() {
             return Err(vds_err("virtual dataset selection out of bounds"));
         }
         out[vo..vo + elem_size].copy_from_slice(&src[so..so + elem_size]);
-        mapped[v as usize] = true;
+        mapped[to_usize(v)?] = true;
     }
     Ok(())
 }
@@ -747,7 +799,7 @@ fn selection_indices(
                     return Err(vds_err("VDS selection blocks overlap"));
                 }
             }
-            let mut out = Vec::with_capacity(volume as usize);
+            let mut out = Vec::with_capacity(to_usize(volume)?);
             for (s, e) in starts.chunks_exact(rank).zip(ends.chunks_exact(rank)) {
                 let mut cur = s.to_vec();
                 'block: loop {
@@ -779,14 +831,17 @@ struct SourceData {
 
 /// Source files and datasets, fetched on demand. The most recently used
 /// external file is kept, since consecutive mappings usually share one.
-struct Sources<'a, 'r> {
+///
-    file_data: &'a [u8],
+/// The virtual dataset's own file (`"."`) is read through its [`Storage`];
+/// an external source file is loaded whole, through the resolver.
+struct Sources<'a, 'r, S: Storage + ?Sized> {
+    file_data: &'a S,
     resolver: Option<&'r VdsFileResolver<'r>>,
     cached_file: Option<(String, Option<Vec<u8>>)>,
 }
 
-impl<'a, 'r> Sources<'a, 'r> {
+impl<'a, 'r, S: Storage + ?Sized> Sources<'a, 'r, S> {
-    fn new(file_data: &'a [u8], resolver: Option<&'r VdsFileResolver<'r>>) -> Self {
+    fn new(file_data: &'a S, resolver: Option<&'r VdsFileResolver<'r>>) -> Self {
         Sources {
             file_data,
             resolver,
@@ -794,16 +849,18 @@ impl<'a, 'r> Sources<'a, 'r> {
         }
     }
 
-    /// The bytes of source file `name`, or `None` if it does not exist.
+    /// The bytes of external source file `name` (not `"."`), or `None` if it
-    fn file(&mut self, name: &str) -> Result<Option<&[u8]>, FormatError> {
+    /// does not exist.
-        if name == "." {
+    fn external(&mut self, name: &str) -> Result<Option<&[u8]>, FormatError> {
-            return Ok(Some(self.file_data));
-        }
         if self.cached_file.as_ref().is_none_or(|(n, _)| n != name) {
             let resolver = self.resolver.ok_or_else(|| {
                 vds_err("external-file virtual dataset sources require a file resolver")
             })?;
-            self.cached_file = Some((String::from(name), resolver(name)?));
+            let mut bytes = resolver(name)?;
+            if let Some(b) = bytes.as_mut() {
+                load_source_file(b)?;
+            }
+            self.cached_file = Some((String::from(name), bytes));
         }
         // An external file is handed over whole; its addresses are relative
         // to its superblock, so skip any user block.
@@ -816,7 +873,10 @@ impl<'a, 'r> Sources<'a, 'r> {
     /// The extent of source dataset `path` in file `file`, or `None` when
     /// either does not exist.
     fn dims(&mut self, file: &str, path: &str) -> Result<Option<Vec<u64>>, FormatError> {
-        let Some(bytes) = self.file(file)? else {
+        if file == "." {
+            return Ok(open_source(self.file_data, path)?.map(|s| s.dataspace.dimensions));
+        }
+        let Some(bytes) = self.external(file)? else {
             return Ok(None);
         };
         Ok(open_source(bytes, path)?.map(|s| s.dataspace.dimensions))
@@ -841,7 +901,13 @@ impl<'a, 'r> Sources<'a, 'r> {
                  from another file is not supported"
             )));
         }
-        let Some(bytes) = self.file(file)? else {
+        if file == "." {
+            let Some(src) = open_source(self.file_data, path)? else {
+                return Ok(None);
+            };
+            return read_source(self.file_data, src, path, datatype).map(Some);
+        }
+        let Some(bytes) = self.external(file)? else {
             return Ok(None);
         };
         let Some(src) = open_source(bytes, path)? else {
@@ -851,6 +917,23 @@ impl<'a, 'r> Sources<'a, 'r> {
     }
 }
 
+/// Check an external source file's superblock extension as libhdf5 does
+/// when it opens the file, and write any metadata cache image over its
+/// metadata in place: libhdf5 reads the image's entries instead of the
+/// file's own, possibly stale, bytes (`crate::superblock_ext`). A source
+/// file whose image cannot be loaded is an error, as other corrupt source
+/// files are here.
+fn load_source_file(whole: &mut [u8]) -> Result<(), FormatError> {
+    let base = crate::signature::find_signature(whole)?;
+    let sb = crate::superblock::Superblock::parse(&whole[base..], 0)?;
+    // The end of file the superblock records; a truncated source file is
+    // read as before, up to its length.
+    let end = sb
+        .data_end(base as u64, whole.len() as u64)
+        .map_or(Ok(whole.len()), |e| to_usize(e).map(|e| base + e))?;
+    crate::superblock_ext::apply_cache_image_in_place(&mut whole[base..end], &sb)
+}
+
 /// Whether elements of `dt` contain addresses into their own file:
 /// variable-length data (global-heap IDs) or references.
 fn holds_file_addresses(dt: &Datatype) -> bool {
@@ -888,19 +971,23 @@ fn source_message<'h>(
 
 /// Open source dataset `path` of the file in `file_data`, or `None` if there
 /// is no such object (libhdf5 reads a missing source as fill).
-fn open_source(file_data: &[u8], path: &str) -> Result<Option<OpenSource>, FormatError> {
+fn open_source<S: Storage + ?Sized>(
+    file_data: &S,
+    path: &str,
+) -> Result<Option<OpenSource>, FormatError> {
     use crate::message_type::MessageType;
-    use crate::shared_message::message_data_with_sohm;
+    use crate::shared_message::message_data_with_sohm_in as message_data_with_sohm;
 
-    // `file_data` starts at the superblock (see `Sources::file`).
+    // `file_data` starts at the superblock (see `Sources::external`).
-    let sb = crate::superblock::Superblock::parse(file_data, 0)?;
+    let sb = crate::superblock::Superblock::parse_in(file_data, 0)?;
     let (os, ls) = (sb.offset_size, sb.length_size);
-    let addr = match crate::group_v2::resolve_path_any(file_data, &sb, path) {
+    let addr = match crate::group_v2::resolve_path_any_in(file_data, &sb, path) {
         Ok(a) => a,
         Err(FormatError::PathNotFound(_)) => return Ok(None),
         Err(e) => return Err(e),
     };
-    let header = crate::object_header::ObjectHeader::parse(file_data, addr as usize, os, ls)?;
+    let header =
+        crate::object_header::ObjectHeader::parse_in(file_data, checked_addr(addr)?, os, ls)?;
     let mut src = OpenSource {
         offset_size: os,
         length_size: ls,
@@ -919,15 +1006,15 @@ fn open_source(file_data: &[u8], path: &str) -> Result<Option<OpenSource>, Forma
 
 /// Read an opened source dataset in full (its own fill value applied to
 /// unallocated chunks).
-fn read_source(
+fn read_source<S: Storage + ?Sized>(
-    file_data: &[u8],
+    file_data: &S,
     src: OpenSource,
     path: &str,
     datatype: &Datatype,
 ) -> Result<SourceData, FormatError> {
     use crate::filter_pipeline::FilterPipeline;
     use crate::message_type::MessageType;
-    use crate::shared_message::message_data_with_sohm;
+    use crate::shared_message::message_data_with_sohm_in as message_data_with_sohm;
 
     let (os, ls) = (src.offset_size, src.length_size);
     let dt_msg = source_message(&src, path, MessageType::Datatype)?;
@@ -965,7 +1052,7 @@ fn read_source(
             message_data_with_sohm(file_data, m, os, ls).and_then(|d| FilterPipeline::parse(&d))
         })
         .transpose()?;
-    let raw = crate::fill_value::read_full_with_fill(
+    let raw = crate::fill_value::read_full_with_fill_in(
         &src.header.messages,
         file_data,
         &layout,
@@ -974,7 +1061,7 @@ fn read_source(
         os,
         ls,
         || {
-            crate::data_read::read_raw_data_full(
+            crate::data_read::read_raw_data_full_in(
                 file_data,
                 &layout,
                 &src.dataspace,

crates/clawhdf5-format/src/vl_data.rs

@@ -5,10 +5,11 @@
 //! `sequence_length(4 LE) + collection_address(offset_size LE) + object_index(4 LE)`.
 
