pyproject.toml

[build-system]
requires = [
  "setuptools >= 61",
  "setuptools_scm[toml] >= 8",
]
build-backend = 'setuptools.build_meta'

[project]
name="INFRA-COMPASS"
dynamic = ["version"]
description = "INFRA-COMPASS is a tool that leverages Large Language Models (LLMs) to create and maintain an inventory of state and local codes and ordinances applicable to energy infrastructure."
readme = "README.rst"
authors = [
  {name = "Paul Pinchuk", email = "ppinchuk@nlr.gov"},
  {name = "Guilherme Castelão", email = "gpimenta@nlr.gov"},
]
maintainers = [
  {name = "Paul Pinchuk", email = "ppinchuk@nlr.gov"},
  {name = "Guilherme Castelão", email = "gpimenta@nlr.gov"},
]
license = "BSD-3-Clause"
keywords = ["COMPASS", "NLR", "Ordinance", "Database"]
requires-python = ">=3.12"
classifiers=[
  "Development Status :: 4 - Beta",
  "Intended Audience :: Science/Research",
  "Natural Language :: English",
  "Operating System :: OS Independent",
  "Programming Language :: Python :: 3",
  "Programming Language :: Python :: 3.12",
  "Programming Language :: Python :: 3.13",
  "Topic :: Software Development :: Libraries :: Application Frameworks",
]
dependencies = [
  "beautifulsoup4>=4.12.3,<5",
  "click>=8.1.7,<9",
  "crawl4ai>=0.8.6,<0.9",
  "docling>=2.90.0,<3",
  "docling-core>=2.73.0,<3.0.0",
  "docling-parse>=5.3.2,<6.0.0",
  "html5lib>=1.1,<2",
  "huggingface_hub<1",
  "langchain-text-splitters>=1.0.0,<2",
  "networkx>=3.4.2,<4",
  "nltk>=3.9.1,<4",
  "nlr-elm>=0.0.44,<1",
  "numpy>=2.4.3,<3",
  "openai>=2.34.0",
  "pandas>=2.2.3,<3",
  "pdftotext>=2.2.2,<3",
  "platformdirs>=4.5.1,<5",
  "playwright>=1.49.0,<1.52",  # version range required for c4ai"
  "pyjson5>=2.0.0,<3",
  "python-dotenv>=1.2.2,<2",
  "rich>=13.9.4,<14",
  "toml>=0.10.2,<0.11",
]

[project.optional-dependencies]
ocr = [
  "pytesseract>=0.3.13,<0.4"
]
dev = [
  "jupyter>=1.0.0,<1.1",
  "pipreqs>=0.4.13,<0.5",
  "ruff>=0.15.22,<0.16",
]
test = [
  "codespell>=2.4,<3",
  "flaky>=3.8.1,<4",
  "pytest>=9.0.3,<9.1",
  "pytest-asyncio>=1.3.0,<2",
  "pytest-cases>=3.10.1,<4",
  "pytest-cov>=7.1.0,<8",
  "pytest-mock>=3.15.1,<4",
  "pytest-profiling>=1.8.1,<2",
  "pytest-xdist>=3.8.0,<4",
  "snakeviz>=2.2.2,<3",
  "statsmodels>=0.14,<0.15",  # Wilson CI for eval metrics
  "tox>=4.52.1,<5",
]
doc = [
  "ghp-import>=2.1.0,<3",
  "myst-parser>=5.0.0,<6",
  "pydata-sphinx-theme>=0.17.1,<0.18",
  "sphinx-click>=6.2.0,<7",
  "sphinx-copybutton>=0.5.2,<0.6",
  "sphinx-tabs>=3.5,<4",
  "sphinxcontrib-mermaid>=2.0.1,<3",
]
anthropic = [
  "anthropic>=0.40.0,<0.41",
  "botocore>=1.35.72,<2",
  "boto3>=1.35.71,<2",
]
build = [
  "build>=1.2.2,<2",
  "pkginfo>=1.12.1.2,<2",
]

[project.scripts]
compass = "compass._cli.main:main"

[project.urls]
homepage = "https://github.com/NatLabRockies/COMPASS"
documentation = "https://natlabrockies.github.io/COMPASS/"
repository = "https://github.com/NatLabRockies/COMPASS"

[tool.setuptools.packages]
find = { include = ["compass*"] }

[tool.setuptools_scm]
fallback_version = "9999"
# Default git command uses "*[0-9]*", which breaks with cX.X.X tags.
# This spec below forces setuptools_scm to look strictly for vX.X.X tags
git_describe_command = [ "git", "describe", "--dirty", "--tags", "--long", "--match", "v*[0-9]*" ]
version_file = "compass/_version.py"

[tool.ruff]
line-length = 79

[tool.ruff.format]
quote-style = "double"
indent-style = "space"

