pyproject.toml

[tool.poetry]
name = "text-dedup"
version = "0.3.1"
description = "All-in-one text deduplication"
authors = ["Chenghao Mou <mouchenghao@gmail.com>"]
license = "Apache-2.0"
readme = "README.md"
packages = [{include = "text_dedup"}]

[tool.poetry.dependencies]
python = "^3.10,<3.12"
numpy = "^1.23.2"
tqdm = "^4.64.1"
datasets = "^2.4.0"
rich = "^12.5.1"
scipy = "1.10.1"
xxhash = "^3.0.0"
pybloom-live = "^4.0.0"
bitarray = "^2.6.2"
pyspark = "^3.3.1"
regex = "^2023.5.5"
urllib3 = "<=2.0"
sphinxcontrib-bibtex = "^2.5.0"
zstandard = "^0.21.0"
ftfy = "^6.1.1"

[tool.poetry.group.dev.dependencies]
pytest = "^7.1.3"
coverage = "^6.4.4"
Sphinx = "^5.1.1"
furo = "^2022.6.21"
pre-commit = "^2.20.0"
isort = "^5.10.1"
livereload = "^2.6.3"
matplotlib = "^3.7.1"
seaborn = "^0.12.2"
ruff = "^0.0.265"
insegel = "^1.3.1"
sphinx-autobuild = "^2021.3.14"
google-cloud-storage = "^2.10.0"

[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"

[tool.ruff]
line-length = 120

[tool.ruff.isort]
force-single-line = true