 #[cfg(not(feature = "std"))]
-use alloc::{collections::BTreeMap, format, string::String, vec, vec::Vec};
+use alloc::{borrow::Cow, collections::BTreeMap, format, string::String, vec, vec::Vec};
 #[cfg(feature = "std")]
-use std::collections::BTreeMap;
+use std::{borrow::Cow, collections::BTreeMap};
 
+use crate::addr::to_usize;
 use crate::error::FormatError;
 use crate::global_heap::{GlobalHeapCollection, GlobalHeapIndex};
 
@@ -55,7 +56,7 @@ pub fn parse_vl_references(
 ) -> Result<Vec<VlElement>, FormatError> {
     let elem_size = 4 + offset_size as usize + 4; // length + address + index
     let total =
-        (num_elements as usize)
+        to_usize(num_elements)?
             .checked_mul(elem_size)
             .ok_or(FormatError::UnexpectedEof {
                 expected: usize::MAX,
@@ -68,7 +69,7 @@ pub fn parse_vl_references(
         });
     }
 
-    let mut elements = Vec::with_capacity(num_elements as usize);
+    let mut elements = Vec::with_capacity(to_usize(num_elements)?);
     let mut pos = 0;
 
     for _ in 0..num_elements {
@@ -136,13 +137,15 @@ pub fn check_element_size(stored_size: u32, offset_size: u8) -> Result<(), Forma
 
 /// A collection's objects, located in the file data but not copied:
 /// `(index, offset, size)` of the first object with each index, sorted by
-/// index.
+/// index. Over a storage without the whole file in memory, also the
-struct CachedCollection {
+/// collection's bytes (`(offset, bytes)`), read once when it is indexed.
+struct CachedCollection<'a> {
     objects: Vec<(u16, usize, usize)>,
+    bytes: Option<(usize, Cow<'a, [u8]>)>,
 }
 
-impl CachedCollection {
+impl<'a> CachedCollection<'a> {
-    fn new(index: GlobalHeapIndex) -> Self {
+    fn new(index: GlobalHeapIndex, bytes: Option<(usize, Cow<'a, [u8]>)>) -> Self {
         let mut objects: Vec<(u16, usize, usize)> = index
             .objects
             .iter()
@@ -151,12 +154,16 @@ impl CachedCollection {
         // Stable, so the first object with a repeated index is kept.
         objects.sort_by_key(|o| o.0);
         objects.dedup_by_key(|o| o.0);
-        Self { objects }
+        Self { objects, bytes }
     }
 
     /// What this entry costs to keep, in bytes (roughly).
     fn cost(&self) -> usize {
-        64 + self.objects.len() * core::mem::size_of::<(u16, usize, usize)>()
+        let held = match &self.bytes {
+            Some((_, Cow::Owned(b))) => b.len(),
+            _ => 0,
+        };
+        64 + self.objects.len() * core::mem::size_of::<(u16, usize, usize)>() + held
     }
 
     fn get(&self, index: u32) -> Option<(usize, usize)> {
@@ -169,6 +176,8 @@ impl CachedCollection {
 /// How many bytes of collection indexes a [`VlResolver`] keeps before it
 /// drops them and starts again. Values are never copied into the cache, so
 /// this bounds what a read retains however many collections it visits.
+/// (Over a storage without the whole file in memory the collections' bytes
+/// are kept too, and count against this.)
 const CACHE_BUDGET: usize = 32 << 20;
 
 /// Resolves variable-length elements against a file's global heap, parsing
@@ -184,11 +193,18 @@ const CACHE_BUDGET: usize = 32 << 20;
 /// that overlap one another are refused (libhdf5 never writes them), so a
 /// file cannot make the resolver parse the same bytes as the objects of
 /// many collections.
-pub struct VlResolver<'a> {
+///
-    file_data: &'a [u8],
+/// The file is any [`Storage`](crate::storage::Storage) (`S`, a slice by default). Over one without
+/// the whole file in memory each collection is read once, when first used,
+/// and kept (within the budget above); [`Self::strings`],
+/// [`Self::string_bytes`] and [`Self::sequences`] work over any storage,
+/// [`Self::element`] and [`Self::string_element`], which borrow from the
+/// file, over a slice.
+pub struct VlResolver<'a, S: crate::storage::Storage + ?Sized = [u8]> {
+    file_data: &'a S,
     offset_size: u8,
     length_size: u8,
-    cache: BTreeMap<u64, CachedCollection>,
+    cache: BTreeMap<u64, CachedCollection<'a>>,
     cached_bytes: usize,
     budget: usize,
     /// Start → end of every collection parsed so far (kept when the cache
@@ -200,6 +216,56 @@ impl<'a> VlResolver<'a> {
     /// A resolver over `file_data` (the file from its superblock on), with
     /// the superblock's offset and length sizes.
     pub fn new(file_data: &'a [u8], offset_size: u8, length_size: u8) -> Self {
+        Self::new_in(file_data, offset_size, length_size)
+    }
+
+    /// One element (the first [`element_size`](Self::element_size) bytes of
+    /// `elem`) of a variable-length sequence whose base type is `base_size`
+    /// bytes: its `length × base_size` bytes, or `None` for a null element
+    /// (heap address 0).
+    pub fn element(
+        &mut self,
+        elem: &[u8],
+        base_size: usize,
+    ) -> Result<Option<&'a [u8]>, FormatError> {
+        let vl = parse_vl_references(elem, 1, self.offset_size)?;
+        let vl = &vl[0];
+        if vl.collection_address == 0 {
+            return Ok(None);
+        }
+        let (start, size) = self.locate(vl)?;
+        let data = &self.file_data[start..start + size];
+        check_object_size(vl, data.len(), base_size)?;
+        Ok(Some(data))
+    }
+
+    /// One variable-length string element: its bytes up to the first NUL,
+    /// or `None` for a null element (h5dump prints it as `NULL`, h5py
+    /// returns it as empty).
+    pub fn string_element(&mut self, elem: &[u8]) -> Result<Option<&'a [u8]>, FormatError> {
+        Ok(self.element(elem, 1)?.map(cut_at_nul))
+    }
+}
+
+/// `data_len`, the size of `vl`'s heap object, against the `length ×
+/// base_size` bytes the element says it holds.
+fn check_object_size(vl: &VlElement, data_len: usize, base_size: usize) -> Result<(), FormatError> {
+    let expected = (vl.length as usize)
+        .checked_mul(base_size)
+        .ok_or_else(|| FormatError::Overflow("variable-length element size".into()))?;
+    if data_len != expected {
+        return Err(FormatError::VlDataError(format!(
+            "global heap object {} in the collection at {} holds {data_len} bytes; the element \
+             says {} × {base_size}",
+            vl.object_index, vl.collection_address, vl.length
+        )));
+    }
+    Ok(())
+}
+
+impl<'a, S: crate::storage::Storage + ?Sized> VlResolver<'a, S> {
+    /// [`VlResolver::new`] over any [`Storage`](crate::storage::Storage).
+    pub fn new_in(file_data: &'a S, offset_size: u8, length_size: u8) -> Self {
         Self {
             file_data,
             offset_size,
@@ -230,49 +296,31 @@ impl<'a> VlResolver<'a> {
 
