rustytorch/crates/data/rtx-data-validation/benches/validation_benchmarks.rs

//! Performance benchmarks for rtx-data-validation
//!
//! This module contains comprehensive benchmarks for all validation operations
//! to measure and track performance characteristics.

use criterion::{BenchmarkId, Criterion, Throughput, black_box, criterion_group, criterion_main};
use std::collections::HashMap;
use std::time::Duration;
use tokio::runtime::Runtime;

use rtx_data_validation::*;

/// Helper function to create sample data for benchmarks
fn create_benchmark_record(id: usize) -> DataRecord {
    let mut fields = HashMap::new();
    fields.insert("id".to_string(), DataValue::Int(id as i64));
    fields.insert(
        "name".to_string(),
        DataValue::String(format!("User {}", id)),
    );
    fields.insert(
        "email".to_string(),
        DataValue::String(format!("user{}@example.com", id)),
    );
    fields.insert("age".to_string(), DataValue::Int(25 + (id % 40) as i64));
    fields.insert(
        "score".to_string(),
        DataValue::Float(50.0 + (id % 50) as f64),
    );
    fields.insert("active".to_string(), DataValue::Bool(id % 2 == 0));

    DataRecord {
        id: format!("record_{}", id),
        timestamp: chrono::Utc::now(),
        fields,
        metadata: HashMap::new(),
    }
}

/// Create multiple benchmark records
fn create_benchmark_records(count: usize) -> Vec<DataRecord> {
    (0..count).map(create_benchmark_record).collect()
}

/// Benchmark basic validation engine performance
fn benchmark_validation_engine(c: &mut Criterion) {
    let rt = Runtime::new().unwrap();

    // Create engine with various rule configurations
    let simple_engine = ValidationEngine::builder()
        .add_rule(ValidationRule::not_null("name"))
        .build()
        .expect("Failed to build simple engine");

    let complex_engine = ValidationEngine::builder()
        .with_statistical_validation(true)
        .add_rule(ValidationRule::not_null("name"))
        .add_rule(ValidationRule::range("age", 0.0, 120.0))
        .add_rule(ValidationRule::email("email"))
        .add_rule(ValidationRule::range("score", 0.0, 100.0))
        .add_rule(ValidationRule::length("name", 2, 100))
        .build()
        .expect("Failed to build complex engine");

    let test_record = create_benchmark_record(1);

    // Benchmark single record validation
    c.bench_function("validation_single_simple", |b| {
        b.to_async(&rt)
            .iter(|| async { black_box(simple_engine.validate(&test_record).await.unwrap()) })
    });

    c.bench_function("validation_single_complex", |b| {
        b.to_async(&rt)
            .iter(|| async { black_box(complex_engine.validate(&test_record).await.unwrap()) })
    });

    // Benchmark batch validation with different sizes
    for size in [10, 100, 1000].iter() {
        let records = create_benchmark_records(*size);

        c.bench_with_input(
            BenchmarkId::new("validation_batch_simple", size),
            &records,
            |b, records| {
                b.to_async(&rt).iter(|| async {
                    black_box(simple_engine.validate_batch(records).await.unwrap())
                })
            },
        );

        c.bench_with_input(
            BenchmarkId::new("validation_batch_complex", size),
            &records,
            |b, records| {
                b.to_async(&rt).iter(|| async {
                    black_box(complex_engine.validate_batch(records).await.unwrap())
                })
            },
        );
    }
}

/// Benchmark validation rules performance
fn benchmark_validation_rules(c: &mut Criterion) {
    let test_values = vec![
        DataValue::String("[email protected]".to_string()),
        DataValue::Int(25),
        DataValue::Float(85.5),
        DataValue::String("Valid Name".to_string()),
    ];

    // Benchmark individual rule types
    let not_null_rule = ValidationRule::not_null("test");
    let range_rule = ValidationRule::range("test", 0.0, 100.0);
    let email_rule = ValidationRule::email("test");
    let length_rule = ValidationRule::length("test", 2, 100);

    c.bench_function("rule_not_null", |b| {
        b.iter(|| {
            for value in &test_values {
                black_box(not_null_rule.validate(value));
            }
        })
    });

    c.bench_function("rule_range", |b| {
        b.iter(|| {
            for value in &test_values {
                black_box(range_rule.validate(value));
            }
        })
    });

    c.bench_function("rule_email", |b| {
        b.iter(|| {
            black_box(email_rule.validate(&test_values[0]));
        })
    });

    c.bench_function("rule_length", |b| {
        b.iter(|| {
            black_box(length_rule.validate(&test_values[3]));
        })
    });

    // Benchmark cross-field validation
    let cross_field_rule = ValidationRule::cross_field(
        "field1",
        "field2",
        crate::rules::CrossFieldOperation::Greater,
    );

    let mut context = HashMap::new();
    context.insert("field2".to_string(), DataValue::Int(20));

    c.bench_function("rule_cross_field", |b| {
        b.iter(|| {
            black_box(cross_field_rule.validate_with_context(&DataValue::Int(30), &context));
        })
    });
}

