[package]
edition = "2024"
name = "xet-data"
version = "1.5.1"
build = false
autolib = false
autobins = false
autoexamples = false
autotests = false
autobenches = false
description = "Data processing pipeline for chunking, deduplication, and file reconstruction; used in the Hugging Face Xet client tools. Intended to be used through the API in the hf-xet package."
homepage = "https://github.com/huggingface/xet-core"
readme = "README.md"
keywords = ["huggingface"]
categories = [
"artificial-intelligence",
"data-structures",
"filesystem",
]
license = "Apache-2.0"
repository = "https://github.com/huggingface/xet-core"
resolver = "2"
[features]
expensive_tests = []
fd-track = [
"xet-runtime/fd-track",
"xet-client/fd-track",
]
python = ["dep:pyo3"]
simulation = [
"xet-client/simulation",
"xet-core-structures/simulation",
]
smoke-test = []
strict = []
[lib]
name = "xet_data"
path = "src/lib.rs"
doctest = false
[[bin]]
name = "x"
path = "src/processing/bin/example.rs"
[[bin]]
name = "xtool"
path = "src/processing/bin/xtool.rs"
[[example]]
name = "chunk"
path = "examples/chunk/main.rs"
[[example]]
name = "hash"
path = "examples/hash/main.rs"
[[example]]
name = "xorb-check"
path = "examples/xorb-check/main.rs"
[[test]]
name = "integration_tests"
path = "tests/integration_tests.rs"
[[test]]
name = "test_clean_smudge"
path = "tests/test_clean_smudge.rs"
[[test]]
name = "test_clean_smudge_multirange"
path = "tests/test_clean_smudge_multirange.rs"
[[test]]
name = "test_full_file_download"
path = "tests/test_full_file_download.rs"
[[test]]
name = "test_range_downloads"
path = "tests/test_range_downloads.rs"
[[test]]
name = "test_session_resume"
path = "tests/test_session_resume.rs"
[[test]]
name = "test_unordered_download"
path = "tests/test_unordered_download.rs"
[[bench]]
name = "reconstruction_bench"
path = "benches/reconstruction_bench.rs"
harness = false
[dependencies.anyhow]
version = "1"
[dependencies.async-trait]
version = "0.1"
[dependencies.bytes]
version = "1.11"
[dependencies.chrono]
version = "0.4"
[dependencies.clap]
version = "4"
features = ["derive"]
[dependencies.gearhash]
version = "0.1"
[dependencies.http]
version = "1"
[dependencies.itertools]
version = "0.14"
[dependencies.lazy_static]
version = "1.5"
[dependencies.more-asserts]
version = "0.3"
[dependencies.pyo3]
version = "0.26"
features = ["abi3-py37"]
optional = true
[dependencies.rand]
version = "0.9"
[dependencies.serde]
version = "1"
features = ["derive"]
[dependencies.serde_json]
version = "1"
[dependencies.tempfile]
version = "3.25"
[dependencies.thiserror]
version = "2.0"
[dependencies.tokio-util]
version = "0.7"
[dependencies.tracing]
version = "0.1"
[dependencies.ulid]
version = "1.2"
[dependencies.url]
version = "2.5"
[dependencies.walkdir]
version = "2"
[dependencies.xet-client]
version = "1.5.1"
[dependencies.xet-core-structures]
version = "1.5.1"
[dependencies.xet-runtime]
version = "1.5.1"
[dev-dependencies.criterion]
version = "0.4"
features = ["async_tokio"]
[dev-dependencies.ctor]
version = "0.6"
[dev-dependencies.dirs]
version = "6.0"
[dev-dependencies.rand]
version = "0.9"
[dev-dependencies.regex]
version = "1"
[dev-dependencies.serial_test]
version = "3"
[dev-dependencies.tempfile]
version = "3.25"
[dev-dependencies.tracing-test]
version = "0.2"
features = ["no-env-filter"]
[target.'cfg(not(target_family = "wasm"))'.dependencies.tokio]
version = "1.49"
features = [
"rt-multi-thread",
"rt",
"time",
]
[target.'cfg(not(target_os = "windows"))'.dependencies.sha2]
version = "0.10"
features = ["asm"]
[target.'cfg(target_family = "wasm")'.dependencies.tokio]
version = "1.49"
features = [
"sync",
"macros",
"io-util",
"rt",
"time",
]
[target.'cfg(target_os = "windows")'.dependencies.sha2]
version = "0.10"