     /// The bytes of one element: `length × base_size` bytes from the heap,
     /// or `None` for a null element.
-    fn resolve(
+    fn resolve(&mut self, vl: &VlElement, base_size: usize) -> Result<Option<&[u8]>, FormatError> {
-        &mut self,
+        if vl.collection_address == 0 {
-        vl: &VlElement,
-        base_size: usize,
-    ) -> Result<Option<&'a [u8]>, FormatError> {
-        let addr = vl.collection_address;
-        if addr == 0 {
             return Ok(None);
         }
         let data = self.object(vl)?;
-        let expected = (vl.length as usize)
+        check_object_size(vl, data.len(), base_size)?;
-            .checked_mul(base_size)
-            .ok_or_else(|| FormatError::Overflow("variable-length element size".into()))?;
-        if data.len() != expected {
-            return Err(FormatError::VlDataError(format!(
-                "global heap object {} in the collection at {addr} holds {} bytes; the element \
-                 says {} × {base_size}",
-                vl.object_index,
-                data.len(),
-                vl.length
-            )));
-        }
         Ok(Some(data))
     }
 
-    /// One element (the first [`element_size`](Self::element_size) bytes of
+    /// [`VlResolver::element`] over any storage: the element's bytes
-    /// `elem`) of a variable-length sequence whose base type is `base_size`
+    /// (borrowed from the resolver's cache of heap collections, so they
-    /// bytes: its `length × base_size` bytes, or `None` for a null element
+    /// live until the next call), or `None` for a null element.
-    /// (heap address 0).
+    pub fn element_in(
-    pub fn element(
         &mut self,
         elem: &[u8],
         base_size: usize,
-    ) -> Result<Option<&'a [u8]>, FormatError> {
+    ) -> Result<Option<&[u8]>, FormatError> {
         let vl = parse_vl_references(elem, 1, self.offset_size)?;
         self.resolve(&vl[0], base_size)
     }
 
-    /// One variable-length string element: its bytes up to the first NUL,
+    /// [`VlResolver::string_element`] over any storage (see
-    /// or `None` for a null element (h5dump prints it as `NULL`, h5py
+    /// [`element_in`](Self::element_in)).
-    /// returns it as empty).
+    pub fn string_element_in(&mut self, elem: &[u8]) -> Result<Option<&[u8]>, FormatError> {
-    pub fn string_element(&mut self, elem: &[u8]) -> Result<Option<&'a [u8]>, FormatError> {
+        Ok(self.element_in(elem, 1)?.map(cut_at_nul))
-        Ok(self.element(elem, 1)?.map(cut_at_nul))
     }
 
     /// The strings of the variable-length string elements in `raw`, as
@@ -329,9 +377,20 @@ pub fn read_vl_strings(
     num_elements: u64,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<String>, FormatError> {
+    read_vl_strings_in(file_data, raw_data, num_elements, offset_size, length_size)
+}
+
+/// [`read_vl_strings`] over any [`Storage`](crate::storage::Storage).
+pub fn read_vl_strings_in<S: crate::storage::Storage + ?Sized>(
+    file_data: &S,
+    raw_data: &[u8],
+    num_elements: u64,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<String>, FormatError> {
     let raw = first_elements(raw_data, num_elements, offset_size)?;
-    VlResolver::new(file_data, offset_size, length_size).strings(raw)
+    VlResolver::new_in(file_data, offset_size, length_size).strings(raw)
 }
 
 /// The first `num_elements` elements of `raw`, or an error if it is shorter.
@@ -363,9 +422,20 @@ pub fn read_vl_bytes(
     num_elements: u64,
     offset_size: u8,
     length_size: u8,
+) -> Result<Vec<Vec<u8>>, FormatError> {
+    read_vl_bytes_in(file_data, raw_data, num_elements, offset_size, length_size)
+}
+
+/// [`read_vl_bytes`] over any [`Storage`](crate::storage::Storage).
+pub fn read_vl_bytes_in<S: crate::storage::Storage + ?Sized>(
+    file_data: &S,
+    raw_data: &[u8],
+    num_elements: u64,
+    offset_size: u8,
+    length_size: u8,
 ) -> Result<Vec<Vec<u8>>, FormatError> {
     let refs = parse_vl_references(raw_data, num_elements, offset_size)?;
-    let mut resolver = VlResolver::new(file_data, offset_size, length_size);
+    let mut resolver = VlResolver::new_in(file_data, offset_size, length_size);
     let mut result = Vec::with_capacity(refs.len());
 
     for vl in &refs {
@@ -384,10 +454,10 @@ pub fn read_vl_bytes(
     Ok(result)
 }
 
-impl<'a> VlResolver<'a> {
+impl<'a, S: crate::storage::Storage + ?Sized> VlResolver<'a, S> {
-    /// The heap object `vl` points to, whatever its size; its collection is
+    /// Where the heap object `vl` points to lies in the file, whatever its
-    /// parsed on first use.
+    /// size (`(offset, size)`); its collection is parsed on first use.
-    fn object(&mut self, vl: &VlElement) -> Result<&'a [u8], FormatError> {
+    fn locate(&mut self, vl: &VlElement) -> Result<(usize, usize), FormatError> {
         let addr = vl.collection_address;
         // libhdf5 writes a null element with address 0, never the undefined
         // address, and fails to read one ("addr undefined") even when its
@@ -401,14 +471,20 @@ impl<'a> VlResolver<'a> {
         if !self.cache.contains_key(&addr) {
             let offset = usize::try_from(addr).map_err(|_| FormatError::UnexpectedEof {
                 expected: usize::MAX,
-                available: self.file_data.len(),
+                available: crate::storage::len_usize(self.file_data),
             })?;
-            let index =
+            let (bytes, base, index) =
-                GlobalHeapCollection::parse_index(self.file_data, offset, self.length_size)?;
+                GlobalHeapCollection::read_collection(self.file_data, addr, self.length_size)?;
-            // parse_index checked that the collection lies in the file.
+            // read_collection checked that the collection lies in the file.
-            let end = offset + index.collection_size as usize;
+            let end = offset + to_usize(index.collection_size)?;
             self.check_overlap(offset, end)?;
-            let coll = CachedCollection::new(index);
+            // With the whole file in memory the objects are sliced from it;
+            // otherwise the collection's bytes are kept.
+            let bytes = match self.file_data.as_contiguous() {
+                Some(_) => None,
+                None => Some((base, bytes)),
+            };
+            let coll = CachedCollection::new(index, bytes);
             if self.cached_bytes.saturating_add(coll.cost()) > self.budget {
                 self.cache.clear();
                 self.cached_bytes = 0;
@@ -416,13 +492,27 @@ impl<'a> VlResolver<'a> {
             self.cached_bytes += coll.cost();
             self.cache.insert(addr, coll);
         }
-        let (start, size) = self.cache[&addr].get(vl.object_index).ok_or(
+        self.cache[&addr]
-            FormatError::GlobalHeapObjectNotFound {
+            .get(vl.object_index)
+            .ok_or(FormatError::GlobalHeapObjectNotFound {
                 collection_address: addr,
                 index: vl.object_index as u16,
-            },
+            })
-        )?;
+    }
-        Ok(&self.file_data[start..start + size])
+
+    /// The heap object `vl` points to, whatever its size; its collection is
+    /// parsed on first use.
+    fn object(&mut self, vl: &VlElement) -> Result<&[u8], FormatError> {
+        let (start, size) = self.locate(vl)?;
+        if let Some(all) = self.file_data.as_contiguous() {
+            return Ok(&all[start..start + size]);
+        }
+        match &self.cache[&vl.collection_address].bytes {
+            Some((base, bytes)) => Ok(&bytes[start - base..start - base + size]),
+            None => Err(FormatError::Storage(
+                "global heap collection bytes were not kept".into(),
+            )),
+        }
     }
 
