Files
clawhdf5/crates/clawhdf5-format/src/file_writer.rs
T
osobhandClaude Opus 5.5 193a5f8a82 writer: track attribute creation order with track_order
h5py's track_order=True orders attributes as well as links; the writer
tracked links only. A tracking object's header now sets the attribute
creation order tracked/indexed flags and carries per-message creation
orders, an Attribute Info message holds the next order (inline too),
and dense storage gets a type-9 creation-order index. The file default
applies to datasets, with DatasetBuilder::track_order per dataset; more
than 65 535 attributes on a tracking object is an error (libhdf5's
counter is 2 bytes). The reader lists such attributes in creation
order.

h5py lists them in order (inline, dense, 20 000 on one dataset) and
keeps numbering in r+ mode, including its inline-to-dense move.

Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
2026-09-26 10:17:37 -05:00

2769 lines
102 KiB
Rust

//! HDF5 file creation (write pipeline).
//!
//! Produces valid HDF5 files with v3 superblock, v2 object headers,
//! link messages, contiguous datasets, inline and dense attributes.
#[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,
};
use crate::data_layout::VdsMapping;
use crate::dataspace::{Dataspace, DataspaceType};
use crate::error::FormatError;
use crate::link_message::{LinkMessage, LinkTarget};
use crate::message_type::MessageType;
use crate::metadata_index::{DatasetMetadata, MetadataBlock, MetadataIndex};
use crate::object_header_writer::ObjectHeaderWriter;
use crate::superblock::Superblock;
use crate::type_builders::{
DatasetBuilder, FinishedGroup, GroupBuilder, build_attr_message, fill_value_message,
};
use crate::writer_tree::{self, LinkTo};
// Re-export public types that moved to type_builders for API compatibility.
#[cfg(feature = "provenance")]
pub use crate::type_builders::ProvenanceConfig;
pub use crate::type_builders::{AttrValue, CompoundTypeBuilder, EnumTypeBuilder};
use crate::datatype::{CharacterSet, Datatype};
pub(crate) const OFFSET_SIZE: u8 = 8;
pub(crate) const LENGTH_SIZE: u8 = 8;
const SUPERBLOCK_SIZE: usize = 48;
/// Largest raw data a compact dataset can hold: the layout message (version,
/// class, 2-byte size, data) must fit an object header message, whose size
/// field is 2 bytes. Bigger "compact" requests fall back to contiguous storage.
const MAX_COMPACT_DATA_SIZE: usize = crate::object_header_writer::MAX_MESSAGE_SIZE - 4;
/// libhdf5's bounds on a file space page size (`H5F_FILE_SPACE_PAGE_SIZE_MIN`
/// and `_MAX`).
const MIN_FILE_SPACE_PAGE_SIZE: u32 = 512;
const MAX_FILE_SPACE_PAGE_SIZE: u32 = 1024 * 1024 * 1024;
/// Superblock extension object header for a file using the paged file-space
/// strategy: a single File Space Info message (0x0017), as libhdf5 writes it
/// for `fs_strategy="page"` without persisted free space.
fn build_paged_superblock_extension(page_size: u32) -> Result<Vec<u8>, FormatError> {
let mut fsinfo = Vec::new();
fsinfo.push(1); // version
fsinfo.push(1); // strategy: H5F_FSPACE_STRATEGY_PAGE
fsinfo.push(0); // persisting free space: no
write_length(&mut fsinfo, 1, LENGTH_SIZE); // free-space section threshold
write_length(&mut fsinfo, u64::from(page_size), LENGTH_SIZE);
fsinfo.extend_from_slice(&0u16.to_le_bytes()); // page end metadata threshold
write_undef_offset(&mut fsinfo, OFFSET_SIZE); // EOA before free-space info
let mut w = ObjectHeaderWriter::new();
// Flags as libhdf5 sets them: bit 2 (never share) and bit 4 (mark if
// unknown). Not constant: libhdf5 rewrites the message when it closes a
// file it opened for writing.
w.add_message_with_flags(MessageType::Unknown(0x0017), fsinfo, 0x14);
w.serialize()
}
/// Threshold for switching from compact (inline) to dense attribute storage.
const DENSE_ATTR_THRESHOLD: usize = 8;
/// Threshold for switching a group from compact (inline Link messages) to dense
/// link storage (fractal heap + v2 B-tree), matching libhdf5's default
/// `max_compact` of 8 links.
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: AttrStorage<'_>,
fill_message: &[u8],
refcount: u32,
) -> Result<Vec<u8>, FormatError> {
let mut w = ObjectHeaderWriter::new();
w.add_message_with_flags(MessageType::Datatype, dt.serialize(), 0x01);
w.add_message(MessageType::Dataspace, ds.serialize(LENGTH_SIZE));
w.add_message_with_flags(MessageType::FillValue, fill_message.to_vec(), 0x01);
w.add_message(MessageType::DataLayout, layout_message.to_vec());
if let Some(pm) = pipeline_message {
w.add_message(MessageType::FilterPipeline, pm.to_vec());
}
attrs.add_to(&mut w);
add_refcount(&mut w, refcount);
w.serialize()
}
#[allow(clippy::too_many_arguments)]
pub(crate) fn build_dataset_oh(
dt: &Datatype,
ds: &Dataspace,
data_addr: u64,
data_size: u64,
attrs: AttrStorage<'_>,
fill_message: &[u8],
refcount: u32,
) -> Result<Vec<u8>, FormatError> {
let mut w = ObjectHeaderWriter::new();
w.add_message_with_flags(MessageType::Datatype, dt.serialize(), 0x01);
w.add_message(MessageType::Dataspace, ds.serialize(LENGTH_SIZE));
w.add_message_with_flags(MessageType::FillValue, fill_message.to_vec(), 0x01);
let mut dl = Vec::new();
dl.push(4); // version
dl.push(1); // class = contiguous
// An empty dataset has no storage: its address must be the undefined
// address, as libhdf5 writes it. A real address with size 0 trips
// libhdf5's `addr + size <= addr` overflow check, and it refuses the
// dataset as "invalid dataset size, likely file corruption" — which made
// every store with no sessions or knowledge graph unreadable by h5py.
let data_addr = if data_size == 0 { u64::MAX } else { data_addr };
dl.extend_from_slice(&data_addr.to_le_bytes());
dl.extend_from_slice(&data_size.to_le_bytes());
w.add_message(MessageType::DataLayout, dl);
attrs.add_to(&mut w);
add_refcount(&mut w, refcount);
w.serialize()
}
/// Build a compact dataset object header where data is stored inline.
pub(crate) fn build_compact_dataset_oh(
dt: &Datatype,
ds: &Dataspace,
data: &[u8],
attrs: AttrStorage<'_>,
fill_message: &[u8],
refcount: u32,
) -> Result<Vec<u8>, FormatError> {
let mut w = ObjectHeaderWriter::new();
w.add_message_with_flags(MessageType::Datatype, dt.serialize(), 0x01);
w.add_message(MessageType::Dataspace, ds.serialize(LENGTH_SIZE));
w.add_message_with_flags(MessageType::FillValue, fill_message.to_vec(), 0x01);
// Compact layout message: version=4, class=0, u16 size, inline data
let mut dl = Vec::new();
dl.push(4); // version
dl.push(0); // class = compact
dl.extend_from_slice(&(data.len() as u16).to_le_bytes());
dl.extend_from_slice(data);
w.add_message(MessageType::DataLayout, dl);
attrs.add_to(&mut w);
add_refcount(&mut w, refcount);
w.serialize()
}
/// Build a group's object header. `link_info` is its Link Info message;
/// with `dense_links` the links live in the fractal heap it points at and are
/// not written inline.
pub(crate) fn build_group_oh(
links: &[LinkMessage],
link_info: &[u8],
dense_links: bool,
attrs: AttrStorage<'_>,
refcount: u32,
) -> Result<Vec<u8>, FormatError> {
let mut w = ObjectHeaderWriter::new();
w.add_message(MessageType::LinkInfo, link_info.to_vec());
// Group Info (version 0, default link-phase thresholds, no estimates).
// Readers don't need it, but libhdf5 reads it before inserting a link:
// without one, adding a link to a group we wrote (h5py in "r+" mode)
// failed with "message type not found".
w.add_message(MessageType::GroupInfo, vec![0, 0]);
if !dense_links {
for link in links {
w.add_message(MessageType::Link, link.serialize(OFFSET_SIZE));
}
}
attrs.add_to(&mut w);
add_refcount(&mut w, refcount);
w.serialize()
}
/// An object with more than one hard link records the count in an Object
/// Reference Count message (libhdf5 omits it for a count of one). Without
/// it, libhdf5 deleting one of the links would free an object that is still
/// linked.
fn add_refcount(w: &mut ObjectHeaderWriter, refcount: u32) {
if refcount > 1 {
let mut msg = vec![0u8]; // version
msg.extend_from_slice(&refcount.to_le_bytes());
w.add_message(MessageType::ObjectReferenceCount, msg);
}
}
/// The character set a link name is written with: UTF-8 when it is not
/// plain ASCII, as h5py writes it.
fn name_charset(name: &str) -> CharacterSet {
if name.is_ascii() {
CharacterSet::Ascii
} else {
CharacterSet::Utf8
}
}
/// The Link message for `link`, whose group and dataset targets are at the
/// given addresses (indexed as in the writer tree).
fn link_message(link: &writer_tree::Link, group_addrs: &[u64], ds_addrs: &[u64]) -> LinkMessage {
let link_target = match &link.to {
LinkTo::Group(g) => LinkTarget::Hard {
object_header_address: group_addrs.get(*g).copied().unwrap_or(0),
},
LinkTo::Dataset(d) => LinkTarget::Hard {
object_header_address: ds_addrs.get(*d).copied().unwrap_or(0),
},
LinkTo::Soft(target_path) => LinkTarget::Soft {
target_path: target_path.clone(),
},
LinkTo::External { file, path } => LinkTarget::External {
filename: file.clone(),
object_path: path.clone(),
},
};
LinkMessage {
name: link.name.clone(),
link_target,
creation_order: link.creation_order,
charset: name_charset(&link.name),
}
}
/// Link Info message for a group with compact (inline) links: no heap, no
/// B-trees. A group tracking creation order records the next order to use.
fn compact_link_info(track_order: bool, nlinks: usize) -> Vec<u8> {
let max_corder = track_order.then_some(nlinks as u64);
serialize_link_info(
max_corder,
u64::MAX,
u64::MAX,
track_order.then_some(u64::MAX),
)
}
// ---- Dense attribute blob ----
/// Pre-built dense attribute storage (fractal heap + B-tree v2 + attribute info message).
pub(crate) struct DenseAttrBlob {
/// Serialized AttributeInfo message data (to embed in the object header).
pub(crate) attr_info_message: Vec<u8>,
/// The combined fractal heap header + direct block + B-tree v2 bytes.
pub(crate) blob: Vec<u8>,
}
/// A fractal heap holding a set of serialized objects, plus the heap IDs that
/// address them. Shared by dense attribute and dense link storage, which differ
/// only in their v2 B-tree record layout.
pub(crate) struct FractalHeapBlock {
/// The complete heap bytes: FRHP header, then either a single root direct
/// block, or a root indirect block (FHIB) followed by its direct blocks.
blob: Vec<u8>,
/// Address of the fractal heap header.
frhp_addr: u64,
/// Address where the v2 B-tree should be placed (right after the heap).
btree_addr: u64,
/// Heap ID for each object, in input order.
heap_ids: Vec<Vec<u8>>,
/// Heap ID length (bytes).
heap_id_length: u16,
}
/// Build a fractal heap for `serialized` objects, laid out at `base_address`.
