pyproject.toml

[build-system]
# mypyc compilation (setup.py, gated by INVOICE2DATA_COMPILE_MYPYC) needs
# setuptools + mypy at build time. mypyc must see the types of every dependency
# used by the compiled modules, so list them here too. (dateparser is NOT here
# because `_dates.py` -- the only module that imports it -- is intentionally
# not compiled; keeping dateparser truly optional at install time.)
requires = [
    "setuptools>=64",
    "mypy>=1.13",
    "click>=8.0.1",
    "PyYAML>=6.0",
    "regex>=2024.4.16",
]
build-backend = "setuptools.build_meta"

[project]
name = "invoice2data"
authors = [{ name = "Manuel Riel" }]
description = "Python parser to extract data from pdf invoice"
version = "1.0.0"
keywords = ["python", "data-mining", "accounting", "invoice", "pdf", "parcing"]
license = { text = "MIT" }
readme = "README.md"
requires-python = ">=3.10"
classifiers = [
    "Programming Language :: Python :: 3",
    "License :: OSI Approved :: MIT License",
    "Operating System :: OS Independent",
    "Programming Language :: Python :: 3.10",
    "Programming Language :: Python :: 3.11",
    "Programming Language :: Python :: 3.12",
    "Programming Language :: Python :: 3.13",
    "Topic :: Office/Business :: Financial",
    "Topic :: Office/Business :: Financial :: Accounting",
    "Environment :: Console",
    "Intended Audience :: Financial and Insurance Industry",
    "Intended Audience :: Developers",
    "Topic :: Office/Business :: Financial :: Accounting",
    "Topic :: Office/Business :: Financial",
    "Topic :: Software Development :: Libraries :: Python Modules",
    "Development Status :: 5 - Production/Stable",
]

dependencies = [
    "click >=8.0.1",
    "python-dateutil >= 2.8",
    "PyYAML >= 6.0",
    "regex >= 2025.2.10; python_version >= '3.12'",
    "regex >= 2024.4.16; python_version < '3.12'",
]


[project.urls]
homepage = "https://github.com/invoice-x/invoice2data"
repository = "https://github.com/invoice-x/invoice2data"
documentation = "https://invoice2data.readthedocs.io"
Changelog = "https://github.com/invoice-x/invoice2data/releases"


[dependency-groups]
dev = [
    "coverage[toml] >= 6.0",
    "pre-commit >=2.16.0",
    "pre-commit-hooks >=4.1.0",
    "pytest >=8.2.2",
    "pytest-mock >= 3.14.0",
    "pygments >=2.10.0",
]
lint = ["ruff >=0.8", "pydoclint >=0.4.2"]
docs = [
    "shibuya >= 2025.1.29",
    "myst-parser >= 4.0",
    "sphinx >= 8.0",
    "sphinx-autobuild >= 2024.10",
    "sphinx-click >= 6.0",
    "sphinx-mermaid >=0.0.7",
    "sphinxcontrib-svg2pdfconverter >=2.1.0",
]
mypy = ["mypy >= 1.13"]
ty = ["ty == 0.0.39"]  # beta + fast-moving: pin so new diagnostics don't surprise CI
typeguard = ["typeguard >= 4.4"]
xdoctest = ["xdoctest[colors] >= 1.2"]

[project.optional-dependencies]
ai = ["httpx >= 0.27"]
camelot = ["camelot-py >= 1.0"]
dateparser = ["dateparser >= 1.2.0"]
defusedxml = ["defusedxml == 0.7.1"]
doctr = ["python-doctr[torch] >= 0.8"]
paddleocr = ["paddleocr >= 2.7", "paddlepaddle >= 2.5", "pypdfium2 >= 4"]
googlevision = [
    "google-cloud-storage == 2.18.2",
    "google-cloud-vision == 3.8.1",
]
hotpdf = ["hotpdf >= 0.1.0"]
ocr = ["ghostscript == 0.7"]
ocrmypdf = ["ocrmypdf >= 14.4.0"]
pdfium = ["pypdfium2 >= 4"]
pdfminer-six = ["pdfminer-six == 20231228"]
pdfoxide = ["pdf-oxide >= 0.3"]
pdfplumber = ["pdfplumber == 0.11.4"]
pyyaml = ["pyyaml == 6.0.2"]

[tool.uv]
package = true
# Published camelot-py requires pdfminer-six>=20240706, but pdfplumber==0.11.4
# and the `pdfminer-six` extra both hard-pin pdfminer.six==20231228 — a genuine
# library-level conflict (not invoice2data's pin). Mark camelot mutually
# exclusive with those pdfminer-bound extras; a camelot user runs it alongside
# the default pdfium/pdftotext cascade. (The next-major camelot drops
# pdfminer.six for playa-pdf, which will retire these once published.)
conflicts = [
    [{ extra = "camelot" }, { extra = "pdfminer-six" }],
    [{ extra = "camelot" }, { extra = "pdfplumber" }],
]

[tool.setuptools.packages.find]
where = ["src"]

[tool.setuptools.package-data]
# Ship py.typed + the bundled templates; test fixtures live outside src/ so
# PDFs/PNGs never enter the wheel, and setuptools omits tests/ from the sdist.
invoice2data = [
    "py.typed",
    "extract/templates/**/*.yml",
    "extract/templates/**/*.json",
]