[tool.ruff.lint]
preview = true
select = [
  "A", # flake8-builtins
  "B", # flake8-bugbear
  "ARG", # flake8-unused-arguments
  "BLE", # flake8-blind-except
  "C4", # flake8-comprehensions
  "C90", # mccabe
  "COM", # flake8-commas
  "D", # pydocstyle
  "E", # pycodestyle
  "EM", # flake8-errmsg
  "F", # Pyflakes
  "FURB", # refurb
  "FLY", #  flynt
  "G", # flake8-logging-format
  "I", # isort
  "INP", # flake8-no-pep420
  "ISC", # flake8-implicit-str-concat
  "LOG", # flake8-logging
  "N", # pep8-naming
  "NPY", # numpy-specific
  "PD", # pandas-vet
  "PERF", # Perflint
  "PIE", # flake8-pie
  "PL", # Pylint
  "PTH", # flake8-use-pathlib
  "PYI", # flake8-pyi
  "Q", # flake8-quotes
  "RET", # flake8-return
  "RSE", # flake8-raise
  "RUF", # Ruff-specific rules
  "S", # flake8-bandit
  "SLF", # flake8-self
  "SIM", # flake8-simplify
  "TID", # flake8-tidy-imports (absolute imports)
  "TRY", # tryceratops
  "UP", # pyupgrade
  "W", # Warning
]
ignore = [
  "no-explicit-stacklevel",
  "missing-trailing-comma",
  "undocumented-magic-method",
  "missing-blank-line-after-summary",
  "blank-line-after-function",
  "surrounding-whitespace",
  "missing-trailing-period",
  "non-imperative-mood",
  "first-word-uncapitalized",
  "reimplemented-starmap",
  "unsorted-imports",
  "collections-named-tuple",
  "mutable-class-default",
  "incorrectly-parenthesized-tuple-in-subscript",
  "non-empty-init-module",
  "assert",
  "hashlib-insecure-hash-function",
  "suspicious-pickle-import",
  "raise-vanilla-args",
]

[tool.ruff.lint.pycodestyle]
max-doc-length = 72

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

[tool.ruff.lint.per-file-ignores]
"__init__.py" = ["unused-import"]
"{tests,evals}/**.py" = [
  "unused-function-argument",
  "unused-method-argument",
  "undocumented-public-class",
  "undocumented-public-method",
  "raw-string-in-exception",
  "logging-f-string",
  "implicit-namespace-package",
  "non-lowercase-variable-in-function",
  "import-private-name",
  "too-many-locals",
  "too-many-statements",
  "magic-value-comparison",
  "no-self-use",
  "private-member-access",
  "raise-within-try",
  "doc-line-too-long",
]

[tool.ruff.lint.pylint]
max-args = 10
max-positional-args = 10

[tool.coverage.run]
branch = true

[tool.coverage.report]
# Regexes for lines to exclude from consideration
exclude_also = [
  # Have to re-enable the standard pragma
  "pragma: no cover",

  # Don't complain about missing debug-only code:
  "if self\\.debug",

  # Don't complain about repr methods
  "def __repr__",
  "def _repr_markdown_",

  # Don't complain about data I/O code
  "def load*",
  "def _load*",

  # Don't complain if tests don't hit defensive assertion code:
  "raise AssertionError",
  "raise NotImplementedError",

  # Don't complain if non-runnable code isn't run:
  "if __name__ == .__main__.:",

  # Don't complain about abstract methods, they aren't run:
  "@(abc\\.)?abstractmethod",

  # Don't complain about logging debugging functions
  "def print_logging_info*",
  "def __cls_name",
]

omit = [
  # omit test files
  "tests/*",
  # omit init files
  "__init__.py",
  # omit version file
  "_version.py",
  # omit log file copied over from other repo
  "compass/utilities/log.py",
  # omit config loading code for now
  "compass/utilities/config.py",
  # omit pixi files
  ".pixi/*",
]


[tool.codespell]
# Applied by the codespell workflow (.github/workflows/ci-codespell.yml)
# to compass/extraction, which is the primary area for new schemas.
# Add project-specific false positives to `ignore-words-list` below as a
# comma-separated list of lowercase words (uncomment when needed).
skip = "*.lock,*.min.js,*.svg,*.pdf,*.png,*.jpg,*.jpeg,*.gif,*.ico"
# ignore-words-list = "word1,word2"

[tool.pytest.ini_options]
addopts = '--durations=10 --disable-warnings -m "not evals"'
asyncio_mode="auto"
asyncio_default_fixture_loop_scope="function"
markers = [
  "evals: accuracy/quality evals (live billable LLM calls). Deselected by default. Run with `pixi run evals <name>` (dev dataset) or `pixi run evals <name> --held-out` (held-out dataset).",
]
testpaths = [
  "tests/python/unit",
  "tests/python/integration",
]