///
/// Uses a single root direct block when the data fits in one (≤ the maximum
/// direct block size), otherwise a root indirect block over multiple direct
/// blocks following the doubling table. The caller builds the matching v2
/// B-tree (type 5 for links, type 8 for attributes) at the returned
/// `btree_addr`.
pub(crate) fn build_single_block_fractal_heap(
serialized: &[Vec<u8>],
base_address: u64,
max_heap_size: u16,
heap_id_length: u16,
) -> Result<FractalHeapBlock, FormatError> {
let os = OFFSET_SIZE as usize;
let ls = LENGTH_SIZE as usize;
let block_offset_bytes = (max_heap_size as usize).div_ceil(8);
let max_direct_block_size: u64 = 65536;
// Direct block layout: sig(4) + ver(1) + heap_addr(os) + block_offset(bo_bytes)
// + checksum(4) [when flags bit 1 set] + data...
let dblock_header_size = 4 + 1 + os + block_offset_bytes + 4; // +4 for checksum
// 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;
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 \
object holds at most {max_managed} bytes (huge heap objects are \
not written)",
big.len()
)));
}
let total_data_size: usize = serialized.iter().map(|s| s.len()).sum();
let dblock_content_size = dblock_header_size + total_data_size;
let starting_block_size = dblock_content_size.next_power_of_two().max(512) as u64;
// When the objects don't fit in a single direct block, fall back to a
// multi-block heap with a root indirect block.
if starting_block_size > max_direct_block_size {
return build_multiblock_fractal_heap(
serialized,
base_address,
max_heap_size,
heap_id_length,
);
}
// Fractal heap header size
let frhp_size = 4
+ 1
+ 2
+ 2
+ 1
+ 4
+ ls
+ os
+ ls
+ os
+ ls
+ ls
+ ls
+ ls
+ ls
+ ls
+ ls
+ ls
+ 2
+ ls
+ ls
+ 2
+ 2
+ os
+ 2
+ 4;
let frhp_addr = base_address;
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 free_space = data_space - total_data_size;
// Build fractal heap header
let mut frhp = Vec::with_capacity(frhp_size);
frhp.extend_from_slice(b"FRHP");
frhp.push(0); // version
frhp.extend_from_slice(&heap_id_length.to_le_bytes());
frhp.extend_from_slice(&0u16.to_le_bytes()); // io_filter_encoded_length
frhp.push(0x02); // flags: bit 1 = checksum direct blocks
frhp.extend_from_slice(&(max_managed as u32).to_le_bytes());
write_length(&mut frhp, 0, LENGTH_SIZE); // next_huge_object_id
write_undef_offset(&mut frhp, OFFSET_SIZE); // btree_huge_objects_address
write_length(&mut frhp, free_space as u64, LENGTH_SIZE); // free_space_managed_blocks
write_undef_offset(&mut frhp, OFFSET_SIZE); // free_space_mgr_addr
write_length(&mut frhp, starting_block_size, LENGTH_SIZE); // managed_space_in_heap
write_length(&mut frhp, starting_block_size, LENGTH_SIZE); // allocated_managed_space
write_length(&mut frhp, 0, LENGTH_SIZE); // dblock_alloc_iter
write_length(&mut frhp, serialized.len() as u64, LENGTH_SIZE); // managed_objects_count
write_length(&mut frhp, 0, LENGTH_SIZE); // huge_objects_size
write_length(&mut frhp, 0, LENGTH_SIZE); // huge_objects_count
write_length(&mut frhp, 0, LENGTH_SIZE); // tiny_objects_size
write_length(&mut frhp, 0, LENGTH_SIZE); // tiny_objects_count
frhp.extend_from_slice(&4u16.to_le_bytes()); // table_width
write_length(&mut frhp, starting_block_size, LENGTH_SIZE);
write_length(&mut frhp, max_direct_block_size, LENGTH_SIZE); // max_direct_block_size
frhp.extend_from_slice(&max_heap_size.to_le_bytes());
let sri: u16 = 1;
frhp.extend_from_slice(&sri.to_le_bytes()); // starting_row_of_indirect_blocks
write_offset(&mut frhp, dblock_addr, OFFSET_SIZE);
frhp.extend_from_slice(&0u16.to_le_bytes()); // root is direct block
let frhp_checksum = crate::checksum::jenkins_lookup3(&frhp);
frhp.extend_from_slice(&frhp_checksum.to_le_bytes());
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);
dblock.extend_from_slice(b"FHDB");
dblock.push(0); // version
write_offset(&mut dblock, frhp_addr, OFFSET_SIZE);
dblock.extend_from_slice(&vec![0u8; block_offset_bytes]); // block_offset = 0 for root
let cksum_pos = dblock.len();
dblock.extend_from_slice(&[0u8; 4]); // checksum placeholder
debug_assert_eq!(dblock.len(), dblock_header_size);
// Data area starts after header
let mut obj_offsets: Vec<(u64, u64)> = Vec::with_capacity(serialized.len());
for s in serialized {
let offset_in_heap = dblock.len() as u64;
obj_offsets.push((offset_in_heap, s.len() as u64));
dblock.extend_from_slice(s);
}
// Pad to full block size
dblock.resize(starting_block_size as usize, 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);
// Build heap IDs
let heap_ids: Vec<Vec<u8>> = obj_offsets
.iter()
.map(|(off, len)| encode_managed_id(*off, *len, max_heap_size, heap_id_length))
.collect();
let mut blob = Vec::with_capacity(frhp.len() + dblock.len());
blob.extend_from_slice(&frhp);
blob.extend_from_slice(&dblock);
Ok(FractalHeapBlock {
blob,
frhp_addr,
btree_addr,
heap_ids,
heap_id_length,
})
}
/// Build a multi-block fractal heap: a root indirect block (FHIB) over direct
/// blocks sized by the doubling table. Used when the objects don't fit in a
/// single direct block.
///
/// Rows of the doubling table whose block size exceeds the maximum direct
/// block size hold child indirect blocks, as the HDF5 spec (and libhdf5)
/// reads them: a child in row `r` spans that row's block size of heap space
/// and has `log2(size) - log2(start * width) + 1` rows of its own, which may
/// in turn hold indirect blocks. Objects are packed into direct blocks in
/// heap-offset order and never span blocks; a block too small for the next
/// object is left unallocated (an undefined address), as libhdf5 skips rows
/// when it needs a bigger block. The caller has checked that every object
/// fits a maximum-size direct block (there is no huge-object path).
fn build_multiblock_fractal_heap(
serialized: &[Vec<u8>],
base_address: u64,
max_heap_size: u16,
heap_id_length: u16,
) -> Result<FractalHeapBlock, FormatError> {
let os = OFFSET_SIZE as usize;
let block_offset_bytes = (max_heap_size as usize).div_ceil(8);
let geom = HeapGeometry {
width: 4,
starting_block_size: 512,
max_direct_block_size: 65536,
dblock_header_size: 4 + 1 + os + block_offset_bytes + 4,
iblock_fixed_size: 5 + os + block_offset_bytes + 4,
max_heap_size,
};
// ---- Pack objects into the doubling table ----
let mut packer = HeapPacker {
geom: &geom,
objects: serialized,
next: 0,
blocks: Vec::new(),
obj_loc: vec![(0, 0); serialized.len()],
};
let root = packer.fill(0, None)?;
let HeapPacker {
blocks, obj_loc, ..
} = packer;
// ---- Addresses ----
let frhp_size = frhp_header_size(os, LENGTH_SIZE as usize);
let frhp_addr = base_address;
let fhib_addr = frhp_addr + frhp_size as u64;
let heap_len = root.subtree_size(&geom, &blocks);
let btree_addr = fhib_addr + heap_len;
// Bookkeeping totals, as libhdf5 keeps them: the managed space is what
// the root's rows span, the allocated space the direct blocks written,
// and the allocation iterator the heap offset after the last of them.
let cur_rows = root.nrows as u16;
let managed_space: u64 = (0..root.nrows).map(|r| geom.row_size(r) * geom.width).sum();
let alloc_space: u64 = blocks.iter().map(|b| b.size).sum();
let used: u64 = blocks
.iter()
.map(|b| geom.dblock_header_size as u64 + b.data.len() as u64)
.sum();
let free_space = alloc_space.saturating_sub(used);
let alloc_iter = blocks.last().map_or(0, |b| b.heap_offset + b.size);
// ---- FRHP header ----
let max_managed = geom.max_managed();
let frhp = write_frhp(WriteFrhp {
heap_id_length,
max_managed,
free_space,
managed_space,
alloc_space,
alloc_iter,
nobjects: serialized.len() as u64,
table_width: geom.width as u16,
starting_block_size: geom.starting_block_size,
max_direct_block_size: geom.max_direct_block_size,
max_heap_size,
root_addr: fhib_addr,
cur_rows,
});
debug_assert_eq!(frhp.len(), frhp_size);
// ---- Indirect and direct blocks, depth first after the root ----
let mut blob = frhp;
root.emit(&geom, &blocks, frhp_addr, fhib_addr, &mut blob);
debug_assert_eq!(blob.len() as u64, frhp_size as u64 + heap_len);
let heap_ids: Vec<Vec<u8>> = obj_loc
.iter()
.map(|(off, len)| encode_managed_id(*off, *len, max_heap_size, heap_id_length))
.collect();
Ok(FractalHeapBlock {
blob,
frhp_addr,
btree_addr,
heap_ids,
heap_id_length,
})
}
/// The doubling table of a heap the writer builds.