/// Benchmark statistical profiling performance
fn benchmark_statistical_profiling(c: &mut Criterion) {
    // Test numerical statistics with different data sizes
    for size in [100, 1000, 10000].iter() {
        let values: Vec<f64> = (0..*size).map(|i| (i as f64) + (i as f64 * 0.1)).collect();

        c.bench_with_input(
            BenchmarkId::new("numerical_statistics", size),
            &values,
            |b, values| b.iter(|| black_box(NumericalStatistics::from_values(values).unwrap())),
        );
    }

    // Benchmark data profile generation
    for size in [100, 1000].iter() {
        let records = create_benchmark_records(*size);

        c.bench_with_input(
            BenchmarkId::new("data_profile_generation", size),
            &records,
            |b, records| {
                b.iter(|| {
                    let mut profile = DataProfile::new();
                    for record in records {
                        profile.add_record(&record.fields).unwrap();
                    }
                    profile.finalize().unwrap();
                    black_box(profile)
                })
            },
        );
    }
}

/// Benchmark quality scoring performance
fn benchmark_quality_scoring(c: &mut Criterion) {
    for size in [100, 1000, 5000].iter() {
        let records = create_benchmark_records(*size);

        c.bench_with_input(
            BenchmarkId::new("quality_score_calculation", size),
            &records,
            |b, records| {
                b.iter(|| black_box(QualityScore::calculate_for_records(records).unwrap()))
            },
        );
    }

    // Benchmark individual quality dimensions
    let records = create_benchmark_records(1000);

    c.bench_function("completeness_scoring", |b| {
        b.iter(|| {
            // Simplified completeness calculation
            let total_fields = records.len() * 6; // 6 fields per record
            let non_null_fields = records
                .iter()
                .flat_map(|r| r.fields.values())
                .filter(|v| !v.is_null())
                .count();
            black_box(non_null_fields as f64 / total_fields as f64)
        })
    });
}

/// Benchmark anomaly detection performance
fn benchmark_anomaly_detection(c: &mut Criterion) {
    // Prepare test data
    let normal_data: Vec<f64> = (0..1000)
        .map(|i| i as f64 + rand::random::<f64>() * 10.0)
        .collect();
    let test_values = vec![500.0, 5000.0]; // Normal and anomalous

    // Benchmark Z-score detection
    let z_score_detector = ZScoreDetector::new(3.0);
    c.bench_function("z_score_detection", |b| {
        b.iter(|| {
            for &value in &test_values {
                black_box(
                    z_score_detector
                        .detect("test", value, &normal_data)
                        .unwrap(),
                );
            }
        })
    });

    // Benchmark IQR detection
    let iqr_detector = IQRDetector::new(1.5);
    c.bench_function("iqr_detection", |b| {
        b.iter(|| {
            for &value in &test_values {
                black_box(iqr_detector.detect("test", value, &normal_data).unwrap());
            }
        })
    });

    // Benchmark Isolation Forest (simplified)
    let mut isolation_forest = IsolationForestDetector::new(10, 100, 0.1);
    let training_data: Vec<Vec<f64>> = (0..100).map(|i| vec![i as f64]).collect();
    isolation_forest.train(&training_data).unwrap();

    c.bench_function("isolation_forest_score", |b| {
        b.iter(|| black_box(isolation_forest.score_multivariate(&[500.0]).unwrap()))
    });
}

/// Benchmark schema operations
fn benchmark_schema_operations(c: &mut Criterion) {
    let mut schema_manager = SchemaManager::new();

    // Benchmark schema inference with different data sizes
    for size in [100, 1000].iter() {
        let records = create_benchmark_records(*size);

        c.bench_with_input(
            BenchmarkId::new("schema_inference", size),
            &records,
            |b, records| {
                b.iter(|| {
                    let mut inference = SchemaInference::new();
                    black_box(
                        inference
                            .infer_schema("benchmark_schema".to_string(), records)
                            .unwrap(),
                    )
                })
            },
        );
    }

    // Benchmark schema validation
    let sample_records = create_benchmark_records(10);
    let schema = schema_manager
        .infer_schema("test_schema".to_string(), &sample_records)
        .unwrap();
    schema_manager.register_schema(schema).unwrap();

    let test_record = create_benchmark_record(1);
    c.bench_function("schema_validation", |b| {
        b.iter(|| black_box(schema_manager.validate_record(&test_record).unwrap()))
    });
}

/// Benchmark pipeline performance
fn benchmark_pipeline_performance(c: &mut Criterion) {
    let rt = Runtime::new().unwrap();

    // Create pipeline components
    let engine = std::sync::Arc::new(
        ValidationEngine::builder()
            .add_rule(ValidationRule::not_null("name"))
            .add_rule(ValidationRule::range("age", 0.0, 120.0))
            .build()
            .expect("Failed to build pipeline engine"),
    );