[project.scripts]
invoice2data = "invoice2data.__main__:main"

[tool.cibuildwheel]
# Build mypyc-compiled CPython wheels (requires-python>=3.10 is honoured
# automatically). Skip PyPy (already JIT-compiled) and musllinux.
build = "cp310-* cp311-* cp312-* cp313-*"
skip = "pp* *-musllinux*"
build-frontend = "build"
environment = { INVOICE2DATA_COMPILE_MYPYC = "1" }
# Smoke-test that the compiled extensions import on every target platform.
test-command = "python -c \"import invoice2data.extract.parsers.regex, invoice2data.extract.parsers.lines, invoice2data.extract.plugins.tables, invoice2data.extract._regex, invoice2data.extract.utils; print('mypyc import OK')\""

[tool.coverage.paths]
source = ["src", "*\\src", "*/site-packages"]
tests = ["tests", "*/tests"]

[tool.coverage.run]
branch = true
source = ["invoice2data", "tests"]
relative_files = true

[tool.coverage.report]
show_missing = true
# Locked in at 90 (combined coverage is ~94%); not the uv-forge template default
# of 100 — that's unrealistic here. Catches regressions, small buffer for variance.
fail_under = 90
exclude_lines = ["pragma: no cover", "if TYPE_CHECKING:"]

[tool.ruff.lint.isort]
force-single-line = true
lines-after-imports = 2

[tool.mypy]
strict = true
warn_unreachable = true
pretty = true
show_column_numbers = true
show_error_context = true

# Optional third-party backends/tools that ship no type information. Centralised
# here so the per-call inline imports stay clean and the result is the same
# whether or not the optional dependency happens to be installed.
[[tool.mypy.overrides]]
module = [
    "pdfminer",
    "pdfminer.*",
    "pdfplumber",
    "pypdfium2",
    "pdf_oxide",
    "hotpdf",
    "ocrmypdf",
    "defusedxml",
    "defusedxml.*",
    "docutils",
    "docutils.*",
    "sphinxmermaid",
    "httpx",
    "doctr",
    "doctr.*",
    "paddleocr",
    "paddleocr.*",
    "dateparser",
    "dateparser.*",
    "dateutil",
    "dateutil.*",
]
ignore_missing_imports = true

# `google.cloud` is a namespace package: mypy partially resolves it (so
# ignore_missing_imports does not apply), but the optional storage/vision
# submodules are absent. camelot is untyped and only re-exports `read_pdf`
# implicitly (tripping strict no-implicit-reexport when it happens to be
# installed). Skip following both entirely so they are treated as Any.
[[tool.mypy.overrides]]
module = ["google", "google.*", "camelot", "camelot.*"]
follow_imports = "skip"
ignore_missing_imports = true

[tool.ty.rules]
# ty (beta, advisory) runs without the optional backend extras installed — some
# of them conflict (see [tool.uv] conflicts), so they can't all be present. mypy
# is the authoritative checker for imports (via its overrides above); let ty skip
# the optional backends it can't resolve rather than fail.
unresolved-import = "ignore"

[tool.ruff]
src = ["src", "tests"]

[tool.ruff.lint]
extend-select = [
    "ASYNC", # flake8-async
    "B",     # flake8-bugbear
    "C4",    # flake8-comprehensions
    "C90",   # mccabe
    "D",     # pydocstyle
    "E",     # pycodestyle
    "F",     # pyflakes
    "FURB",  # refurb
    "I",     # isort
    "LOG",   # flake8-logging
    "N",     # pep8-naming
    "PERF",  # perflint
    "PIE",   # flake8-pie
    "PTH",   # flake8-use-pathlib
    "RET",   # flake8-return
    "RUF",   # Ruff-specific rules
    "S",     # flake8-bandit
    "SIM",   # flake8-simplify
    "TRY",   # tryceratops
    "UP",    # pyupgrade
    "W",     # pycodestyle
]
# ignore checks that conflict with ruff formatter (https://docs.astral.sh/ruff/formatter/#conflicting-lint-rules)
extend-ignore = [
    "COM812",
    "COM819",
    "D107",   # undocumented-public-init conflicts with DOC301
    "D206",
    "D300",
    "E111",
    "E114",
    "E117",
    "ISC001",
    "ISC002",
    "Q000",
    "Q001",
    "Q002",
    "Q003",
    "W191",
    "RUF013", # activate after implement typing
    "D103",   # fix docstrings in public functions later
    "S101",   # use of assert Activate later
    "S603",
    "UP031",  # Use fstring instead of % identifier
    "E501",   # fix this when updating docstrings
    "TRY003", # long messages in raise (would need custom exception classes)
    # "DOC106",
    # "DOC107",
    # "DOC111",
]

[tool.ruff.lint.per-file-ignores]
"*/test_*.py" = ["S101", "D100", "D101", "D102"]
"noxfile.py" = ["S101"]
"**/conftest.py" = ["S101"]
# Test scaffolding keeps os.path; pathlib there is churn with no runtime benefit.
"tests/**" = ["PTH"]

[tool.ruff.lint.pydocstyle]
convention = "google"

[tool.pydoclint]
style = 'google'
exclude = '\.git|\.nox|tests*|noxfile.py|tables.py'
arg-type-hints-in-docstring = true