struct HeapGeometry {
width: u64,
starting_block_size: u64,
max_direct_block_size: u64,
dblock_header_size: usize,
/// An indirect block's size without its child entries.
iblock_fixed_size: usize,
max_heap_size: u16,
}
impl HeapGeometry {
fn row_size(&self, row: usize) -> u64 {
block_size_for_row(self.starting_block_size, row)
}
/// Rows holding direct blocks: `log2(max_direct / start) + 2`.
fn max_direct_rows(&self) -> usize {
(self.max_direct_block_size / self.starting_block_size).ilog2() as usize + 2
}
/// `log2(start * width)`, libhdf5's `first_row_bits`.
fn first_row_bits(&self) -> u32 {
(self.starting_block_size * self.width).ilog2()
}
/// Rows of an indirect block spanning `size` bytes of heap space
/// (libhdf5's `H5HF__dtable_size_to_rows`).
fn rows_for_size(&self, size: u64) -> usize {
(size.ilog2() - self.first_row_bits() + 1) as usize
}
/// Rows the root indirect block can have: enough to span the heap's
/// whole `2^max_heap_size` address space.
fn max_root_rows(&self) -> usize {
(u32::from(self.max_heap_size) - self.first_row_bits() + 1) as usize
}
/// The largest object a direct block holds.
fn max_managed(&self) -> u32 {
(self.max_direct_block_size - self.dblock_header_size as u64) as u32
}
}
/// A direct block the packer filled.
struct HeapDirectBlock {
size: u64,
heap_offset: u64,
data: Vec<u8>,
}
/// One entry of an indirect block.
enum HeapSlot {
/// Not allocated (undefined address).
Empty,
/// Index into the packer's direct blocks.
Direct(usize),
Indirect(HeapIndirectBlock),
}
struct HeapIndirectBlock {
heap_offset: u64,
nrows: usize,
/// `nrows * width` entries, row-major.
slots: Vec<HeapSlot>,
}
impl HeapIndirectBlock {
fn own_size(&self, geom: &HeapGeometry) -> u64 {
(geom.iblock_fixed_size + self.slots.len() * OFFSET_SIZE as usize) as u64
}
/// Bytes of this block and everything below it.
fn subtree_size(&self, geom: &HeapGeometry, blocks: &[HeapDirectBlock]) -> u64 {
self.own_size(geom)
+ self
.slots
.iter()
.map(|s| match s {
HeapSlot::Empty => 0,
HeapSlot::Direct(i) => blocks[*i].size,
HeapSlot::Indirect(ib) => ib.subtree_size(geom, blocks),
})
.sum::<u64>()
}
/// Append this block at `addr` (= `out`'s current end, relative to the
/// same base as `frhp_addr`), then its children in entry order.
fn emit(
&self,
geom: &HeapGeometry,
blocks: &[HeapDirectBlock],
frhp_addr: u64,
addr: u64,
out: &mut Vec<u8>,
) {
let block_offset_bytes = (geom.max_heap_size as usize).div_ceil(8);
let start = out.len();
out.extend_from_slice(b"FHIB");
out.push(0); // version
write_offset(out, frhp_addr, OFFSET_SIZE);
out.extend_from_slice(&self.heap_offset.to_le_bytes()[..block_offset_bytes]);
let mut child = addr + self.own_size(geom);
for s in &self.slots {
match s {
HeapSlot::Empty => write_undef_offset(out, OFFSET_SIZE),
HeapSlot::Direct(i) => {
write_offset(out, child, OFFSET_SIZE);
child += blocks[*i].size;
}
HeapSlot::Indirect(ib) => {
write_offset(out, child, OFFSET_SIZE);
child += ib.subtree_size(geom, blocks);
}
}
}
let checksum = crate::checksum::jenkins_lookup3(&out[start..]);
out.extend_from_slice(&checksum.to_le_bytes());
let mut child = addr + self.own_size(geom);
for s in &self.slots {
match s {
HeapSlot::Empty => {}
HeapSlot::Direct(i) => {
let b = &blocks[*i];
let d = out.len();
out.extend_from_slice(b"FHDB");
out.push(0); // version
write_offset(out, frhp_addr, OFFSET_SIZE);
out.extend_from_slice(&b.heap_offset.to_le_bytes()[..block_offset_bytes]);
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);
let cksum = crate::checksum::jenkins_lookup3(&out[d..]);
out[cksum_pos..cksum_pos + 4].copy_from_slice(&cksum.to_le_bytes());
child += b.size;
}
HeapSlot::Indirect(ib) => {
ib.emit(geom, blocks, frhp_addr, child, out);
child += ib.subtree_size(geom, blocks);
}
}
}
}
}
/// Packs objects into a heap's doubling table in heap-offset order.
struct HeapPacker<'a> {
geom: &'a HeapGeometry,
objects: &'a [Vec<u8>],
/// The next object to place.
next: usize,
blocks: Vec<HeapDirectBlock>,
/// Each object's (heap offset, length).
obj_loc: Vec<(u64, u64)>,
}
impl HeapPacker<'_> {
/// Fill an indirect block at `heap_offset` with `nrows` rows, or, for the
/// root (`None`), with as many rows as the objects need.
fn fill(
&mut self,
heap_offset: u64,
nrows: Option<usize>,
) -> Result<HeapIndirectBlock, FormatError> {
let geom = self.geom;
let width = geom.width as usize;
let mut slots = Vec::new();
let mut off = heap_offset;
let mut row = 0usize;
while self.next < self.objects.len() && nrows.is_none_or(|n| row < n) {
if nrows.is_none() && row >= geom.max_root_rows() {
return Err(FormatError::SerializationError(format!(
"fractal heap: {} objects do not fit its {}-bit address space",
self.objects.len(),
geom.max_heap_size
)));
}
let size = geom.row_size(row);
for _ in 0..width {
if self.next == self.objects.len() {
slots.push(HeapSlot::Empty);
} else if row < geom.max_direct_rows() {
slots.push(self.fill_direct(off, size));
} else {
let child_rows = geom.rows_for_size(size);
// 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)
{
slots.push(HeapSlot::Empty);
off += size;
continue;
}
let child = self.fill(off, Some(child_rows))?;
let used = child.slots.iter().any(|s| !matches!(s, HeapSlot::Empty));
slots.push(if used {
HeapSlot::Indirect(child)
} else {
HeapSlot::Empty
});
}
off += size;
}
row += 1;
}
let nrows = nrows.unwrap_or(row);
slots.resize_with(nrows * width, || HeapSlot::Empty);
Ok(HeapIndirectBlock {
heap_offset,
nrows,
slots,
})
}
/// Fill the direct block at `heap_offset` with as many of the next
/// 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 mut data = Vec::new();
while let Some(obj) = self.objects.get(self.next) {
if data.len() + obj.len() > capacity {
break;
}
self.obj_loc[self.next] =
(heap_offset + (header + data.len()) as u64, obj.len() as u64);
data.extend_from_slice(obj);
self.next += 1;
}
if data.is_empty() && self.objects.get(self.next).is_some_and(|o| !o.is_empty()) {
return HeapSlot::Empty;
}
self.blocks.push(HeapDirectBlock {
size,
heap_offset,
data,
});
HeapSlot::Direct(self.blocks.len() - 1)
}
}
/// Doubling-table block size for `row`: rows 0 and 1 share the starting size;
/// row r (r ≥ 1) is `start * 2^(r-1)`.
fn block_size_for_row(starting_block_size: u64, row: usize) -> u64 {
if row <= 1 {
starting_block_size
} else {
starting_block_size << (row - 1)
}
}
/// Size in bytes of the FRHP header for the given offset/length sizes.
fn frhp_header_size(os: usize, ls: usize) -> usize {
4 + 1
+ 2
+ 2
+ 1
+ 4
+ ls
+ os
+ ls
+ os
+ ls
+ ls
+ ls
+ ls
+ ls
+ ls
+ ls
+ ls
+ 2
+ ls
+ ls
+ 2
+ 2
+ os
+ 2
+ 4
}
/// Parameters for [`write_frhp`].
struct WriteFrhp {
heap_id_length: u16,
max_managed: u32,
free_space: u64,
managed_space: u64,
alloc_space: u64,
/// Heap offset of the next direct block to allocate.
alloc_iter: u64,
nobjects: u64,
table_width: u16,
starting_block_size: u64,
max_direct_block_size: u64,
max_heap_size: u16,
root_addr: u64,
cur_rows: u16,
}
/// Serialize a fractal heap header (FRHP).
fn write_frhp(p: WriteFrhp) -> Vec<u8> {
let mut frhp = Vec::with_capacity(frhp_header_size(OFFSET_SIZE as usize, LENGTH_SIZE as usize));
frhp.extend_from_slice(b"FRHP");
frhp.push(0); // version
frhp.extend_from_slice(&p.heap_id_length.to_le_bytes());
frhp.extend_from_slice(&0u16.to_le_bytes()); // io_filter_encoded_length
frhp.push(0x02); // flags: bit 1 = checksum direct blocks
frhp.extend_from_slice(&p.max_managed.to_le_bytes());
write_length(&mut frhp, 0, LENGTH_SIZE); // next_huge_object_id
write_undef_offset(&mut frhp, OFFSET_SIZE); // btree_huge_objects_address
write_length(&mut frhp, p.free_space, LENGTH_SIZE); // free_space_managed_blocks
write_undef_offset(&mut frhp, OFFSET_SIZE); // free_space_mgr_addr
write_length(&mut frhp, p.managed_space, LENGTH_SIZE); // managed_space_in_heap
write_length(&mut frhp, p.alloc_space, LENGTH_SIZE); // allocated_managed_space
write_length(&mut frhp, p.alloc_iter, LENGTH_SIZE); // dblock_alloc_iter
write_length(&mut frhp, p.nobjects, LENGTH_SIZE); // managed_objects_count
write_length(&mut frhp, 0, LENGTH_SIZE); // huge_objects_size
write_length(&mut frhp, 0, LENGTH_SIZE); // huge_objects_count
write_length(&mut frhp, 0, LENGTH_SIZE); // tiny_objects_size
write_length(&mut frhp, 0, LENGTH_SIZE); // tiny_objects_count
frhp.extend_from_slice(&p.table_width.to_le_bytes());
write_length(&mut frhp, p.starting_block_size, LENGTH_SIZE);
write_length(&mut frhp, p.max_direct_block_size, LENGTH_SIZE);
frhp.extend_from_slice(&p.max_heap_size.to_le_bytes());
frhp.extend_from_slice(&1u16.to_le_bytes()); // starting # rows in root indirect block
write_offset(&mut frhp, p.root_addr, OFFSET_SIZE);
frhp.extend_from_slice(&p.cur_rows.to_le_bytes());
let checksum = crate::checksum::jenkins_lookup3(&frhp);
frhp.extend_from_slice(&checksum.to_le_bytes());
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();
let name_hashes: Vec<u32> = attrs
.iter()
.map(|a| crate::checksum::jenkins_lookup3(a.name.as_bytes()))
.collect();
// Attribute heaps use max_heap_size 40 / heap ID length 8 (matching libhdf5).