    // Benchmark streaming validation
    let stream_config = crate::pipeline::StreamConfig::default();
    let streaming_validator =
        crate::pipeline::StreamingValidator::new(engine.clone(), stream_config);

    c.bench_function("streaming_add_record", |b| {
        b.to_async(&rt).iter(|| async {
            let record = create_benchmark_record(1);
            black_box(streaming_validator.add_record(record).await.unwrap())
        })
    });

    // Benchmark batch validation
    let batch_config = crate::pipeline::BatchConfig::default();
    let batch_validator = crate::pipeline::BatchValidator::new(engine, batch_config);

    for size in [100, 1000].iter() {
        let batch_records = create_benchmark_records(*size);

        c.bench_with_input(
            BenchmarkId::new("batch_validation", size),
            &batch_records,
            |b, records| {
                b.to_async(&rt).iter(|| async {
                    black_box(
                        batch_validator
                            .process_batch(records.clone())
                            .await
                            .unwrap(),
                    )
                })
            },
        );
    }
}

/// Benchmark metrics collection
fn benchmark_metrics_collection(c: &mut Criterion) {
    let mut metrics = ValidationMetrics::new();

    c.bench_function("metrics_record_validation", |b| {
        b.iter(|| black_box(metrics.record_validation(Duration::from_millis(100), true)))
    });

    c.bench_function("metrics_record_error", |b| {
        b.iter(|| black_box(metrics.record_error("test_error", "warning")))
    });

    c.bench_function("metrics_update_resources", |b| {
        b.iter(|| black_box(metrics.update_resource_metrics(75.0, 1_000_000, 10)))
    });

    c.bench_function("metrics_calculate_throughput", |b| {
        b.iter(|| black_box(metrics.calculate_throughput(60.0)))
    });

    // Benchmark metrics with large history
    let mut large_metrics = ValidationMetrics::new();
    for i in 0..10000 {
        large_metrics.record_validation(Duration::from_millis(50 + i % 100), i % 10 != 0);
    }

    c.bench_function("metrics_summary_large", |b| {
        b.iter(|| black_box(large_metrics.summary()))
    });
}

/// Benchmark throughput with different configurations
fn benchmark_throughput_scaling(c: &mut Criterion) {
    let rt = Runtime::new().unwrap();

    // Test different engine configurations
    let engines = vec![
        (
            "minimal",
            ValidationEngine::builder()
                .add_rule(ValidationRule::not_null("name"))
                .build()
                .unwrap(),
        ),
        (
            "standard",
            ValidationEngine::builder()
                .add_rule(ValidationRule::not_null("name"))
                .add_rule(ValidationRule::range("age", 0.0, 120.0))
                .add_rule(ValidationRule::email("email"))
                .build()
                .unwrap(),
        ),
        (
            "comprehensive",
            ValidationEngine::builder()
                .with_statistical_validation(true)
                .add_rule(ValidationRule::not_null("name"))
                .add_rule(ValidationRule::range("age", 0.0, 120.0))
                .add_rule(ValidationRule::email("email"))
                .add_rule(ValidationRule::range("score", 0.0, 100.0))
                .add_rule(ValidationRule::length("name", 2, 100))
                .build()
                .unwrap(),
        ),
    ];

    // Test different record sizes
    for (engine_name, engine) in engines {
        for size in [100, 1000, 10000].iter() {
            let records = create_benchmark_records(*size);

            let mut group = c.benchmark_group(format!("throughput_{}_{}", engine_name, size));
            group.throughput(Throughput::Elements(*size as u64));

            group.bench_function("batch", |b| {
                b.to_async(&rt)
                    .iter(|| async { black_box(engine.validate_batch(&records).await.unwrap()) })
            });

            group.finish();
        }
    }
}

/// Benchmark memory efficiency
fn benchmark_memory_efficiency(c: &mut Criterion) {
    // Test memory usage with different data sizes
    for size in [1000, 10000, 100000].iter() {
        let records = create_benchmark_records(*size);

        c.bench_with_input(
            BenchmarkId::new("memory_profile_generation", size),
            &records,
            |b, records| {
                b.iter(|| {
                    let mut profile = DataProfile::new();
                    for record in records {
                        profile.add_record(&record.fields).unwrap();
                    }
                    // Measure peak memory usage during profile generation
                    black_box(profile)
                })
            },
        );
    }
}

criterion_group!(
    benches,
    benchmark_validation_engine,
    benchmark_validation_rules,
    benchmark_statistical_profiling,
    benchmark_quality_scoring,
    benchmark_anomaly_detection,
    benchmark_schema_operations,
    benchmark_pipeline_performance,
    benchmark_metrics_collection,
    benchmark_throughput_scaling,
    benchmark_memory_efficiency
);

criterion_main!(benches);