aprender-core 0.29.3

[package]
name = "aprender-core"
version.workspace = true
edition = "2021"
rust-version = "1.89"
authors = ["Noah Gift <noah@paiml.com>"]
license = "MIT"
description = "Next-generation machine learning library in pure Rust"
repository = "https://github.com/paiml/aprender"
documentation = "https://docs.rs/aprender"
readme = "README.md"
keywords = ["machine-learning", "classification", "clustering", "statistics", "graph-algorithms"]
categories = ["science", "algorithms"]
exclude = [
    # Build/IDE artifacts
    "target/",
    "*.profraw",
    "*.profdata",
    ".vscode/",
    ".idea/",
    "proptest-regressions/",
    # Dev tool artifacts (CB-510 class: recursive patterns required)
    ".pmat/",
    ".pmat-metrics/",
    ".pmat-metrics.toml",
    "*.bak",
    # CI/CD and dev infrastructure (not needed by library consumers)
    ".github/",
    ".githooks/",
    ".bashrsignore",
    "Makefile",
    "scripts/",
    # Documentation (published on GitHub Pages, not crates.io)
    "docs/",
    "book/",
    # Test data and traces
    "golden_traces/",
    "tokenizer.json",
    "defect-report-*.json",
    "trace_*.json",
    "fuzz/",
    # Model files (root-anchored: /models/ not models/ per CB-510)
    "/models/",
    # Contracts not consumed by build.rs (model-families/ IS needed)
    "contracts/chat-template-semantics-v1.yaml",
    "contracts/classification-finetune-v1.yaml",
    "contracts/kernel-fusion-v1.yaml",
    "contracts/layer-parity-v1.yaml",
    "contracts/model-metadata-bounds-v1.yaml",
    "contracts/quantized-dot-product-v1.yaml",
    "contracts/special-tokens-registry-v1.yaml",
    "contracts/tensor-layout-v1.yaml",
    "contracts/tokenizer-vocab-v1.yaml",
    "contracts/publish-safety-v1.yaml",
]

[lib]
name = "aprender"

[lints]
workspace = true

[dependencies]
# Serialization
serde = { version = "1.0", features = ["derive"] }
serde_json = "1.0"  # SafeTensors JSON metadata
bincode = "1.3"
rmp-serde = "1.3"  # MessagePack for .apr metadata (spec §2)
tempfile = "3.14"  # Streaming APR writer temp file (realizar#136)

# Random number generation for model_selection
rand = { version = "0.9", features = ["small_rng"] }
rand_chacha = "0.9"  # ChaCha20 PRNG for Monte Carlo simulations

# Audio processing
rustfft = { version = "6.2", optional = true }  # FFT for mel spectrogram computation
thiserror = { version = "2.0", optional = true }  # Error handling for audio module

# Native audio capture (Linux ALSA)
alsa = { version = "0.9", optional = true }  # ALSA bindings for Linux audio capture

# Parallelization for graph algorithms (optional for WASM compatibility)
rayon = { version = "1.10", optional = true }

# Shared formatting and system utilities (Batuta stack)
batuta-common = "0.1"

# Core compute primitives - SIMD-accelerated tensor operations
# v0.14.5: wgpu adapter limits fix (buffer >256MB for 7B+ models)
trueno = "0.17"

# K-quantization formats (Q4_K, Q5_K, Q6_K) - Toyota Way: ONE source of truth
# Path dependency until published to crates.io
trueno-quant = "0.1"

# RAG pipeline for document-based ML (optional, GH-125)
trueno-rag = { version = "0.2", optional = true }

# Compression for .apr format (optional, spec §3.3)
lz4_flex = { version = "0.11", optional = true }
zstd = { version = "0.13", optional = true }

# Half-precision floats for quantization (spec §6.2)
half = { version = "2.4", optional = true, default-features = false, features = ["std"] }

# Digital signatures for .apr format (optional, spec §4.2)
ed25519-dalek = { version = "2.1", optional = true, default-features = false, features = ["std", "zeroize", "rand_core"] }