let heap = build_single_block_fractal_heap(&serialized, base_address, 40, 8)?;
let frhp_addr = heap.frhp_addr;
let btree_addr = heap.btree_addr;
let heap_id_length = heap.heap_id_length;
let heap_ids = &heap.heap_ids;
// 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 order: Vec<usize> = (0..attrs.len()).collect();
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_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
rec
})
.collect();
let bthd_addr = btree_addr;
let mut blob = heap.blob;
blob.extend_from_slice(&dense_v2_btree(8, record_size, &records, 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,
blob,
})
}
// ---- Dense link blob ----
/// Pre-built dense link storage (fractal heap + B-tree v2 + link-info message).
pub(crate) struct DenseLinkBlob {
/// Serialized LinkInfo message (to embed in the group's object header).
pub(crate) link_info_message: Vec<u8>,
/// The combined fractal heap, name-index B-tree and (when creation order
/// is tracked) creation-order-index B-tree bytes.
pub(crate) blob: Vec<u8>,
}
/// libhdf5's node size for the dense link and attribute indexes it creates
/// (`H5G_NAME_BT2_NODE_SIZE`, `H5A_NAME_BT2_NODE_SIZE`, and the
/// creation-order indexes'), with their split and merge percentages.
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,
) -> Result<Vec<u8>, FormatError> {
let rs = usize::from(record_size);
let n = records.len();
let node_size = if n <= usize::from(u16::MAX) {
let btlf_size = 4 + 1 + 1 + n * rs + 4;
let max_node = 4 + 1 + 1 + usize::from(u16::MAX) * rs + 4;
u32::try_from(btlf_size.next_power_of_two().max(512).min(max_node))
.map_err(|_| FormatError::Overflow("B-tree v2 node size".into()))?
} else {
DENSE_BT2_NODE_SIZE
};
let flat: Vec<u8> = records
.iter()
.inspect(|r| debug_assert_eq!(r.len(), rs))
.flat_map(|r| r.iter().copied())
.collect();
build_btree_v2(
BTreeV2Params {
tree_type: btree_type,
node_size,
record_size,
split_percent: DENSE_BT2_SPLIT_PERCENT,
merge_percent: DENSE_BT2_MERGE_PERCENT,
},
&flat,
addr,
OFFSET_SIZE,
LENGTH_SIZE,
)
}
/// Build dense link storage for a group's links, laid out at `base_address`.
///
/// Mirrors [`build_dense_attrs`]: each link is stored as a serialized Link
/// message in a fractal heap, indexed by a v2 B-tree of **type 5** (link-name
/// index, record = name hash + heap ID). With `track_order` (every link then
/// carries its creation order) a **type 6** B-tree (creation-order index,
/// record = creation order + heap ID) follows, as libhdf5 writes for a group
/// created with an indexed creation order. The returned LinkInfo message
/// points at the heap and the B-trees.
pub(crate) fn build_dense_links(
links: &[LinkMessage],
base_address: u64,
track_order: bool,
) -> Result<DenseLinkBlob, FormatError> {
let serialized: Vec<Vec<u8>> = links.iter().map(|l| l.serialize(OFFSET_SIZE)).collect();
// libhdf5's link heap uses max_heap_size 32 / heap ID length 7 (vs 40/8 for
// attributes), giving a 7-byte heap ID and an 11-byte type-5 record.
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
// 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(|&(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)| {
let mut rec = hash.to_le_bytes().to_vec();
rec.extend_from_slice(&heap.heap_ids[i]);
rec
})
.collect();
let name_bt_addr = heap.btree_addr;
let mut blob = heap.blob;
blob.extend_from_slice(&dense_v2_btree(
5,
4 + heap_id_length,
&name_records,
name_bt_addr,
)?);
let link_info_message = if track_order {
// Type 6 records: creation order(8) + heap_id, sorted by order.
let mut by_order: Vec<(u64, usize)> = links
.iter()
.enumerate()
.map(|(i, l)| (l.creation_order.unwrap_or(i as u64), i))
.collect();
by_order.sort_unstable();
let order_records: Vec<Vec<u8>> = by_order
.iter()
.map(|&(order, i)| {
let mut rec = order.to_le_bytes().to_vec();
rec.extend_from_slice(&heap.heap_ids[i]);
rec
})
.collect();
let order_bt_addr = base_address + blob.len() as u64;
blob.extend_from_slice(&dense_v2_btree(
6,
8 + heap_id_length,
&order_records,
order_bt_addr,
)?);
let next_order = by_order.last().map_or(0, |&(o, _)| o + 1);
serialize_link_info(
Some(next_order),
heap.frhp_addr,
name_bt_addr,
Some(order_bt_addr),
)
} else {
serialize_link_info(None, heap.frhp_addr, name_bt_addr, None)
};
Ok(DenseLinkBlob {
link_info_message,
blob,
})
}
/// Serialize a LinkInfo message (version 0). `max_creation_order` (the next
/// creation order to assign) is present when creation order is tracked, and
/// `btree_corder_addr` when it is indexed; both set flag bits.
fn serialize_link_info(
max_creation_order: Option<u64>,
fh_addr: u64,
btree_name_addr: u64,
btree_corder_addr: Option<u64>,
) -> Vec<u8> {
let mut data = Vec::new();
data.push(0); // version
let mut flags = 0u8;
if max_creation_order.is_some() {
flags |= 0x01; // creation order tracked
}
if btree_corder_addr.is_some() {
flags |= 0x02; // creation order indexed
}
data.push(flags);
if let Some(m) = max_creation_order {
data.extend_from_slice(&m.to_le_bytes());
}
write_offset(&mut data, fh_addr, OFFSET_SIZE);
write_offset(&mut data, btree_name_addr, OFFSET_SIZE);
if let Some(a) = btree_corder_addr {
write_offset(&mut data, a, OFFSET_SIZE);
}
data
}
fn encode_managed_id(offset: u64, length: u64, max_heap_size: u16, id_length: u16) -> Vec<u8> {
let mut id = vec![0u8; id_length as usize];
id[0] = 0x00; // type = 0 (managed)
let combined = offset | (length << max_heap_size);
let payload_len = (id_length as usize) - 1;
for i in 0..payload_len.min(8) {
id[1 + i] = ((combined >> (i * 8)) & 0xFF) as u8;
}
id
}
/// 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(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
}
// ---- VDS helpers ----
/// Serialize VDS mappings for storage in a global heap object.
///
/// Delegates to `data_layout_write::serialize_vds_mappings` (the canonical
/// implementation with full version/external-file handling), then appends a
/// trailing 4-byte Jenkins lookup3 checksum that parsers skip after consuming
/// all `nused` entries.
pub(crate) fn serialize_vds_mappings(mappings: &[VdsMapping]) -> Vec<u8> {
let mut buf = crate::data_layout_write::serialize_vds_mappings(mappings, 8);
let cksum = crate::checksum::jenkins_lookup3(&buf);
buf.extend_from_slice(&cksum.to_le_bytes());
buf
}
/// Build a minimal global heap collection containing a single object.
///
/// Returns the serialized collection bytes. The object index is always 1.
///
/// Global heap collection layout:
/// ```text
/// "GCOL"(4) · version(1) · reserved(3) · collection_size(8)
/// · [index(2) · ref_count(2) · reserved(4) · object_size(8) · data · padding]
/// · free-space-marker(2)
/// ```
pub(crate) fn build_global_heap_collection(object_data: &[u8]) -> Vec<u8> {
let ls = LENGTH_SIZE as usize;
let header_size = 8 + ls; // sig(4)+ver(1)+rsv(3)+coll_size(ls)
let obj_header_size = 8 + ls; // idx(2)+rc(2)+rsv(4)+obj_size(ls)
let padded_data_len = pad8(object_data.len());
let free_marker_size = 2;
let collection_size = header_size + obj_header_size + padded_data_len + free_marker_size;
let mut buf = Vec::with_capacity(collection_size);
buf.extend_from_slice(b"GCOL");
buf.push(1); // version
buf.extend_from_slice(&[0u8; 3]); // reserved
buf.extend_from_slice(&(collection_size as u64).to_le_bytes()); // collection_size
// Object 1
buf.extend_from_slice(&1u16.to_le_bytes()); // index
buf.extend_from_slice(&1u16.to_le_bytes()); // reference count
buf.extend_from_slice(&[0u8; 4]); // reserved
buf.extend_from_slice(&(object_data.len() as u64).to_le_bytes()); // object size
buf.extend_from_slice(object_data);
// Pad object data to 8-byte boundary
let pad = padded_data_len - object_data.len();
buf.extend_from_slice(&vec![0u8; pad]);
// Free space marker
buf.extend_from_slice(&0u16.to_le_bytes());
debug_assert_eq!(buf.len(), collection_size);
buf
}
/// Round up to the next multiple of 8.
fn pad8(x: usize) -> usize {
(x + 7) & !7
}
/// Build a Virtual Dataset object header.
///
/// The layout message for a VDS dataset is:
/// ```text
/// version(1=4) · class(1=3) · global_heap_address(8) · global_heap_index(4)
/// ```
pub(crate) fn build_vds_dataset_oh(
dt: &Datatype,
ds: &Dataspace,
global_heap_addr: u64,
attrs: AttrStorage<'_>,
fill_message: &[u8],
refcount: u32,
) -> Result<Vec<u8>, FormatError> {
let mut w = ObjectHeaderWriter::new();
w.add_message_with_flags(MessageType::Datatype, dt.serialize(), 0x01);
w.add_message(MessageType::Dataspace, ds.serialize(LENGTH_SIZE));
w.add_message_with_flags(MessageType::FillValue, fill_message.to_vec(), 0x01);
// VDS layout message: version=4, class=3, global_heap_address(8), global_heap_index=1(4)
let mut dl = Vec::new();
dl.push(4u8); // version
dl.push(3u8); // class = virtual
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);
attrs.add_to(&mut w);
add_refcount(&mut w, refcount);
w.serialize()
}
fn write_offset(buf: &mut Vec<u8>, val: u64, offset_size: u8) {
match offset_size {
2 => buf.extend_from_slice(&(val as u16).to_le_bytes()),
4 => buf.extend_from_slice(&(val as u32).to_le_bytes()),
8 => buf.extend_from_slice(&val.to_le_bytes()),
_ => {}
}
}
fn write_length(buf: &mut Vec<u8>, val: u64, length_size: u8) {
write_offset(buf, val, length_size);
}
fn write_undef_offset(buf: &mut Vec<u8>, offset_size: u8) {
for _ in 0..offset_size {
buf.push(0xFF);
}
}
// ---- FileWriter ----
/// The main file creation API.