     /// Record the collection at `start..end`, refusing one that overlaps a
@@ -573,6 +663,32 @@ mod tests {
         }
     }
 
+    #[test]
+    fn element_in_over_a_storage_matches_element_over_a_slice() {
+        let mut file_data = vec![0u8; 512];
+        build_gcol_at(&mut file_data, 256, &[(1, b"Alice\0x"), (2, b"Bob")]);
+        let mut raw = build_vl_refs(&["Alice\0x", "Bob"], 256, 1, 8);
+        raw.extend(element(0, 0, 0, 8)); // null
+        raw.extend(element(9, 256, 1, 8)); // wrong length: an error
+        let storage = crate::storage::CountingStorage::new(file_data.clone());
+        let dynamic: &dyn crate::storage::Storage = &storage;
+        let mut slice = VlResolver::new(&file_data, 8, 8);
+        let mut any = VlResolver::new_in(dynamic, 8, 8);
+        for e in raw.chunks(16) {
+            let want = slice.element(e, 1).map(|o| o.map(<[u8]>::to_vec));
+            let got = any.element_in(e, 1).map(|o| o.map(<[u8]>::to_vec));
+            assert_eq!(format!("{want:?}"), format!("{got:?}"));
+            let want = slice.string_element(e).map(|o| o.map(<[u8]>::to_vec));
+            let got = any.string_element_in(e).map(|o| o.map(<[u8]>::to_vec));
+            assert_eq!(format!("{want:?}"), format!("{got:?}"));
+        }
+        assert_eq!(
+            any.string_element_in(&raw[..16]).unwrap(),
+            Some(&b"Alice"[..])
+        );
+        assert!(storage.reads() > 0);
+    }
+
     #[test]
     fn null_vl_element_zero_address() {
         let mut raw = Vec::new();
@@ -708,6 +824,7 @@ mod tests {
         let mut r = VlResolver::new(&file_data, 8, 8);
         let one = CachedCollection {
             objects: vec![(0, 0, 0); 3],
+            bytes: None,
         }
         .cost();
         r.budget = 2 * one + 1;