# Encryption for .apr format (optional, spec §4.1)
aes-gcm = { version = "0.10", optional = true }
argon2 = { version = "0.5", optional = true, default-features = false, features = ["std"] }
x25519-dalek = { version = "2.0", optional = true, default-features = false, features = ["static_secrets"] }
hkdf = { version = "0.12", optional = true }
sha2 = { version = "0.10", optional = true }  # For HKDF-SHA256

# Data loading
alimentar = { version = "0.2.8", optional = true }

# entrenar: removed as runtime dep — explainable types now live in aprender (GH-305)
# entrenar remains as dev-dependency for InferenceMonitor integration tests

# Syscall/GPU profiling for showcase benchmarks (dev-only to avoid circular dep)
# renacer depends on aprender, so it cannot be a runtime dep
# renacer = { version = "0.10", optional = true }

# SIMD-accelerated compression for KV cache (optional)
trueno-zram-core = { version = "0.3.0", optional = true }

# Hugging Face Hub integration (optional, spec §11.8, hf-hub-upload-spec.md)
hf-hub = { version = "0.4", optional = true, default-features = false, features = ["ureq"] }
dirs = { version = "6.0", optional = true }
ureq = { version = "2.12", optional = true, features = ["json"] }  # Direct HTTP for HF Hub upload (APR-PUB-001)

# SafeTensors format parsing (optional, for HF model comparison)
safetensors = { version = "0.4", optional = true }

# WASM bindings for noise generator (optional, spec: noise-generator-apr-wasm-spec.md)
wasm-bindgen = { version = "0.2", optional = true }
js-sys = { version = "0.3", optional = true }
minijinja = { version = "2.14.0", features = ["loader", "serde"] }

# UCBD: compile-time contract enforcement via #[contract] proc macro
provable-contracts-macros = "0.3"

# Toyota Way: ONE source of truth for quantization (Step E)
# NOTE: Currently blocked by cyclic dependency (realizar optionally depends on aprender).
# Resolution requires creating a separate quantization crate.
# See docs/specifications/qwen2.5-coder-showcase-demo.md Section E.7 for roadmap.
# realizar = { version = "0.6", default-features = false, optional = true }  # BLOCKED

[build-dependencies]
serde = { version = "1", features = ["derive"] }
serde_yaml_ng = "0.10"

[dev-dependencies]
proptest = "1.6"
criterion = { workspace = true }
renacer = "0.10"
tempfile = "3.14"  # For format module tests
jugar-probar = "0.5"  # TUI/GUI testing framework with coverage tracking (spec §8)
ctrlc = "3.4"  # Signal handling for SIGINT/SIGTERM (PMAT-098-PF: zombie process mitigation)
provable-contracts = "0.3"  # Contract enforcement (dev-only)
entrenar = "0.7"  # Integration tests for InferenceMonitor (GH-305: was runtime dep, now dev-only)