///
/// Groups nest to any depth: a name may be a path (`"a/b/x"`), and missing
/// intermediate groups are created, as h5py does; groups also nest through
/// [`GroupBuilder::add_group`]. See [`GroupBuilder`] for how names and
/// repeated groups are handled, and for soft, hard and external links.
pub struct FileWriter {
/// The root group's contents (its name is unused).
root: GroupBuilder,
/// 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`.
alignment_threshold: usize,
/// Global alignment boundary in bytes (0 = disabled).
alignment_bytes: usize,
/// File space page size. When set, the file uses libhdf5's paged
/// file-space strategy (a File Space Info message in the superblock
/// extension).
page_size: Option<u32>,
}
impl Default for FileWriter {
fn default() -> Self {
Self::new()
}
}
/// A dataset ready for layout.
struct DsFlat {
dt: Datatype,
ds: Dataspace,
raw: Vec<u8>,
attrs: Vec<AttributeMessage>,
chunk_options: ChunkOptions,
maxshape: Option<Vec<u64>>,
/// Serialized Fill Value message.
fill_message: Vec<u8>,
compact: bool,
alignment: usize,
/// VDS source mappings (set for Virtual datasets).
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,
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();
let raw = if is_vds {
// VDS datasets have no raw data stored in this file.
db.data.unwrap_or_default()
} else {
db.data.ok_or(FormatError::DatasetMissingData)?
};
let max_dimensions = db.maxshape.clone();
let dspace = Dataspace {
space_type: if shape.is_empty() {
DataspaceType::Scalar
} else {
DataspaceType::Simple
},
rank: shape.len() as u8,
dimensions: shape,
max_dimensions,
};
let mut attrs = Vec::new();
for (n, v) in &db.attrs {
attrs.push(build_attr_message(n, v));
}
#[cfg(feature = "provenance")]
if let Some(ref prov) = db.provenance {
let p = crate::provenance::Provenance {
creator: prov.creator.clone(),
timestamp: prov.timestamp.clone(),
source: prov.source.clone(),
};
// The provenance attributes replace any the caller set by hand.
let prov = p.build_attrs(&raw);
attrs.retain(|a| prov.iter().all(|b| b.name != a.name));
attrs.extend(prov);
}
let fill_message = fill_value_message(db.fill_time, db.fill_value.as_deref(), &dt)?;
Ok(DsFlat {
dt,
ds: dspace,
raw,
attrs,
chunk_options: db.chunk_options,
maxshape: db.maxshape,
fill_message,
compact: db.compact,
alignment: db.alignment,
virtual_sources: db.virtual_sources,
refcount,
track_order,
})
}
/// A group ready for layout.
struct GrpFlat {
attrs: Vec<AttributeMessage>,
links: Vec<writer_tree::Link>,
track_order: bool,
refcount: u32,
}
impl GrpFlat {
fn link_messages(&self, group_addrs: &[u64], ds_addrs: &[u64]) -> Vec<LinkMessage> {
self.links
.iter()
.map(|l| link_message(l, group_addrs, ds_addrs))
.collect()
}
}
impl FileWriter {
pub fn new() -> Self {
Self {
root: GroupBuilder::new("/"),
track_order: false,
alignment_threshold: 0,
alignment_bytes: 0,
page_size: None,
}
}
/// Set global file alignment: datasets with raw data >= `threshold` bytes
/// will have their data aligned to `bytes` boundary.
///
/// For example, `.alignment(1, 4096)` aligns all datasets to 4KB pages.
pub fn alignment(&mut self, threshold: usize, bytes: usize) -> &mut Self {
self.alignment_threshold = threshold;
self.alignment_bytes = bytes;
self
}
/// Write the file with libhdf5's *paged* file-space strategy and the given
/// page size, as `H5Pset_file_space_strategy(H5F_FSPACE_STRATEGY_PAGE)` +
/// `H5Pset_file_space_page_size` (h5py: `fs_strategy="page"`,
/// `fs_page_size=...`) do: a v3 superblock with an extension holding a
/// File Space Info message, and the file padded to a whole number of
/// pages. Readers with a page buffer can then fetch metadata page by page.
///
/// `page_size` must be between 512 bytes and 1 GiB (libhdf5's limits);
/// [`Self::finish`] fails otherwise. This used to write a "version 4"
/// superblock, which does not exist and no HDF5 library can open.
pub fn with_page_size(&mut self, page_size: u32) -> &mut Self {
self.page_size = Some(page_size);
self
}
/// Track (and index) creation order — of links and attributes in every
/// group that does not set its own [`GroupBuilder::track_order`], the
/// root included, and of attributes on every dataset that does not set
/// 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
}
/// Start a group. The builder is detached: fill it, then pass
/// `finish()`'s result to [`Self::add_group`]. `name` may be a path
/// (`"a/b"`); missing intermediate groups are created.
pub fn create_group(&mut self, name: &str) -> GroupBuilder {
GroupBuilder::new(name)
}
/// Add a finished group to the root group.
pub fn add_group(&mut self, group: FinishedGroup) {
self.root.add_group(group);
}
/// Create a dataset. `name` may be a path (`"a/b/x"`, or `"/a/b/x"`);
/// missing intermediate groups are created.
pub fn create_dataset(&mut self, name: &str) -> &mut DatasetBuilder {
self.root.create_dataset(name)
}
pub fn set_root_attr(&mut self, name: &str, value: AttrValue) {
self.root.set_attr(name, value);
}
/// Add a soft link `name` (a path from the root) to `target`. See
/// [`GroupBuilder::add_soft_link`].
pub fn add_soft_link(&mut self, name: &str, target: &str) -> &mut Self {
self.root.add_soft_link(name, target);
self
}
/// Add another hard link `name` (a path from the root) to the object at
/// `target`. See [`GroupBuilder::add_hard_link`].
pub fn add_hard_link(&mut self, name: &str, target: &str) -> &mut Self {
self.root.add_hard_link(name, target);
self
}
/// Add an external link `name` (a path from the root) to `target_path`
/// in the file `target_file`.
pub fn add_external_link(
&mut self,
name: &str,
target_file: &str,
target_path: &str,
) -> &mut Self {
self.root.add_external_link(name, target_file, target_path);
self
}
pub fn finish(self) -> Result<Vec<u8>, FormatError> {
let page_size = self.page_size;
if let Some(ps) = page_size
&& !(MIN_FILE_SPACE_PAGE_SIZE..=MAX_FILE_SPACE_PAGE_SIZE).contains(&ps)
{
return Err(FormatError::SerializationError(format!(
"file space page size {ps} is outside libhdf5's \
{MIN_FILE_SPACE_PAGE_SIZE}..={MAX_FILE_SPACE_PAGE_SIZE} bytes"
)));
}
// The group tree, in layout order: groups depth-first from the root,
// then every group's datasets in the same order.
let tree = writer_tree::build(self.root, self.track_order)?;
let all_ds: Vec<DsFlat> = tree
.datasets
.into_iter()
.map(|(db, refcount)| flatten_ds(db, refcount, self.track_order))
.collect::<Result<_, _>>()?;
let groups: Vec<GrpFlat> = tree
.groups
.into_iter()
.map(|g| GrpFlat {
attrs: g
.attrs
.iter()
.map(|(n, v)| build_attr_message(n, v))
.collect(),
links: g.links,
track_order: g.track_order,
refcount: g.refcount,
})
.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);
let ds_attrs = all_ds.iter().flat_map(|d| &d.attrs);
for a in group_attrs.chain(ds_attrs) {
a.datatype.check_encodable()?;
}
for d in &all_ds {
d.dt.check_encodable()?;
}
let is_vds: Vec<bool> = all_ds.iter().map(|d| d.virtual_sources.is_some()).collect();
let is_chunked: Vec<bool> = all_ds
.iter()
.enumerate()
.map(|(i, d)| {
// Only a dataset that can grow needs chunks; a maxshape equal
// to the shape is as fixed as no maxshape at all.
let resizable = d.maxshape.as_ref().is_some_and(|m| *m != d.ds.dimensions);
!is_vds[i] && (d.chunk_options.is_chunked() || resizable)
})
.collect();
// Determine which datasets use compact storage
let is_compact: Vec<bool> = all_ds
.iter()
.enumerate()
.map(|(i, d)| {
!is_vds[i] && !is_chunked[i] && d.compact && d.raw.len() <= MAX_COMPACT_DATA_SIZE
})
.collect();
let group_dense: Vec<bool> = groups
.iter()
.map(|g| g.attrs.len() > DENSE_ATTR_THRESHOLD)
.collect();
let ds_dense: Vec<bool> = all_ds
.iter()
.map(|d| d.attrs.len() > DENSE_ATTR_THRESHOLD)
.collect();
// Dense link decision: a group with more than the compact threshold of
// links stores them in a fractal heap + v2 B-tree instead of inline.
let group_links_dense: Vec<bool> = groups
.iter()
.map(|g| g.links.len() > DENSE_LINK_THRESHOLD)
.collect();
// Pass 1: compute OH sizes with dummy addresses. Link messages and
// the Link Info message are the same size whatever the addresses.
let group_oh_sizes: Vec<usize> = groups
.iter()
.enumerate()
.map(|(gi, g)| {
let dummy_links = g.link_messages(&[], &[]);
let attr_blob = group_dense[gi]
.then(|| build_dense_attrs(&g.attrs, 0, g.track_order))
.transpose()?;
let li = if group_links_dense[gi] {
serialize_link_info(
g.track_order.then_some(0),
0,
0,
g.track_order.then_some(0),
)
} else {
compact_link_info(g.track_order, g.links.len())
};
build_group_oh(
&dummy_links,
&li,
group_links_dense[gi],
AttrStorage {
attrs: &g.attrs,
dense: attr_blob.as_ref(),
track_order: g.track_order,
},
g.refcount,
)
.map(|oh| oh.len())
})
.collect::<Result<_, _>>()?;
struct DataBlob {
data: Vec<u8>,
oh_bytes: Vec<u8>,
/// Cached compressed chunks for chunked datasets; reused in Pass 2
/// to avoid re-compressing the same data.