crates/clawhdf5-format/tests/blosc2_alloc_bounds.rs

@@ -0,0 +1,641 @@
+//! Crafted Blosc2 frames and chunks cannot make the decoder allocate out of
+//! proportion to the HDF5 chunk it decodes.
+//!
+//! A frame's header, its offsets chunk and its chunk headers all declare
+//! sizes, and the decoder used to allocate what they declared: a 173-byte
+//! frame whose offsets chunk claimed 2 GiB was decoded in full before any
+//! check failed. Every allocation is now bounded by the output limit (the
+//! HDF5 chunk's size) and the input's length.
+//!
+//! Peak heap use is measured with a counting global allocator; the tests
+//! share it, so each holds `SERIAL` for its whole run.
+#![cfg(feature = "blosc2")]
+
+use std::alloc::{GlobalAlloc, Layout, System};
+use std::sync::Mutex;
+use std::sync::atomic::{AtomicUsize, Ordering};
+
+use clawhdf5_format::filters_blosc2::{blosc2_decompress, blosc2_decompress_chunk};
+
+struct Counting;
+
+static CURRENT: AtomicUsize = AtomicUsize::new(0);
+static PEAK: AtomicUsize = AtomicUsize::new(0);
+static SERIAL: Mutex<()> = Mutex::new(());
+
+unsafe impl GlobalAlloc for Counting {
+    unsafe fn alloc(&self, layout: Layout) -> *mut u8 {
+        let p = unsafe { System.alloc(layout) };
+        if !p.is_null() {
+            let now = CURRENT.fetch_add(layout.size(), Ordering::Relaxed) + layout.size();
+            PEAK.fetch_max(now, Ordering::Relaxed);
+        }
+        p
+    }
+
+    unsafe fn alloc_zeroed(&self, layout: Layout) -> *mut u8 {
+        let p = unsafe { System.alloc_zeroed(layout) };
+        if !p.is_null() {
+            let now = CURRENT.fetch_add(layout.size(), Ordering::Relaxed) + layout.size();
+            PEAK.fetch_max(now, Ordering::Relaxed);
+        }
+        p
+    }
+
+    unsafe fn dealloc(&self, ptr: *mut u8, layout: Layout) {
+        unsafe { System.dealloc(ptr, layout) };
+        CURRENT.fetch_sub(layout.size(), Ordering::Relaxed);
+    }
+}
+
+#[global_allocator]
+static ALLOC: Counting = Counting;
+
+/// Bytes allocated at the peak of `f`, above what was live when it started.
+fn peak_during<T>(f: impl FnOnce() -> T) -> (T, usize) {
+    let base = CURRENT.load(Ordering::Relaxed);
+    PEAK.store(base, Ordering::Relaxed);
+    let out = f();
+    (out, PEAK.load(Ordering::Relaxed).saturating_sub(base))
+}
+
+/// What decoding one HDF5 chunk of `limit` bytes from `input` may hold at
+/// once: the output, a few blocks of scratch (each no larger than the
+/// output), the offsets table, and the Zstandard decoder's state, which has
+/// a fixed ceiling: a window of at most 128 KiB (or twice the stream) and a
+/// block's table of sequences (up to 98,303 of 12 bytes, 1.2 MB).
+fn bound(limit: usize, input: &[u8]) -> usize {
+    6 * limit + 2 * input.len() + (2 << 20)
+}
+
+fn lock() -> std::sync::MutexGuard<'static, ()> {
+    SERIAL.lock().unwrap_or_else(|e| e.into_inner())
+}
+
+/// A 32-byte (extended) Blosc2 chunk header.
+fn chunk_header(ts: u8, nbytes: i32, blocksize: i32, cbytes: i32, special: u8) -> Vec<u8> {
+    let mut c = vec![5u8, 1, 0x05, ts];
+    for v in [nbytes, blocksize, cbytes] {
+        c.extend_from_slice(&v.to_le_bytes());
+    }
+    c.resize(32, 0);
+    c[31] = special << 4;
+    c
+}
+
+/// A chunk of `nbytes` bytes that repeats one value (special type 3).
+fn repeated(value: &[u8], nbytes: i32, blocksize: i32) -> Vec<u8> {
+    let mut c = chunk_header(
+        value.len() as u8,
+        nbytes,
+        blocksize,
+        32 + value.len() as i32,
+        3,
+    );
+    c.extend_from_slice(value);
+    c
+}
+
+/// A frame offset recording a special chunk of `kind` (1 zeros, 2 NaN).
+fn special_offset(kind: u8) -> [u8; 8] {
+    (((0x80 | kind) as i64) << 56).to_le_bytes()
+}
+
+/// A B2ND metalayer.
+fn nd_meta(shape: &[i64], chunks: &[i32], blocks: &[i32]) -> Vec<u8> {
+    let n = shape.len() as u8;
+    let mut m = vec![0x95, 0, n, 0x90 | n];
+    for s in shape {
+        m.push(0xd3);
+        m.extend_from_slice(&s.to_be_bytes());
+    }
+    for dims in [chunks, blocks] {
+        m.push(0x90 | n);
+        for d in dims {
+            m.push(0xd2);
+            m.extend_from_slice(&d.to_be_bytes());
+        }
+    }
+    m
+}
+
+/// A contiguous frame: header (with a `b2nd` metalayer if given), the data
+/// chunks, then the offsets chunk.
+fn frame(
+    meta: Option<&[u8]>,
+    nbytes: i64,
+    typesize: i32,
+    chunksize: i32,
+    data: &[u8],
+    offsets: &[u8],
+) -> Vec<u8> {
+    let mut h = vec![0u8; 91];
+    h[0] = 0x9e;
+    h[1] = 0xa8;
+    h[2..10].copy_from_slice(b"b2frame\0");
+    h[25] = 2;
+    match meta {
+        Some(m) => {
+            h.extend_from_slice(&[0xde, 0, 1, 0xa4]);
+            h.extend_from_slice(b"b2nd");
+            let at = h.len() as i32 + 5;
+            h.push(0xd2);
+            h.extend_from_slice(&at.to_be_bytes());
+            h.push(0xc6);
+            h.extend_from_slice(&(m.len() as u32).to_be_bytes());
+            h.extend_from_slice(m);
+        }
+        None => h.extend_from_slice(&[0xde, 0, 0]),
+    }
+    let header_len = h.len() as i32;
+    h[11..15].copy_from_slice(&header_len.to_be_bytes());
+    h[30..38].copy_from_slice(&nbytes.to_be_bytes());
+    h[39..47].copy_from_slice(&(data.len() as i64).to_be_bytes());
+    h[48..52].copy_from_slice(&typesize.to_be_bytes());
+    h[58..62].copy_from_slice(&chunksize.to_be_bytes());
+    h.extend_from_slice(data);
+    h.extend_from_slice(offsets);
+    let len = h.len() as u64;
+    h[16..24].copy_from_slice(&len.to_be_bytes());
+    h
+}
+
+/// The frame header's own sizes must not size the offsets chunk: a frame
+/// declaring 32 Mi chunks of 4 bytes, whose offsets chunk (40 bytes) says
+/// "one repeated offset, 256 MiB of them", made the decoder build all
+/// 256 MiB of offsets for a 1 MiB HDF5 chunk and then return 4 bytes.
+#[test]
+fn offsets_chunk_is_bounded_by_the_output_limit() {
+    let _g = lock();
+    let limit = 1 << 20;
+    let offsets_len: i32 = 256 << 20;
+    let nchunks = offsets_len as i64 / 8;
+    let offsets = repeated(&special_offset(1), offsets_len, 64 << 20);
+    let f = frame(None, nchunks * 4, 4, 4, &[], &offsets);
+    let (r, peak) = peak_during(|| blosc2_decompress(&f, limit));
+    assert!(r.is_err(), "decoded {:?} bytes", r.map(|v| v.len()));
+    assert!(
+        peak <= bound(limit, &f),
+        "peak {peak} bytes for a {}-byte frame",
+        f.len()
+    );
+    // The same frame with a variable chunk size (0): the offsets chunk
+    // alone says how many chunks there are.
+    let f = frame(None, nchunks * 4, 4, 0, &[], &offsets);
+    let (r, peak) = peak_during(|| blosc2_decompress(&f, limit));
+    assert!(r.is_err());
+    assert!(peak <= bound(limit, &f), "chunksize 0: peak {peak} bytes");
+}
+
+/// A legitimate frame of this shape (one chunk, its offset special) still
+/// decodes.
+#[test]
+fn small_frames_still_decode() {
+    let _g = lock();
+    let offsets = repeated(&special_offset(1), 8, 8);
+    let f = frame(None, 64, 4, 64, &[], &offsets);
+    assert_eq!(blosc2_decompress(&f, 64).unwrap(), vec![0; 64]);
+    let _ = blosc2_decompress_chunk;
+}
+
+/// A chunk that decodes to nothing kept its declared block size (up to
+/// 512 MiB) and allocated two scratch blocks of it: about 1 GiB for a
+/// 20-byte chunk.
+#[test]
+fn empty_chunk_does_not_allocate_its_block_size() {
+    let _g = lock();
+    let mut c = vec![5u8, 1, 0x01, 1];
+    for v in [0i32, 0x1FFF_F000, 20] {
+        c.extend_from_slice(&v.to_le_bytes());
+    }
+    c.resize(20, 0);
+    let (r, peak) = peak_during(|| blosc2_decompress_chunk(&c, 1 << 20));
+    assert_eq!(r.map(|v| v.len()).unwrap_or(0), 0);
+    assert!(
+        peak <= bound(0, &c),
+        "peak {peak} bytes for a 20-byte chunk"
+    );