[features]
default = ["parallel"]
parallel = ["rayon"]  # Enable parallel graph algorithms (disable for WASM)
datasets = ["alimentar"]  # Enable data loading from alimentar
format-compression = ["lz4_flex", "zstd"]  # Enable LZ4/ZSTD compression for .apr format (spec §3.3, GH-146)
format-signing = ["ed25519-dalek"]  # Enable Ed25519 signatures for .apr format (spec §4.2)
format-encryption = ["aes-gcm", "argon2", "x25519-dalek", "hkdf", "sha2"]  # Enable encryption for .apr format (spec §4.1)
format-quantize = ["half"]  # Enable quantization for .apr format (spec §6.2)
format-homomorphic = []  # Enable homomorphic encryption for .apr format (spec: homomorphic-encryption-spec.md)
# Note: mmap is automatic on native platforms, no feature needed (spec: bundle-mmap-spec.md)
hf-hub-integration = ["hf-hub", "dirs", "ureq", "sha2"]  # Enable Hugging Face Hub integration (GH-100, APR-PUB-001)
audio = ["rustfft", "thiserror"]  # Enable audio processing (mel spectrogram, resampling)
audio-capture = ["audio"]  # Enable audio capture base functionality
audio-alsa = ["audio-capture", "alsa"]  # Enable ALSA audio capture (Linux only)
audio-coreaudio = ["audio-capture"]  # Enable CoreAudio capture (macOS only)
audio-wasapi = ["audio-capture"]  # Enable WASAPI capture (Windows only)
audio-webaudio = ["audio-capture"]  # Enable WebAudio capture (WASM only)
audio-playback = ["audio"]  # Enable audio playback
audio-codec = ["audio"]  # Enable audio codec decoding (WAV, MP3, AAC, FLAC, Opus)
audio-noise = ["audio"]  # Enable ML-based noise generation (GH-144)
audio-noise-wasm = ["audio-noise", "wasm-bindgen", "js-sys"]  # Enable WASM bindings for noise generator
safetensors-compare = ["safetensors", "hf-hub-integration", "half"]  # Enable SafeTensors comparison (GH-121)
rag = ["trueno-rag"]  # Enable RAG pipeline for document-based ML (GH-125)
# inference-monitoring feature removed — explainable types now unconditional (GH-305)
gpu = ["trueno/gpu"]  # Enable GPU acceleration via trueno wgpu backend
model-tests = []  # Enable heavy model/inference tests (requires models/ dir, ollama, GPU)
cuda = ["trueno/cuda-monitor"]  # Enable CUDA monitoring via trueno-gpu (NVIDIA GPUs)
cpu-only = []
showcase-profile = []  # Renacer profiling (disabled: renacer→aprender circular dep)
showcase-zram = ["trueno-zram-core"]  # Enable trueno-zram KV cache compression for showcase benchmarks (PAR-040)
explainable-monitor-integration = []  # Enable integration tests for aprender explainability + entrenar InferenceMonitor (GH-305)
# Chaos engineering features (from renacer)
chaos-basic = []
chaos-network = ["chaos-basic"]
chaos-byzantine = ["chaos-basic"]
chaos-full = ["chaos-network", "chaos-byzantine"]

# WASM support: enable getrandom's "js" feature for browser environments
[target.'cfg(target_arch = "wasm32")'.dependencies]
getrandom = { version = "0.2", features = ["js"] }

# Memory-mapped I/O for native platforms (spec: bundle-mmap-spec.md)
[target.'cfg(not(target_arch = "wasm32"))'.dependencies]
memmap2 = "0.9"

[[test]]
name = "book"
path = "tests/book/mod.rs"

[[bench]]
name = "linear_regression"
harness = false

[[bench]]
name = "kmeans"
harness = false

[[bench]]
name = "dataframe"
harness = false

[[bench]]
name = "graph"
harness = false

[[bench]]
name = "recommend"
harness = false

[[bench]]
name = "citl"
harness = false

[[bench]]
name = "bpe"
harness = false

[[bench]]
name = "ollama_parity"
harness = false
required-features = ["format-quantize"]

[[example]]
name = "shell_encryption_demo"
required-features = ["format-encryption"]

[[example]]
name = "chat_template"

[[example]]
name = "text_preprocessing"

[[example]]
name = "time_series_forecasting"

[profile.release]
lto = true
codegen-units = 1
panic = "abort"
strip = "none"
debug = true  # Enable debug info for flamegraph/profiling

[profile.dev]
lto = false
panic = "abort"

[package.metadata.docs.rs]
all-features = true
rustdoc-args = ["--generate-link-to-definition"]
targets = ["x86_64-unknown-linux-gnu"]

[package.metadata.release]
sign-commit = false
sign-tag = false
push = true
publish = true
shared-version = true
tag-name = "v{{version}}"
pre-release-commit-message = "release: aprender v{{version}}"

[[package.metadata.release.pre-release-replacements]]
file = "CHANGELOG.md"
search = "## \\[Unreleased\\]"
replace = "## [{{version}}] - {{date}}"

# PMAT-262: Self-patch so transitive deps (realizar, entrenar) use the local
# workspace aprender instead of a stale crates.io version. This prevents type
# mismatches when building apr-cli from the workspace.
# GH-344: Sibling patches (realizar, trueno, etc.) moved to .cargo/config.toml.dev-overrides
# so that `git clone && cargo check` works without sibling repos.
# See .cargo/config.toml.dev-overrides for full-stack development setup.