precompressed: Option<PrecompressedChunks>,
}
let mut dummy_blobs: Vec<DataBlob> = Vec::new();
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, d.track_order))
.transpose()?;
if is_vds[i] {
// VDS: dummy OH with address 0 to get the OH size. The global
// heap blob will be placed after the OHs in pass 2.
let oh = build_vds_dataset_oh(
&d.dt,
&d.ds,
0, // dummy address
AttrStorage {
attrs: &d.attrs,
dense: dense_blob.as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
// Global heap blob size is address-independent; compute it now
// so pass 2 can place it correctly.
let vds_mappings = d.virtual_sources.as_deref().unwrap_or(&[]);
let gcol_bytes =
build_global_heap_collection(&serialize_vds_mappings(vds_mappings));
dummy_blobs.push(DataBlob {
data: gcol_bytes, // store heap blob here temporarily
oh_bytes: oh,
precompressed: None,
});
} else if is_chunked[i] {
let elem_size = d.dt.type_size() as usize;
let chunk_dims = d
.chunk_options
.resolve_chunk_dims_for(&d.ds.dimensions, elem_size);
// Compress once in Pass 1; cache the result so Pass 2 can skip
// re-compression and just rebuild the index with real addresses.
let pre = precompress_chunks(
&d.raw,
&d.ds.dimensions,
&chunk_dims,
elem_size,
&d.chunk_options,
)?;
let result = build_chunked_data_from_precompressed(
&pre,
dummy_cursor,
d.maxshape.as_deref(),
)?;
dummy_cursor += result.data_bytes.len() as u64;
let oh = build_chunked_dataset_oh(
&d.dt,
&d.ds,
&result.layout_message,
result.pipeline_message.as_deref(),
AttrStorage {
attrs: &d.attrs,
dense: dense_blob.as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
dummy_blobs.push(DataBlob {
data: result.data_bytes,
oh_bytes: oh,
precompressed: Some(pre),
});
} else if is_compact[i] {
let oh = build_compact_dataset_oh(
&d.dt,
&d.ds,
&d.raw,
AttrStorage {
attrs: &d.attrs,
dense: dense_blob.as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
dummy_blobs.push(DataBlob {
data: vec![],
oh_bytes: oh,
precompressed: None,
});
} else {
let oh = build_dataset_oh(
&d.dt,
&d.ds,
0,
d.raw.len() as u64,
AttrStorage {
attrs: &d.attrs,
dense: dense_blob.as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
dummy_blobs.push(DataBlob {
data: vec![],
oh_bytes: oh,
precompressed: None,
});
}
}
let actual_ds_oh_sizes: Vec<usize> = dummy_blobs.iter().map(|b| b.oh_bytes.len()).collect();
// Pass 2: compute real addresses
// A paged file carries its File Space Info in a superblock extension
// object header, placed right after the superblock.
let sb_ext = page_size
.map(build_paged_superblock_extension)
.transpose()?;
let superblock_size = SUPERBLOCK_SIZE + sb_ext.as_ref().map_or(0, Vec::len);
let mut cursor2 = superblock_size;
// Each group (the root first) is laid out as: object header, then (if
// dense) its link blob, then (if dense) its attribute blob. Link blobs
// are sized with dummy target addresses here — link message size is
// address-independent — and rebuilt with real addresses when written.
let mut group_link_blob_addrs: Vec<Option<u64>> = Vec::new();
let mut group_dense_blobs: Vec<Option<DenseAttrBlob>> = Vec::new();
let mut group_addrs2: Vec<u64> = Vec::with_capacity(groups.len());
for (gi, g) in groups.iter().enumerate() {
group_addrs2.push(cursor2 as u64);
cursor2 += group_oh_sizes[gi];
if group_links_dense[gi] {
let blob_addr = cursor2 as u64;
let dummy = g.link_messages(&[], &[]);
cursor2 += build_dense_links(&dummy, blob_addr, g.track_order)?
.blob
.len();
group_link_blob_addrs.push(Some(blob_addr));
} else {
group_link_blob_addrs.push(None);
}
if group_dense[gi] {
let blob = build_dense_attrs(&g.attrs, cursor2 as u64, g.track_order)?;
cursor2 += blob.blob.len();
group_dense_blobs.push(Some(blob));
} else {
group_dense_blobs.push(None);
}
}
let root_group_addr = group_addrs2[0];
let mut ds_dense_blobs: Vec<Option<DenseAttrBlob>> = Vec::new();
let ds_oh_addrs2: Vec<u64> = actual_ds_oh_sizes
.iter()
.enumerate()
.map(|(i, &sz)| {
let addr = cursor2 as u64;
cursor2 += sz;
if ds_dense[i] {
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 {
ds_dense_blobs.push(None);
}
Ok(addr)
})
.collect::<Result<_, FormatError>>()?;
let mut ds_blobs2: Vec<DataBlob> = Vec::new();
let global_align_threshold = self.alignment_threshold;
let global_align_bytes = self.alignment_bytes;
for (i, d) in all_ds.iter().enumerate() {
if is_vds[i] {
// VDS: place the global heap collection right after the OHs,
// then rebuild the OH with the real heap address.
let gcol_bytes = &dummy_blobs[i].data; // pre-computed in pass 1
let heap_addr = cursor2 as u64;
cursor2 += gcol_bytes.len();
let oh = build_vds_dataset_oh(
&d.dt,
&d.ds,
heap_addr,
AttrStorage {
attrs: &d.attrs,
dense: ds_dense_blobs[i].as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
ds_blobs2.push(DataBlob {
data: gcol_bytes.clone(),
oh_bytes: oh,
precompressed: None,
});
} else if is_chunked[i] {
let base_address = cursor2 as u64;
// Reuse precompressed chunks from Pass 1 — avoids re-compressing
// the same data a second time.
let result = build_chunked_data_from_precompressed(
dummy_blobs[i]
.precompressed
.as_ref()
.expect("chunked dataset missing precompressed cache"),
base_address,
d.maxshape.as_deref(),
)?;
cursor2 += result.data_bytes.len();
let oh = build_chunked_dataset_oh(
&d.dt,
&d.ds,
&result.layout_message,
result.pipeline_message.as_deref(),
AttrStorage {
attrs: &d.attrs,
dense: ds_dense_blobs[i].as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
ds_blobs2.push(DataBlob {
data: result.data_bytes,
oh_bytes: oh,
precompressed: None,
});
} else if is_compact[i] {
// Compact: data is inline in the object header, no external blob
let oh = build_compact_dataset_oh(
&d.dt,
&d.ds,
&d.raw,
AttrStorage {
attrs: &d.attrs,
dense: ds_dense_blobs[i].as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
ds_blobs2.push(DataBlob {
data: vec![],
oh_bytes: oh,
precompressed: None,
});
} else {
// Determine alignment: per-dataset overrides global
let align = if d.alignment > 0 {
d.alignment
} else if global_align_bytes > 0 && d.raw.len() >= global_align_threshold {
global_align_bytes
} else {
8 // default: 8-byte alignment for zero-copy read support
};
let padding = (align - (cursor2 % align)) % align;
cursor2 += padding;
let oh = build_dataset_oh(
&d.dt,
&d.ds,
cursor2 as u64,
d.raw.len() as u64,
AttrStorage {
attrs: &d.attrs,
dense: ds_dense_blobs[i].as_ref(),
track_order: d.track_order,
},
&d.fill_message,
d.refcount,
)?;
let mut data = vec![0u8; padding];
data.extend_from_slice(&d.raw);
cursor2 += d.raw.len();
ds_blobs2.push(DataBlob {
data,
oh_bytes: oh,
precompressed: None,
});
}
}
let actual_ds_oh_sizes2: Vec<usize> = ds_blobs2.iter().map(|b| b.oh_bytes.len()).collect();
debug_assert_eq!(actual_ds_oh_sizes, actual_ds_oh_sizes2);
// libhdf5 ends a paged file on a page boundary.
let data_end = cursor2;
if let Some(ps) = page_size {
cursor2 = cursor2.next_multiple_of(ps as usize);
}
let eof_addr2 = cursor2 as u64;
let mut buf = Vec::with_capacity(cursor2);
let sb = Superblock {
version: 3,
offset_size: OFFSET_SIZE,
length_size: LENGTH_SIZE,
base_address: 0,
eof_address: eof_addr2,
root_group_address: root_group_addr,
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(if sb_ext.is_some() {
SUPERBLOCK_SIZE as u64
} else {
u64::MAX
}),
checksum: None,
page_size: None,
};
buf.extend_from_slice(&sb.serialize());
if let Some(ref ext) = sb_ext {
buf.extend_from_slice(ext);
}
// Group OHs + dense blobs (link blob, then attr blob, matching pass 2)
for (gi, g) in groups.iter().enumerate() {
let links = g.link_messages(&group_addrs2, &ds_oh_addrs2);
// Rebuild the link blob with real target addresses (same size as
// the dummy used for layout); its LinkInfo goes in the OH.
let link_blob = group_link_blob_addrs[gi]
.map(|addr| build_dense_links(&links, addr, g.track_order))
.transpose()?;
let li = match &link_blob {
Some(b) => b.link_info_message.clone(),
None => compact_link_info(g.track_order, links.len()),
};
let oh = build_group_oh(
&links,
&li,
link_blob.is_some(),
AttrStorage {
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]);
debug_assert_eq!(buf.len() as u64, group_addrs2[gi]);
buf.extend_from_slice(&oh);
if let Some(ref b) = link_blob {
buf.extend_from_slice(&b.blob);
}
if let Some(ref blob) = group_dense_blobs[gi] {
buf.extend_from_slice(&blob.blob);
}
}
// Dataset OHs + dense blobs
for (i, blob) in ds_blobs2.iter().enumerate() {
buf.extend_from_slice(&blob.oh_bytes);
if let Some(ref dense) = ds_dense_blobs[i] {
buf.extend_from_slice(&dense.blob);
}
}
// Data
for blob in &ds_blobs2 {
buf.extend_from_slice(&blob.data);
}
debug_assert_eq!(buf.len(), data_end);
buf.resize(cursor2, 0);
Ok(buf)
}
}
// ---- Independent parallel dataset creation ----
/// Builder that creates datasets without locking the file header.