+    // Inside a frame for a non-empty HDF5 chunk it is an error, not data.
+    let offsets = repeated(&0i64.to_le_bytes(), 8, 8);
+    let f = frame(None, 64, 4, 64, &c, &offsets);
+    let (r, peak) = peak_during(|| blosc2_decompress(&f, 64));
+    assert!(r.is_err(), "decoded {:?}", r.map(|v| v.len()));
+    assert!(peak <= bound(64, &f), "in a frame: peak {peak} bytes");
+}
+
+/// B2ND chunks were decoded whole, padding included, with up to 16x the
+/// HDF5 chunk size as their limit. Blocks are now placed as they are
+/// decoded, so the padding is never held.
+///
+/// Ten dimensions: nine of 3 split into blocks of 2 (padded to 4) and one
+/// of 4, so each chunk is 13x the array. One chunk, stored three ways: as a
+/// NaN chunk in the frame's offsets, as a repeated-value chunk, and as a
+/// chunk of stored (uncompressed) blocks.
+#[test]
+fn b2nd_padding_is_never_held() {
+    let _g = lock();
+    let ts = 4usize;
+    let mut shape = vec![3i64; 9];
+    shape.push(4);
+    let chunks: Vec<i32> = shape.iter().map(|&s| s as i32).collect();
+    let mut blocks = vec![2i32; 9];
+    blocks.push(4);
+    let meta = nd_meta(&shape, &chunks, &blocks);
+    let items: usize = shape.iter().product::<i64>() as usize;
+    let limit = items * ts;
+    let block_bytes = ts * blocks.iter().product::<i32>() as usize;
+    let ext_bytes = ts * 4usize.pow(9) * 4;
+    assert!(ext_bytes > 13 * limit);
+    let offsets = |off: [u8; 8]| repeated(&off, 8, 8);
+
+    let value = 1.5f32.to_le_bytes();
+    let stored = {
+        // Every block stored raw: block k holds the value k.
+        let mut c = chunk_header(4, ext_bytes as i32, block_bytes as i32, 0, 0);
+        c[2] = 0x02 | 0x10; // memcpyed, not split
+        c.truncate(16);
+        for k in 0..ext_bytes / block_bytes {
+            c.extend((k as f32).to_le_bytes().repeat(block_bytes / 4));
+        }
+        let n = c.len() as i32;
+        c[12..16].copy_from_slice(&n.to_le_bytes());
+        c
+    };
+    let cases: Vec<(&str, Vec<u8>)> = vec![
+        (
+            "NaN offset",
+            frame(
+                Some(&meta),
+                ext_bytes as i64,
+                4,
+                ext_bytes as i32,
+                &[],
+                &offsets(special_offset(2)),
+            ),
+        ),
+        (
+            "repeated value",
+            frame(
+                Some(&meta),
+                ext_bytes as i64,
+                4,
+                ext_bytes as i32,
+                &repeated(&value, ext_bytes as i32, block_bytes as i32),
+                &offsets(0i64.to_le_bytes()),
+            ),
+        ),
+        (
+            "stored blocks",
+            frame(
+                Some(&meta),
+                ext_bytes as i64,
+                4,
+                ext_bytes as i32,
+                &stored,
+                &offsets(0i64.to_le_bytes()),
+            ),
+        ),
+    ];
+    for (name, f) in cases {
+        let (r, peak) = peak_during(|| blosc2_decompress(&f, limit));
+        let out = r.unwrap_or_else(|e| panic!("{name}: {e}"));
+        assert_eq!(out.len(), limit, "{name}");
+        match name {
+            "NaN offset" => assert!(
+                out.chunks(4)
+                    .all(|v| f32::from_le_bytes(v.try_into().unwrap()).is_nan())
+            ),
+            "repeated value" => assert!(out.chunks(4).all(|v| v == value)),
+            _ => {
+                // Element (i0..i9) lies in block (i0/2, .., i8/2), numbered
+                // in C order over a 2x..x2x1 grid of blocks.
+                let mut idx = [0usize; 10];
+                for (e, v) in out.chunks(4).enumerate() {
+                    let mut n = e;
+                    for d in (0..10).rev() {
+                        idx[d] = n % shape[d] as usize;
+                        n /= shape[d] as usize;
+                    }
+                    let k = idx[..9].iter().fold(0, |k, &i| k * 2 + i / 2);
+                    assert_eq!(
+                        f32::from_le_bytes(v.try_into().unwrap()),
+                        k as f32,
+                        "{name} {e}"
+                    );
+                }
+            }
+        }
+        assert!(
+            peak <= bound(limit, &f),
+            "{name}: peak {peak} bytes for a {limit}-byte chunk ({}-byte frame)",
+            f.len()
+        );
+    }
+}
+
+/// A B2ND chunk larger than the array (here 16x, the old cap) is refused,
+/// or at least never allocated.
+#[test]
+fn b2nd_chunk_larger_than_the_array_is_not_allocated() {
+    let _g = lock();
+    let limit = 1 << 20;
+    let c = 16 * limit as i32;
+    let meta = nd_meta(&[limit as i64], &[c], &[c]);
+    let offsets = repeated(&special_offset(1), 8, 8);
+    let f = frame(Some(&meta), c as i64, 1, c, &[], &offsets);
+    let (r, peak) = peak_during(|| blosc2_decompress(&f, limit));
+    assert!(
+        peak <= bound(limit, &f),
+        "peak {peak} bytes ({:?})",
+        r.map(|v| v.len())
+    );
+}
+
+/// ruzstd reserves a frame's declared window (up to 100 MiB) before it
+/// decodes a frame with a decoder it has used before: a Blosc2 chunk of
+/// two 16-byte Zstandard streams, each declaring a 96 MiB window,
+/// allocated 96 MiB. c-blosc2 compresses each block with its size known,
+/// so its windows never exceed the block.
+#[test]
+fn zstd_window_is_bounded_by_the_output() {
+    let _g = lock();
+    let mut z = 0xfd2f_b528u32.to_le_bytes().to_vec();
+    // No single segment, no checksum; window 2^26 + 4/8 of it = 96 MiB.
+    z.extend_from_slice(&[0x00, (16 << 3) | 4]);
+    // One raw block, last, of 16 bytes.
+    let h = 1 | (16 << 3);
+    z.extend_from_slice(&[h as u8, (h >> 8) as u8, 0]);
+    z.extend_from_slice(&[7; 16]);
+    // Two blocks of 16 bytes, one stream each (not split), Zstandard
+    // (codec 4).
+    let chunk = |z: &[u8]| {
+        let mut c = vec![5u8, 1, 0x10 | (4 << 5), 1];
+        for v in [32i32, 16, 0] {
+            c.extend_from_slice(&v.to_le_bytes());
+        }
+        let first = 24 + 4 + z.len();
+        c.extend_from_slice(&24i32.to_le_bytes());
+        c.extend_from_slice(&(first as i32).to_le_bytes());
+        for _ in 0..2 {
+            c.extend_from_slice(&(z.len() as i32).to_le_bytes());
+            c.extend_from_slice(z);
+        }
+        let n = c.len() as i32;
+        c[12..16].copy_from_slice(&n.to_le_bytes());
+        c
+    };
+    let c = chunk(&z);
+    let (r, peak) = peak_during(|| blosc2_decompress_chunk(&c, 32));
+    assert!(peak <= bound(32, &c), "peak {peak} bytes ({r:?})");
+    assert!(r.is_err(), "{r:?}");
+    // The same streams with a window they can use read.
+    z[5] = 0;
+    assert_eq!(
+        blosc2_decompress_chunk(&chunk(&z), 32).unwrap(),
+        vec![7; 32]
+    );
+}
+
+/// xorshift64*: deterministic, so a failure reproduces.
+struct Rng(u64);
+
+impl Rng {
+    fn next(&mut self) -> u64 {
+        let mut x = self.0;
+        x ^= x >> 12;
+        x ^= x << 25;
+        x ^= x >> 27;
+        self.0 = x;
+        x.wrapping_mul(0x2545_F491_4F6C_DD1D)
+    }
+
+    fn below(&mut self, n: usize) -> usize {
+        (self.next() % n.max(1) as u64) as usize
+    }
+
+    /// A size that tends to the edges: small, a power of two, huge.
+    fn size(&mut self) -> i64 {
+        match self.below(6) {
+            0 => self.below(64) as i64,
+            1 => 1 << self.below(31),
+            2 => i32::MAX as i64 - self.below(4096) as i64,
+            3 => (1i64 << self.below(62)) + self.below(8) as i64,
+            4 => MAX_BLOCK - self.below(3) as i64,
+            _ => self.next() as i32 as i64,
+        }
+    }
+}
+
+const MAX_BLOCK: i64 = 536_866_816;
+
+/// One to four edits: bytes, or a size field written little-endian (chunk
+/// headers) or big-endian (frame headers), most often at a header's size
