aprender-data 0.29.0

[package]
exclude = ["tests/", "benches/", "docs/", ".github/", "book/", ".pmat/"]
name = "aprender-data"
version.workspace = true
edition = "2021"
authors = ["paiml"]
description = "Data Loading, Distribution and Tooling in Pure Rust"
license = "MIT"
repository = "https://github.com/paiml/alimentar"
keywords = ["data", "ml", "arrow", "parquet", "dataset"]
categories = ["data-structures", "science"]
rust-version = "1.75"

[features]
default = ["local", "tokio-runtime", "cli", "repl", "mmap", "shuffle", "provenance"]
local = []
mmap = ["dep:memmap2"]
shuffle = ["dep:rand"]
provenance = ["dep:sha2"]
tokio-runtime = ["dep:tokio"]
s3 = ["dep:aws-sdk-s3", "dep:aws-config", "tokio-runtime"]
http = ["dep:reqwest"]
hf-hub = ["http", "dep:base64", "dep:hex", "dep:sha2"]
cli = ["dep:clap", "dep:crossterm"]
repl = ["cli", "dep:reedline", "dep:nu-ansi-term"]
wasm = ["dep:wasm-bindgen", "dep:wasm-bindgen-futures", "dep:js-sys", "dep:getrandom"]
format-encryption = ["dep:aes-gcm", "dep:argon2", "dep:x25519-dalek", "dep:hkdf", "dep:sha2", "dep:rand_core", "dep:getrandom"]
format-signing = ["dep:ed25519-dalek", "dep:getrandom"]
format-streaming = ["mmap"]
doctest = ["dep:chrono", "dep:regex", "dep:walkdir"]
dhat-heap = ["dep:dhat"]

[dependencies]
# Core (always)
# NOTE: arrow 57+ fixes comfy-table let_chains compatibility
arrow = { version = "57", default-features = false, features = ["prettyprint", "ipc"] }
arrow-csv = { version = "57" }
arrow-json = { version = "57" }
parquet = { version = "57", default-features = false, features = ["arrow", "snap", "zstd"] }
thiserror = "2"
bytes = "1"
serde = { version = "1", features = ["derive"] }
serde_json = "1"

# Format serialization
rmp-serde = "1"
zstd = "0.13"
lz4_flex = "0.11"

# Random (for shuffle) - with JS support for WASM
rand = { version = "0.8", optional = true }
getrandom = { version = "0.2", features = ["js"], optional = true }

# Memory mapping (not available on WASM)
memmap2 = { version = "0.9", optional = true }

# CLI (optional, not for WASM)
clap = { version = "4", features = ["derive"], optional = true }
crossterm = { version = "0.28", optional = true }

# REPL (feature-gated)
reedline = { version = "0.37", optional = true }
nu-ansi-term = { version = "0.50", optional = true }

# Async (feature-gated)
tokio = { version = "1", features = ["rt", "fs", "sync", "rt-multi-thread"], optional = true }

# S3 (feature-gated)
aws-sdk-s3 = { version = "1", optional = true }
aws-config = { version = "1", optional = true }

# HTTP (feature-gated)
reqwest = { version = "0.12", default-features = false, features = ["rustls-tls", "blocking", "json", "multipart"], optional = true }
base64 = { version = "0.22", optional = true }
hex = { version = "0.4", optional = true }

# WASM (feature-gated)
wasm-bindgen = { version = "0.2", optional = true }
wasm-bindgen-futures = { version = "0.4", optional = true }
js-sys = { version = "0.3", optional = true }

# Encryption (feature-gated, WASM-safe)
aes-gcm = { version = "0.10", optional = true }
argon2 = { version = "0.5", optional = true }
x25519-dalek = { version = "2", features = ["static_secrets"], optional = true }
hkdf = { version = "0.12", optional = true }
sha2 = { version = "0.10", optional = true }
rand_core = { version = "0.6", optional = true }

# Signing (feature-gated, WASM-safe)
ed25519-dalek = { version = "2", features = ["rand_core"], optional = true }
serde_yaml = "0.9.34"

# Heap profiling (feature-gated)
dhat = { version = "0.3", optional = true }

# TUI support (unicode width calculation)
unicode-width = "0.2"

# Doctest extraction (feature-gated)
chrono = { version = "0.4", features = ["serde"], optional = true }
regex = { version = "1", optional = true }
walkdir = { version = "2", optional = true }

[dev-dependencies]
tempfile = "3"
tokio = { version = "1", features = ["rt", "rt-multi-thread", "macros"] }
criterion = { version = "0.5", features = ["html_reports"] }
proptest = "1.6"
assert_cmd = "2"
predicates = "3"
jugar-probar = "1.0.1"

[[bin]]
name = "alimentar"
path = "src/main.rs"
required-features = ["cli"]

[[bin]]
name = "generate_fixtures"
path = "scripts/generate_fixtures.rs"

[lib]
name = "alimentar"
path = "src/lib.rs"

[profile.release]
lto = true
codegen-units = 1
strip = true

[lints.rust]
unsafe_code = "deny"
unsafe_op_in_unsafe_fn = "warn"
missing_docs = "warn"
rust_2018_idioms = { level = "warn", priority = -1 }

[lints.clippy]
correctness = { level = "deny", priority = -1 }
suspicious = { level = "warn", priority = -1 }
perf = { level = "warn", priority = -1 }
complexity = { level = "warn", priority = -1 }
style = { level = "warn", priority = -1 }
pedantic = { level = "warn", priority = -1 }
nursery = { level = "warn", priority = -1 }
unwrap_used = "allow"
expect_used = "allow"
panic = "allow"
unreachable = "warn"
checked_conversions = "warn"
fallible_impl_from = "warn"
missing_errors_doc = "allow"
missing_panics_doc = "allow"
module_name_repetitions = "allow"
must_use_candidate = "allow"
too_many_lines = "allow"
cognitive_complexity = "allow"
significant_drop_tightening = "allow"
redundant_pub_crate = "allow"
missing_const_for_fn = "allow"
option_if_let_else = "allow"
uninlined_format_args = "allow"
cast_precision_loss = "allow"
cast_sign_loss = "allow"
cast_possible_truncation = "allow"
cast_possible_wrap = "allow"
len_zero = "allow"
redundant_closure = "allow"
redundant_closure_for_method_calls = "allow"
bool_comparison = "allow"
approx_constant = "allow"

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

[[example]]
name = "doctest_extraction"
required-features = ["doctest"]

[[example]]
name = "hub_publishing"
required-features = ["hf-hub"]

[package.metadata.docs.rs]
all-features = true
rustdoc-args = ["--generate-link-to-definition"]

[package.metadata.release]
shared-version = true

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

[profile.dev]
panic = "abort"

[profile.test]
opt-level = 1

[profile.bench]
lto = true
codegen-units = 1