///
/// Each `IndependentDatasetBuilder` accumulates its own [`MetadataBlock`]
/// independently. On [`IndependentDatasetBuilder::finish`], the block is
/// returned for later merging.
///
/// Thread-safety: each thread should own its own builder instance.
pub struct IndependentDatasetBuilder {
block: MetadataBlock,
}
impl IndependentDatasetBuilder {
/// Create a new independent builder with the given creator id.
pub fn new(creator_id: u32) -> Self {
Self {
block: MetadataBlock::new(creator_id),
}
}
/// Add a dataset specification to this builder.
pub fn add_dataset(&mut self, meta: DatasetMetadata) {
self.block.add_dataset(meta);
}
/// Consume the builder and return the metadata block.
pub fn finish(self) -> MetadataBlock {
self.block
}
}
/// Finalize multiple independently-created metadata blocks into a complete HDF5 file.
///
/// This implements the write-ahead approach: each block's data is laid out
/// sequentially, then the index table (root group with links) is written last
/// to point at all the dataset object headers.
pub fn finalize_parallel(blocks: Vec<MetadataBlock>) -> Result<Vec<u8>, FormatError> {
let index = MetadataIndex::merge_blocks(&blocks)?;
finalize_from_index(index)
}
/// Build a complete HDF5 file from a merged MetadataIndex.
fn finalize_from_index(index: MetadataIndex) -> Result<Vec<u8>, FormatError> {
// Convert DatasetMetadata into the internal DsFlat representation and
// delegate to the same two-pass algorithm used by FileWriter.
let mut fw = FileWriter::new();
for ds_meta in &index.datasets {
let db = fw.create_dataset(&ds_meta.name);
// Set the datatype and raw data directly via internal fields
db.datatype = Some(ds_meta.datatype.clone());
db.shape = Some(ds_meta.dataspace.dimensions.clone());
db.maxshape = ds_meta.maxshape.clone();
db.data = Some(ds_meta.raw_data.clone());
db.chunk_options = ds_meta.chunk_options.clone();
for (name, val) in &ds_meta.attrs {
db.set_attr(name, val.clone());
}
}
fw.finish()
}
#[cfg(test)]
mod tests {
use super::*;
use crate::group_v2::resolve_path_any;
use crate::object_header::ObjectHeader;
use crate::signature;
fn parse_file(bytes: &[u8]) -> (Superblock, ObjectHeader) {
let sig = signature::find_signature(bytes).unwrap();
let sb = Superblock::parse(bytes, sig).unwrap();
let oh = ObjectHeader::parse(
bytes,
sb.root_group_address as usize,
sb.offset_size,
sb.length_size,
)
.unwrap();
(sb, oh)
}
fn read_dataset_f64(bytes: &[u8], path: &str) -> Vec<f64> {
let sig = signature::find_signature(bytes).unwrap();
let sb = Superblock::parse(bytes, sig).unwrap();
let addr = resolve_path_any(bytes, &sb, path).unwrap();
let hdr =
ObjectHeader::parse(bytes, addr as usize, sb.offset_size, sb.length_size).unwrap();
let dt_data = &hdr
.messages
.iter()
.find(|m| m.msg_type == MessageType::Datatype)
.unwrap()
.data;
let ds_data = &hdr
.messages
.iter()
.find(|m| m.msg_type == MessageType::Dataspace)
.unwrap()
.data;
let dl_data = &hdr
.messages
.iter()
.find(|m| m.msg_type == MessageType::DataLayout)
.unwrap()
.data;
let (dt, _) = Datatype::parse(dt_data).unwrap();
let ds = Dataspace::parse(ds_data, sb.length_size).unwrap();
let dl =
crate::data_layout::DataLayout::parse(dl_data, sb.offset_size, sb.length_size).unwrap();
let raw = crate::data_read::read_raw_data(bytes, &dl, &ds, &dt).unwrap();
crate::data_read::read_as_f64(&raw, &dt).unwrap()
}
#[test]
fn empty_file_root_group_only() {
let fw = FileWriter::new();
let bytes = fw.finish().unwrap();
let (sb, oh) = parse_file(&bytes);
assert_eq!(sb.version, 3);
assert_eq!(oh.version, 2);
}
#[test]
fn file_with_f64_dataset() {
let mut fw = FileWriter::new();
fw.create_dataset("data").with_f64_data(&[1.0, 2.0, 3.0]);
let bytes = fw.finish().unwrap();
assert_eq!(read_dataset_f64(&bytes, "data"), vec![1.0, 2.0, 3.0]);
}
#[test]
fn file_with_dataset_attrs() {
let mut fw = FileWriter::new();
fw.create_dataset("data")
.with_f64_data(&[1.0, 2.0])
.set_attr("scale", AttrValue::F64(0.5));
let bytes = fw.finish().unwrap();
assert_eq!(read_dataset_f64(&bytes, "data"), vec![1.0, 2.0]);
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let addr = resolve_path_any(&bytes, &sb, "data").unwrap();
let hdr =
ObjectHeader::parse(&bytes, addr as usize, sb.offset_size, sb.length_size).unwrap();
let attrs = crate::attribute::extract_attributes(&hdr, sb.length_size).unwrap();
assert_eq!(attrs.len(), 1);
assert_eq!(attrs[0].name, "scale");
}
#[test]
fn file_with_group_and_dataset() {
let mut fw = FileWriter::new();
let mut gb = fw.create_group("grp");
gb.create_dataset("vals").with_f64_data(&[10.0, 20.0]);
fw.add_group(gb.finish());
let bytes = fw.finish().unwrap();
assert_eq!(read_dataset_f64(&bytes, "grp/vals"), vec![10.0, 20.0]);
}
#[test]
fn file_with_root_attr() {
let mut fw = FileWriter::new();
fw.set_root_attr("version", AttrValue::I64(42));
let bytes = fw.finish().unwrap();
let (sb, oh) = parse_file(&bytes);
let attrs = crate::attribute::extract_attributes(&oh, sb.length_size).unwrap();
assert_eq!(attrs[0].name, "version");
}
#[test]
fn dense_attrs_self_roundtrip() {
let mut fw = FileWriter::new();
let ds = fw.create_dataset("data");
ds.with_f64_data(&[1.0, 2.0, 3.0]);
for i in 0..20 {
ds.set_attr(&format!("attr_{i:03}"), AttrValue::F64(i as f64 * 1.5));
}
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let addr = resolve_path_any(&bytes, &sb, "data").unwrap();
let hdr =
ObjectHeader::parse(&bytes, addr as usize, sb.offset_size, sb.length_size).unwrap();
let attrs =
crate::attribute::extract_attributes_full(&bytes, &hdr, sb.offset_size, sb.length_size)
.unwrap();
assert_eq!(attrs.len(), 20);
for i in 0..20 {
let attr = attrs
.iter()
.find(|a| a.name == format!("attr_{i:03}"))
.unwrap();
let v = attr.read_as_f64().unwrap();
assert!((v[0] - i as f64 * 1.5).abs() < 1e-10);
}
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();
fw.create_dataset("dummy").with_f64_data(&[0.0]);
for i in 0..15 {
fw.set_root_attr(&format!("root_{i:02}"), AttrValue::F64(i as f64 * 2.0));
}
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let oh = ObjectHeader::parse(
&bytes,
sb.root_group_address as usize,
sb.offset_size,
sb.length_size,
)
.unwrap();
let attrs =
crate::attribute::extract_attributes_full(&bytes, &oh, sb.offset_size, sb.length_size)
.unwrap();
assert_eq!(attrs.len(), 15);
}
#[test]
fn inline_attrs_below_threshold() {
let mut fw = FileWriter::new();
let ds = fw.create_dataset("data");
ds.with_f64_data(&[1.0]);
for i in 0..5 {
ds.set_attr(&format!("a{i}"), AttrValue::F64(i as f64));
}
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let addr = resolve_path_any(&bytes, &sb, "data").unwrap();
let hdr =
ObjectHeader::parse(&bytes, addr as usize, sb.offset_size, sb.length_size).unwrap();
assert!(
!hdr.messages
.iter()
.any(|m| m.msg_type == MessageType::AttributeInfo)
);
let attrs = crate::attribute::extract_attributes(&hdr, sb.length_size).unwrap();
assert_eq!(attrs.len(), 5);
}
#[test]
fn encode_decode_managed_id_roundtrip() {
let id = encode_managed_id(100, 42, 40, 8);
let fh = crate::fractal_heap::FractalHeapHeader {
heap_id_length: 8,
io_filter_encoded_length: 0,
max_managed_object_size: 1024,
table_width: 4,
starting_block_size: 4096,
max_direct_block_size: 65536,
max_heap_size: 40,
starting_row_of_indirect_blocks: 1,
root_block_address: 0,
current_rows_in_root_indirect_block: 0,
managed_objects_count: 0,
huge_btree_address: u64::MAX,
filter_pipeline: None,
root_direct_block_filtered_size: 0,
root_direct_block_filter_mask: 0,
offset_size: 8,
length_size: 8,
};
let (off, len) = fh.decode_managed_id(&id).unwrap();
assert_eq!(off, 100);
assert_eq!(len, 42);
}
#[test]
fn finalize_parallel_basic() {
use crate::chunked_write::ChunkOptions;
use crate::metadata_index::{MetadataBlock, build_dataset_metadata};
use crate::type_builders::make_f64_type;
let mut b0 = MetadataBlock::new(0);
let data_a: Vec<u8> = [1.0f64, 2.0, 3.0]
.iter()
.flat_map(|v| v.to_le_bytes())
.collect();
b0.add_dataset(build_dataset_metadata(
"alpha",
make_f64_type(),
vec![3],
data_a,
ChunkOptions::default(),
None,
vec![],
));
let mut b1 = MetadataBlock::new(1);
let data_b: Vec<u8> = [10.0f64, 20.0]
.iter()
.flat_map(|v| v.to_le_bytes())
.collect();
b1.add_dataset(build_dataset_metadata(
"beta",
make_f64_type(),
vec![2],
data_b,
ChunkOptions::default(),
None,
vec![],
));
let bytes = finalize_parallel(vec![b0, b1]).unwrap();
assert_eq!(read_dataset_f64(&bytes, "alpha"), vec![1.0, 2.0, 3.0]);
assert_eq!(read_dataset_f64(&bytes, "beta"), vec![10.0, 20.0]);
}
#[test]
fn finalize_parallel_duplicate_error() {
use crate::chunked_write::ChunkOptions;
use crate::metadata_index::{MetadataBlock, build_dataset_metadata};
use crate::type_builders::make_f64_type;
let mut b0 = MetadataBlock::new(0);
b0.add_dataset(build_dataset_metadata(
"dup",
make_f64_type(),
vec![1],
vec![0u8; 8],
ChunkOptions::default(),
None,
vec![],
));
let mut b1 = MetadataBlock::new(1);
b1.add_dataset(build_dataset_metadata(
"dup",
make_f64_type(),
vec![1],
vec![0u8; 8],
ChunkOptions::default(),
None,
vec![],
));
let err = finalize_parallel(vec![b0, b1]).unwrap_err();
assert!(matches!(err, FormatError::DuplicateDatasetName(_)));
}
// ---- Virtual Dataset (VDS) round-trip tests ----
/// Serialize an H5S ALL selection (type=3, version=1, 16 bytes).
fn sel_all() -> Vec<u8> {
vec![
3, 0, 0, 0, // type = ALL
1, 0, 0, 0, // version
0, 0, 0, 0, // reserved
0, 0, 0, 0, // length (unused for ALL)
]
}
/// Serialize an H5S HYPER selection (version 3, rank 1, enc_size 2).