+/// fields.
+fn mutate(rng: &mut Rng, seed: &[u8], data_at: usize) -> Vec<u8> {
+    let mut v = seed.to_vec();
+    for _ in 0..1 + rng.below(4) {
+        let len = v.len();
+        if len < 16 {
+            v.push(rng.next() as u8);
+            continue;
+        }
+        match rng.below(8) {
+            0 => {
+                let i = rng.below(len);
+                v[i] ^= 1 << rng.below(8);
+            }
+            1 => {
+                let i = rng.below(len);
+                v[i] = rng.next() as u8;
+            }
+            2 => {
+                // Frame header: nbytes, cbytes (i64), typesize, chunksize.
+                let x = rng.size();
+                match rng.below(4) {
+                    0 if len >= 38 => v[30..38].copy_from_slice(&x.to_be_bytes()),
+                    1 if len >= 47 => v[39..47].copy_from_slice(&x.to_be_bytes()),
+                    2 if len >= 52 => v[48..52].copy_from_slice(&(x as i32).to_be_bytes()),
+                    _ if len >= 62 => v[58..62].copy_from_slice(&(x as i32).to_be_bytes()),
+                    _ => {}
+                }
+            }
+            3 | 4 => {
+                // A chunk header's nbytes, blocksize or cbytes: in the first
+                // data chunk, or anywhere (the offsets chunk comes last).
+                let at = if rng.below(2) == 0 && data_at + 16 <= len {
+                    data_at + 4 * (1 + rng.below(3))
+                } else {
+                    rng.below(len - 3)
+                };
+                let x = rng.size() as i32;
+                v[at..at + 4].copy_from_slice(&x.to_le_bytes());
+            }
+            5 => v.truncate(rng.below(len)),
+            6 => {
+                let at = rng.below(len);
+                v[at] = [0x10, 0x20, 0x30, 0x40, 0x05, 0x07, 0x02][rng.below(7)];
+            }
+            _ => {
+                let i = rng.below(len - 3);
+                let x = rng.size() as i32;
+                v[i..i + 4].copy_from_slice(&x.to_be_bytes());
+            }
+        }
+    }
+    v
+}
+
+/// Every fixture frame that decodes, with its decoded size.
+fn seeds() -> Vec<(Vec<u8>, usize)> {
+    let dir = std::path::Path::new(env!("CARGO_MANIFEST_DIR")).join("tests/fixtures/blosc2");
+    let mut v = Vec::new();
+    for e in std::fs::read_dir(dir).unwrap() {
+        let p = e.unwrap().path();
+        if p.extension().is_some_and(|x| x == "b2f")
+            && let Ok(out) = std::fs::read(p.with_extension("out"))
+        {
+            v.push((std::fs::read(&p).unwrap(), out.len()));
+        }
+    }
+    v.sort();
+    assert!(v.len() >= 20, "fixtures missing");
+    v
+}
+
+fn header_len(frame: &[u8]) -> usize {
+    i32::from_be_bytes(frame[11..15].try_into().unwrap()) as usize
+}
+
+/// Mutated fixture frames, decoded with their HDF5 chunk size as the
+/// limit, and their first chunks on their own: whatever they declare, no
+/// decode holds more than a small multiple of the output and the input.
+#[test]
+fn fuzzed_frames_and_chunks_stay_within_the_allocation_bound() {
+    let _g = lock();
+    let seeds = seeds();
+    let mut rng = Rng(0xb2a1);
+    let mut worst = (0.0f64, String::new());
+    for i in 0..20_000 {
+        let (seed, limit) = &seeds[rng.below(seeds.len())];
+        let f = mutate(&mut rng, seed, header_len(seed));
+        let (r, peak) = peak_during(|| blosc2_decompress(&f, *limit));
+        if let Ok(out) = &r {
+            assert!(out.len() <= *limit, "iteration {i}: output past the limit");
+        }
+        assert!(
+            peak <= bound(*limit, &f),
+            "iteration {i}: peak {peak} bytes for a {limit}-byte chunk from {} bytes ({:?})",
+            f.len(),
+            r.map(|v| v.len())
+        );
+        let ratio = peak as f64 / bound(*limit, &f) as f64;
+        if ratio > worst.0 {
+            worst = (
+                ratio,
+                format!(
+                    "frame iteration {i}: peak {peak}, limit {limit}, input {}",
+                    f.len()
+                ),
+            );
+        }
+    }
+    for i in 0..20_000 {
+        let (seed, _) = &seeds[rng.below(seeds.len())];
+        let at = header_len(seed);
+        let chunk = &seed[at..];
+        let c = mutate(&mut rng, chunk, 0);
+        let limit = 1 << 16;
+        let (r, peak) = peak_during(|| blosc2_decompress_chunk(&c, limit));
+        assert!(
+            peak <= bound(limit, &c),
+            "chunk iteration {i}: peak {peak} bytes from {} bytes ({:?})",
+            c.len(),
+            r.map(|v| v.len())
+        );
+    }
+    eprintln!("worst peak / bound: {:.2} ({})", worst.0, worst.1);
+}
+
+/// Frames built from random header sizes, offsets chunks and B2ND shapes
+/// (chunk and block shapes that pad, special and repeated-value chunks).
+#[test]
+fn random_frames_stay_within_the_allocation_bound() {
+    let _g = lock();
+    let mut rng = Rng(0xb2a2);
+    for i in 0..5_000 {
+        let ts = [1usize, 2, 4, 8][rng.below(4)];
+        let ndim = 1 + rng.below(8);
+        let mut shape = Vec::new();
+        let mut chunks = Vec::new();
+        let mut blocks = Vec::new();
+        for _ in 0..ndim {
+            let s = 1 + rng.below(if ndim > 3 { 4 } else { 40 });
+            let c = if rng.below(8) == 0 {
+                s * (1 + rng.below(4))
+            } else {
+                1 + rng.below(s)
+            };
+            let b = 1 + rng.below(c);
+            shape.push(s as i64);
+            chunks.push(c as i32);
+            blocks.push(b as i32);
+        }
+        let items: usize = shape.iter().product::<i64>() as usize;
+        let limit = items * ts;
+        let meta = nd_meta(&shape, &chunks, &blocks);
+        let ext: usize = ts
+            * chunks
+                .iter()
+                .zip(&blocks)
+                .map(|(&c, &b)| (c as usize).div_ceil(b as usize) * b as usize)
+                .product::<usize>();
+        let nchunks: usize = shape
+            .iter()
+            .zip(&chunks)
+            .map(|(&s, &c)| (s as usize).div_ceil(c as usize))
+            .product();
+        let block_bytes = ts * blocks.iter().product::<i32>() as usize;
+        let chunksize = if rng.below(4) == 0 {
+            rng.size()
+        } else {
+            ext as i64
+        };
+        let nbytes = if rng.below(4) == 0 {
+            rng.size()
+        } else {
+            (nchunks * ext) as i64
+        };
+        let off_n = if rng.below(4) == 0 {
+            rng.size() as i32
+        } else {
+            8 * nchunks as i32
+        };
+        let (data, off) = match rng.below(3) {
+            0 => (Vec::new(), special_offset(1 + rng.below(2) as u8)),
+            _ => {
+                let bs = if rng.below(4) == 0 {
+                    rng.size() as i32
+                } else {
+                    block_bytes as i32
+                };
+                let value: Vec<u8> = (0..ts).map(|_| rng.next() as u8).collect();
+                let n = if rng.below(4) == 0 {
+                    rng.size() as i32
+                } else {
+                    ext as i32
+                };
+                (repeated(&value, n, bs), 0i64.to_le_bytes())
+            }
+        };
+        let offsets = repeated(&off, off_n, off_n.clamp(1, 8));
+        let meta = (rng.below(4) != 0).then_some(meta.as_slice());
+        let f = frame(meta, nbytes, ts as i32, chunksize as i32, &data, &offsets);
+        let (r, peak) = peak_during(|| blosc2_decompress(&f, limit));
+        assert!(
+            peak <= bound(limit, &f),
+            "iteration {i}: peak {peak} bytes for a {limit}-byte chunk ({:?}, shape {shape:?} \
+             chunks {chunks:?} blocks {blocks:?})",
+            r.map(|v| v.len())
+        );
+    }
+}

