pyproject.toml

[tool.poetry]
name = "libpdf"
version = "0.1.0"
description = "Extract structured data from PDFs."
authors = [
    "Marco Heinemann <marco.heinemann@useblocks.com>",
]
maintainers = [
    "Jui-Wen Chen <jui-wen.chen@useblocks.com>",
    "Haiyang Zhang <haiyang.zhang@useblocks.com>",
]
license = 'MIT'
readme = 'README.rst'
homepage = 'http://pypi.python.org/pypi/libpdf'
repository = 'http://github.com/useblocks/libpdf'
documentation = 'http://libpdf.readthedocs.io/en/latest'
classifiers = [
    'Development Status :: 3 - Alpha',
    'Intended Audience :: Developers',
    'License :: OSI Approved :: MIT License',
    'Operating System :: OS Independent',
    'Programming Language :: Python',
    'Programming Language :: Python :: 3',
    'Programming Language :: Python :: 3.8',
    'Programming Language :: Python :: 3.9',
    'Programming Language :: Python :: 3.10',
    'Programming Language :: Python :: 3.11',
    'Programming Language :: Python :: 3.12',
    'Topic :: Scientific/Engineering',
    'Topic :: Scientific/Engineering :: Information Analysis',
    'Topic :: Software Development',
    'Topic :: Software Development :: Libraries',
    'Topic :: Software Development :: Libraries :: Python Modules',
    'Topic :: Text Processing',
    'Topic :: Utilities',
]
include = [
    'deps/*'
]

[tool.poetry.dependencies]
python = "^3.8"
chardet = "^4"
click = "^8"
PyYAML = "^6"
"ruamel.yaml" = "^0.17"

# optional deps for progress bars
tqdm = { version = "^4.50.0", optional = true }
colorama = { version = "^0.4.4", optional = true }

# dependencies needed by pdfminer.six and pdfplumber which are deliverd as wheels in the deps folder
# the libs were patched and no upstream PR has been raised yet
# see [tool.poetry.dev-dependencies] for the forked libraries
pycryptodome = "^3.9.9"
sortedcontainers = "^2.3.0"
pillow = "~9.0.1"  # cannot go higher because of this bug https://github.com/jsvine/pdfplumber/discussions/637
unicodecsv = "^0.14.1"
wand = "^0.6.5"

# optional deps for docs, needed to make RTD work with pyproject.toml
# see https://github.com/readthedocs/readthedocs.org/issues/4912#issuecomment-664002569
sphinx = { version = "*", optional = true }
sphinx_rtd_theme = { version = "*", optional = true }
sphinxcontrib-plantuml = { version = "*", optional = true }

[tool.poetry.extras]
tqdm = ["tqdm"]
colorama = ["colorama"]
docs = ["sphinx", "sphinx_rtd_theme", "sphinxcontrib-plantuml"]

[tool.poetry.dev-dependencies]
# forked libraries;
# they must be dev deps so make them invisible to PyPI and the egg/wheel requires section
# 'poetry install' installs those by default, so the Git deps are actually used instead of the wheels in the deps
# folder; see also libpdf/_import_forks.py
pdfplumber = { git = "https://github.com/useblocks/pdfplumber.git" }
"pdfminer.six" = {git = "https://github.com/useblocks/pdfminer.six", rev = "develop"}

# testing
pytest = "^7"
pytest-cov = "*"
tox = "*"
pytest-xdist = "*"  # parallelisation

# linting, formatting
ruff = "^0.2.0"

# docs
sphinx = "*"
sphinx_rtd_theme = "*"
sphinxcontrib-plantuml = "*"

[tool.ruff]
# Exclude a variety of commonly ignored directories.
exclude = [
    ".bzr",
    ".direnv",
    ".eggs",
    ".git",
    ".hg",
    ".mypy_cache",
    ".nox",
    ".pants.d",
    ".pytype",
    ".ruff_cache",
    ".svn",
    ".tox",
    ".venv",
    "__pypackages__",
    "_build",
    "buck-out",
    "build",
    "dist",
    "node_modules",
    "venv",
    ".venv",
]

# use Black's default
line-length = 88

# lint according to oldest supported version
target-version = "py38"

# try out unstable rules, fixes, and formatting
# changes are reviewed anyway
preview = true

[tool.ruff.lint]
# concept: select all rules by default and ignore unwanted (maybe just per file)
# this comes at a cost: new added rules are applied automatically, so a CI that
# runs green today might fail tomorow in case ruff gets updated
# (which brings more good than bad)
# TODO ENABLE NEXT LINE
select = ["ALL"]
# select = [] # commented out (don't want to fix all issues with this PR)
ignore = [
    "COM812", # may conflict with Ruff formatter
    "D107",   # undocumented-public-init - commonly the docstring provides no helpful information
    "D203",   # D203 and D211 are mutually exclusive: https://github.com/PyCQA/pydocstyle/issues/141
    "D212",   # D212 and D213 are mutually exclusive: https://stackoverflow.com/a/45990465
    "FIX002", # line-contains-todo - TODOs are considered helpful
    "ISC001", # may conflict with Ruff formatter
    "TD003",  # missing-todo-link - no big processes for us
    "ANN101", # missing-type-self - infer typing is simple for this library
    "CPY001", # missing-copyright-notice - specified on project level
    "FBT001", # boolean-type-hint-positional-argument - good message, but need a concept how to deal with this
]

# Allow autofix for all enabled rules (when `--fix`) is provided.
fixable = ["ALL"]
unfixable = [
    "D400",   # ends-in-period - dangerous to use (ruff may confuse summary line and content)
    "ERA001", # commented-out-code - commented code is commonly there for a reason
]

[tool.ruff.lint.pylint]
max-args = 8

[tool.ruff.lint.flake8-annotations]
mypy-init-return = true  # no init return type needed if at least one arg is annotated

[tool.ruff.lint.mccabe]
max-complexity = 20

[tool.ruff.lint.per-file-ignores]
# "docs/conf.py" = ["E402"]
"**/__init__.py" = [
    "D104",
] # Missing docstring in public package - not needed for inits
"tests/**/*.py" = [
    "S101", # Use of `assert` detected
]

[tool.poetry.scripts]
libpdf = 'libpdf.core:main_cli'

[build-system]
requires = ["poetry>=0.12"]
build-backend = "poetry.masonry.api"