/// Encodes start=`start`, stride=1, count=1, block=`block`.
fn sel_hyper_1d(start: u16, block: u16) -> Vec<u8> {
let mut v = vec![
2, 0, 0, 0, // type = HYPER
3, 0, 0, 0, // version 3
0x01, // flags = regular
0x02, // enc_size = 2 (u16 per coordinate)
1, 0, 0, 0, // rank = 1
];
v.extend_from_slice(&start.to_le_bytes()); // start
v.extend_from_slice(&1u16.to_le_bytes()); // stride
v.extend_from_slice(&1u16.to_le_bytes()); // count
v.extend_from_slice(&block.to_le_bytes()); // block
v
}
#[test]
fn vds_write_read_virtual_layout() {
use crate::data_layout::DataLayout;
// A virtual dataset /vds of shape [8] backed by two same-file sources:
// /src_a maps to virtual[0:4] and /src_b maps to virtual[4:8].
let mapping_a = VdsMapping {
source_file: ".".into(),
source_dataset: "src_a".into(),
source_selection: sel_all(),
virtual_selection: sel_hyper_1d(0, 4),
};
let mapping_b = VdsMapping {
source_file: ".".into(),
source_dataset: "src_b".into(),
source_selection: sel_all(),
virtual_selection: sel_hyper_1d(4, 4),
};
let mut fw = FileWriter::new();
// Source datasets (real data in this file)
fw.create_dataset("src_a")
.with_f64_data(&[1.0, 2.0, 3.0, 4.0]);
fw.create_dataset("src_b")
.with_f64_data(&[5.0, 6.0, 7.0, 8.0]);
// Virtual dataset
fw.create_dataset("vds")
.with_shape(&[8])
.with_f64_data(&[]) // shape hint; raw data is ignored for VDS
.with_virtual_sources(vec![mapping_a, mapping_b]);
let bytes = fw.finish().unwrap();
// Verify the virtual dataset resolves to DataLayout::Virtual
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let vds_addr = resolve_path_any(&bytes, &sb, "vds").unwrap();
let hdr =
ObjectHeader::parse(&bytes, vds_addr as usize, sb.offset_size, sb.length_size).unwrap();
let dl_data = &hdr
.messages
.iter()
.find(|m| m.msg_type == MessageType::DataLayout)
.unwrap()
.data;
let mut layout = DataLayout::parse(dl_data, sb.offset_size, sb.length_size).unwrap();
// Before resolution, mappings field is empty.
assert!(
matches!(layout, DataLayout::Virtual { .. }),
"expected Virtual layout, got {layout:?}"
);
// Resolve VDS mappings from the global heap.
layout.resolve_vds_mappings(&bytes, sb.length_size).unwrap();
match &layout {
DataLayout::Virtual { mappings, .. } => {
assert_eq!(mappings.len(), 2, "expected 2 VDS mappings");
assert_eq!(mappings[0].source_file, ".");
assert_eq!(mappings[0].source_dataset, "src_a");
assert_eq!(mappings[1].source_file, ".");
assert_eq!(mappings[1].source_dataset, "src_b");
// Verify the virtual selections cover [0:4] and [4:8].
use crate::selection::Selection;
let (vsel_a, _) =
Selection::decode_serialized(&mappings[0].virtual_selection).unwrap();
let (vsel_b, _) =
Selection::decode_serialized(&mappings[1].virtual_selection).unwrap();
assert_eq!(vsel_a.iter_linear_1d(8).unwrap(), vec![0, 1, 2, 3]);
assert_eq!(vsel_b.iter_linear_1d(8).unwrap(), vec![4, 5, 6, 7]);
}
other => panic!("expected Virtual layout after resolution, got {other:?}"),
}
// Source datasets still readable normally.
assert_eq!(read_dataset_f64(&bytes, "src_a"), vec![1.0, 2.0, 3.0, 4.0]);
assert_eq!(read_dataset_f64(&bytes, "src_b"), vec![5.0, 6.0, 7.0, 8.0]);
}
#[test]
fn vds_external_source_file() {
use crate::data_layout::DataLayout;
// A VDS mapping referencing an external file ("other.h5").
let mapping_ext = VdsMapping {
source_file: "other.h5".into(),
source_dataset: "data".into(),
source_selection: sel_all(),
virtual_selection: sel_all(),
};
let mut fw = FileWriter::new();
fw.create_dataset("ext_vds")
.with_shape(&[10])
.with_f64_data(&[]) // shape hint only
.with_virtual_sources(vec![mapping_ext]);
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let addr = resolve_path_any(&bytes, &sb, "ext_vds").unwrap();
let hdr =
ObjectHeader::parse(&bytes, addr as usize, sb.offset_size, sb.length_size).unwrap();
let dl_data = &hdr
.messages
.iter()
.find(|m| m.msg_type == MessageType::DataLayout)
.unwrap()
.data;
let mut layout = DataLayout::parse(dl_data, sb.offset_size, sb.length_size).unwrap();
layout.resolve_vds_mappings(&bytes, sb.length_size).unwrap();
match &layout {
DataLayout::Virtual { mappings, .. } => {
assert_eq!(mappings.len(), 1);
assert_eq!(mappings[0].source_file, "other.h5");
assert_eq!(mappings[0].source_dataset, "data");
}
other => panic!("expected Virtual, got {other:?}"),
}
}
#[test]
fn vds_empty_mapping_list() {
// Calling with_virtual_sources([]) is silently ignored — the dataset
// falls back to a normal contiguous layout rather than writing an empty VDS.
use crate::data_layout::DataLayout;
let mut fw = FileWriter::new();
fw.create_dataset("empty_vds")
.with_shape(&[0])
.with_f64_data(&[])
.with_virtual_sources(vec![]);
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let addr = resolve_path_any(&bytes, &sb, "empty_vds").unwrap();
let hdr =
ObjectHeader::parse(&bytes, addr as usize, sb.offset_size, sb.length_size).unwrap();
let dl_data = &hdr
.messages
.iter()
.find(|m| m.msg_type == MessageType::DataLayout)
.unwrap()
.data;
let layout = DataLayout::parse(dl_data, sb.offset_size, sb.length_size).unwrap();
// Empty mapping list → no VDS layout; should be Contiguous or Compact.
assert!(
!matches!(layout, DataLayout::Virtual { .. }),
"empty with_virtual_sources should NOT produce a VDS layout, got {layout:?}"
);
}
#[test]
fn external_link_write_roundtrip() {
let mut fw = FileWriter::new();
let mut grp = fw.create_group("sensors");
grp.create_dataset("local_ds").with_f64_data(&[1.0, 2.0]);
grp.add_external_link("remote_temp", "other_file.h5", "/temperature");
fw.add_group(grp.finish());
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
let sensors_addr = resolve_path_any(&bytes, &sb, "sensors").unwrap();
let hdr = ObjectHeader::parse(
&bytes,
sensors_addr as usize,
sb.offset_size,
sb.length_size,
)
.unwrap();
// Find the external LinkMessage directly in the object header.
let ext_link = hdr
.messages
.iter()
.filter(|m| m.msg_type == MessageType::Link)
.filter_map(|m| crate::link_message::LinkMessage::parse(&m.data, sb.offset_size).ok())
.find(|l| l.name == "remote_temp")
.expect("external link 'remote_temp' not found in group OH");
match &ext_link.link_target {
crate::link_message::LinkTarget::External {
filename,
object_path,
} => {
assert_eq!(filename, "other_file.h5");
assert_eq!(object_path, "/temperature");
}
other => panic!("expected External link, got {other:?}"),
}
}
#[test]
fn file_writer_paged_file_uses_v3_superblock_and_fsinfo_extension() {
// This used to write superblock "version 4", which does not exist.
let mut fw = FileWriter::new();
fw.with_page_size(4096);
fw.create_dataset("data").with_f64_data(&[1.0, 2.0]);
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
assert_eq!(sb.version, 3);
assert_eq!(sb.superblock_extension_address, Some(48));
assert_eq!(bytes.len() % 4096, 0);
assert_eq!(sb.eof_address, bytes.len() as u64);
let ext = ObjectHeader::parse(&bytes, 48, 8, 8).unwrap();
let fsinfo = &ext.messages[0];
assert_eq!(fsinfo.msg_type, MessageType::Unknown(0x0017));
// Byte-for-byte what HDF5 2.0 writes for fs_strategy="page",
// fs_page_size=4096.
let mut expected = vec![1u8, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0];
expected.extend_from_slice(&4096u64.to_le_bytes());
expected.extend_from_slice(&[0, 0]);
expected.extend_from_slice(&[0xff; 8]);
assert_eq!(fsinfo.data, expected);
assert_eq!(fsinfo.flags, 0x14);
assert_eq!(read_dataset_f64(&bytes, "data"), vec![1.0, 2.0]);
}
#[test]
fn file_writer_rejects_page_sizes_libhdf5_would() {
for ps in [0u32, 511, MAX_FILE_SPACE_PAGE_SIZE + 1] {
let mut fw = FileWriter::new();
fw.with_page_size(ps);
assert!(fw.finish().is_err(), "page size {ps}");
}
}
#[test]
fn file_writer_default_superblock_is_v3() {
let mut fw = FileWriter::new();
fw.create_dataset("data").with_f64_data(&[1.0, 2.0]);
let bytes = fw.finish().unwrap();
let sig = signature::find_signature(&bytes).unwrap();
let sb = Superblock::parse(&bytes, sig).unwrap();
assert_eq!(sb.version, 3);
assert_eq!(sb.page_size, None);
}
}