crates/clawhdf5-format/tests/fixtures/blosc2/bytedelta.err

@@ -0,0 +1 @@
+filter 35

crates/clawhdf5-format/tests/fixtures/blosc2/generate.py

@@ -0,0 +1,129 @@
+"""Generate the Blosc2 frames the `filters_blosc2` unit tests decode.
+
+Each case is `<name>.b2f` (a Blosc2 contiguous frame, what the HDF5 Blosc2
+filter stores per chunk) and `<name>.out` (what decoding it must give: the
+first chunk of a plain frame, or the whole array in C order for a B2ND
+frame), or `<name>.err` (a frame clawhdf5 must refuse; the file holds a word
+the error message must contain).
+
+These cover what files written by h5py + hdf5plugin never contain, but a
+Blosc2 frame may: special chunks (repeated value, NaN, uninitialised), the
+delta filter over many blocks and odd type sizes, bit shuffle of blocks that
+are not a multiple of 8 elements, shuffle with a byte-group size, forced
+stream splitting, multi-chunk B2ND arrays with padded edge chunks and a
+chunk of zeros, and features clawhdf5 refuses (dictionaries, registered
+filters).
+
+Written with python-blosc2 4.13.1 (c-blosc2 3.3.4) in a scratch venv
+(`pip install blosc2`). Re-run only to regenerate:
+
+    python generate.py <this directory>
+"""
+import os
+import sys
+
+import blosc2
+import numpy as np
+
+out = sys.argv[1]
+
+
+def save(name, frame, expected):
+    with open(os.path.join(out, name + ".b2f"), "wb") as f:
+        f.write(frame)
+    with open(os.path.join(out, name + ".out"), "wb") as f:
+        f.write(expected)
+
+
+def save_err(name, frame, word):
+    with open(os.path.join(out, name + ".b2f"), "wb") as f:
+        f.write(frame)
+    with open(os.path.join(out, name + ".err"), "w") as f:
+        f.write(word)
+
+
+def plain(data, **cparams):
+    """A one-chunk super-chunk frame of `data`, as hdf5-blosc2 writes."""
+    data = np.ascontiguousarray(data)
+    cp = blosc2.CParams(typesize=data.dtype.itemsize, **cparams)
+    sc = blosc2.SChunk(chunksize=data.nbytes, cparams=cp)
+    sc.append_data(data)
+    return sc.to_cframe(), data.tobytes()
+
+
+def special(nitems, dtype, kind, value=None):
+    dt = np.dtype(dtype)
+    sc = blosc2.SChunk(chunksize=nitems * dt.itemsize,
+                       cparams=blosc2.CParams(typesize=dt.itemsize))
+    sc.fill_special(nitems, kind, value)
+    return sc.to_cframe()
+
+
+# Special chunks. A repeated value stays in the frame as a 33+ byte chunk;
+# NaN and uninitialised chunks become special offsets.
+save("value_i4", special(300, "<i4", blosc2.SpecialValue.VALUE, 123456),
+     np.full(300, 123456, "<i4").tobytes())
+save("value_f8", special(250, "<f8", blosc2.SpecialValue.VALUE, -2.5),
+     np.full(250, -2.5, "<f8").tobytes())
+save("nan_f4", special(500, "<f4", blosc2.SpecialValue.NAN),
+     np.full(500, np.nan, "<f4").tobytes())
+save("nan_f8", special(300, "<f8", blosc2.SpecialValue.NAN),
+     np.full(300, np.nan, "<f8").tobytes())
+save("zero_u2", special(2000, "<u2", blosc2.SpecialValue.ZERO), bytes(4000))
+# Uninitialised values: libhdf5 would hand back whatever memory it had;
+# clawhdf5 returns zeros.
+save("uninit_i8", special(64, "<i8", blosc2.SpecialValue.UNINIT), bytes(512))
+
+rng = np.random.default_rng(11)
+ramp = lambda n, dt: ((np.arange(n) * 7) % 1000 + rng.integers(0, 3, n)).astype(dt)
+# Slowly varying: what the delta filter is for (noise would be stored raw).
+smooth = lambda n, dt: (np.arange(n) // 3 + 1000).astype(dt)
+
+# Delta over many blocks, for type sizes 1, 2, 4, 8, 3 (bytes) and 16 (u64
+# pairs).
+for dt, n, codec in [("<u1", 2000, blosc2.Codec.LZ4), ("<i2", 1000, blosc2.Codec.BLOSCLZ),
+                     ("<i4", 700, blosc2.Codec.LZ4), ("<u8", 400, blosc2.Codec.BLOSCLZ)]:
+    save(f"delta_{np.dtype(dt).name}_{codec.name.lower()}",
+         *plain(smooth(n, dt), codec=codec, blocksize=256,
+                filters=[blosc2.Filter.DELTA], filters_meta=[0]))
+rec3 = np.frombuffer(smooth(3 * 300, "<u1").tobytes(), dtype="V3")
+save("delta_v3", *plain(rec3, blocksize=300, filters=[blosc2.Filter.DELTA], filters_meta=[0]))
+rec16 = np.frombuffer(smooth(2 * 200, "<u8").tobytes(), dtype="V16")
+save("delta_shuffle_v16", *plain(rec16, blocksize=512,
+                                 filters=[blosc2.Filter.DELTA, blosc2.Filter.SHUFFLE],
+                                 filters_meta=[0, 0]))
+
+# Bit shuffle of blocks whose element count is not a multiple of 8 (44-byte
+# blocks of 4-byte elements: 8 transposed, 3 copied).
+save("bitshuffle_odd_blocks", *plain(ramp(500, "<i4"), codec=blosc2.Codec.ZSTD, blocksize=44,
+                                     filters=[blosc2.Filter.BITSHUFFLE], filters_meta=[0]))
+# Shuffle in groups of 2 bytes of an 8-byte type (filters_meta).
+save("shuffle_meta2", *plain(ramp(500, "<i8"), codec=blosc2.Codec.ZLIB,
+                             filters=[blosc2.Filter.SHUFFLE], filters_meta=[2]))
+# Streams split per byte, and never split.
+save("always_split", *plain(ramp(1000, "<f4"), codec=blosc2.Codec.LZ4HC,
+                            splitmode=blosc2.SplitMode.ALWAYS_SPLIT))
+save("never_split", *plain(ramp(1500, "<u2"), codec=blosc2.Codec.ZSTD,
+                           splitmode=blosc2.SplitMode.NEVER_SPLIT))
+
+# B2ND arrays of several chunks whose edge chunks and blocks are padded, and
+# one whose middle chunk is all zeros (a special offset).
+for name, shape, chunks, blocks, dt in [
+    ("b2nd_2d", (37, 29), (10, 16), (4, 6), "<i4"),
+    ("b2nd_3d", (9, 10, 7), (4, 5, 3), (3, 2, 2), "<f4"),
+    ("b2nd_4d", (5, 7, 5, 6), (3, 2, 5, 4), (2, 2, 3, 3), "<u2"),
+]:
+    a = ramp(int(np.prod(shape)), dt).reshape(shape)
+    arr = blosc2.asarray(a, chunks=chunks, blocks=blocks)
+    save(name, arr.to_cframe(), a.tobytes())
+a = ramp(60 * 20, "<i2").reshape(60, 20)
+a[20:40, :] = 0
+arr = blosc2.asarray(a, chunks=(20, 20), blocks=(8, 16))
+save("b2nd_zero_chunk", arr.to_cframe(), a.tobytes())
+
+# Refused: a dictionary, and a registered filter (bytedelta).
+frame, _ = plain(ramp(4000, "<i4"), codec=blosc2.Codec.ZSTD, use_dict=True, blocksize=2048)
+save_err("zstd_dict", frame, "dictionar")
+frame, _ = plain(ramp(500, "<i4"), filters=[blosc2.Filter.SHUFFLE, blosc2.Filter.BYTEDELTA],
+                 filters_meta=[0, 4])
+save_err("bytedelta", frame, "filter 35")