pyproject.toml

[build-system]
requires = ["hatchling>=1.26.3"]
build-backend = "hatchling.build"

[project]
name = "edgartools"
description = 'Navigate Edgar filings with ease'
readme = "README.md"
requires-python = ">=3.10"
license = "MIT"
keywords = ["sec", "edgar", "filings", "company", "python", "finance", "financial", "reports"]
authors = [
  { name = "Dwight Gunning", email = "dgunning@gmail.com" },
]
classifiers = [
  "Development Status :: 4 - Beta",
  "Programming Language :: Python",
  "Programming Language :: Python :: 3.10",
  "Programming Language :: Python :: 3.11",
  "Programming Language :: Python :: 3.12",
  "Programming Language :: Python :: 3.13",
  "Programming Language :: Python :: 3.14",
  "Programming Language :: Python :: Implementation :: CPython",
  "Programming Language :: Python :: Implementation :: PyPy",
  "License :: OSI Approved :: MIT License",
]
dependencies = [
  "httpx>=0.25.0",
  "pandas>=2.0.0",
  "tabulate>=0.9.0",
  "pyarrow>=17.0.0",
  "beautifulsoup4>=4.10.0",
  "lxml>=4.4",
  "rich>=13.8.0",
  "humanize>=4.0.0",
  "stamina>=24.2.0",
  "orjson>=3.6.0",
  "textdistance>=4.5.0",
  "rank_bm25>=0.2.1",
  "rapidfuzz>=3.5.0",
  "unidecode>=1.2.0",
  "pydantic>=2.0.0",
  "tqdm>=4.62.0",
  "nest-asyncio>=1.5.1",
  "jinja2>=3.1.0",
  "hishel==0.1.3",
  "httpxthrottlecache>=0.1.6",
]
dynamic = ["version"]

[project.optional-dependencies]
ai = [
    "mcp>=1.12.3",          # Model Context Protocol
    "tiktoken>=0.10.0",     # Token counting for OpenAI models
]

ai-dev = [
    "pytest-mock>=3.12.0",       # For mocking AI responses
    "responses>=0.24.0",         # HTTP response mocking
]

data = [
    "duckdb>=1.0.0",            # SQL interface for company datasets
]

test-harness = [
    "click>=8.0.0",              # CLI interface for test harness
]

# Cloud storage support via fsspec
cloud = [
    "fsspec>=2023.1.0",          # Core filesystem abstraction
]

s3 = [
    "fsspec>=2023.1.0",
    "s3fs>=2023.1.0",            # AWS S3, Cloudflare R2, MinIO
]

gcs = [
    "fsspec>=2023.1.0",
    "gcsfs>=2023.1.0",           # Google Cloud Storage
]

azure = [
    "fsspec>=2023.1.0",
    "adlfs>=2023.1.0",           # Azure Blob Storage
]

all-cloud = [
    "fsspec>=2023.1.0",
    "s3fs>=2023.1.0",
    "gcsfs>=2023.1.0",
    "adlfs>=2023.1.0",
]

[project.urls]
Documentation = "https://dgunning.github.io/edgartools/"
Issues = "https://github.com/dgunning/edgartools/issues"
Source = "https://github.com/dgunning/edgartools"

[project.scripts]
edgartools-mcp = "edgar.ai.mcp_server:main"
edgar-test = "tests.harness.cli:cli"

[tool.hatch.version]
path = "edgar/__about__.py"

[tool.hatch.build]
include = [
  "edgar/**/*.py",
  "edgar/**/templates/*.html",
  "edgar/**/docs/*.md",
  "edgar/ai/skills/core/*.md",
  "edgar/reference/data/*",
  "edgar/entity/data/*",
  "edgar/xbrl/standardization/*",
  "LICENSE.txt"
]
artifacts = [
  "LICENSE.txt"  # And this line
]

[tool.hatch.envs.default]
dependencies = [
  "pytest",
  "pytest-cov",
  "pytest-env",
  "pytest-xdist",
  "pytest-asyncio",
  "pytest-retry",
  "pytest-vcr",
  "vcrpy",
  "filelock",
  "pyinstrument",
  "pyright",
  "ruff",
  "tqdm",
  "xlsxwriter",
  "openpyxl",
  "jupyter",
  "freezegun==1.5.1",
  "mkdocs",
  "mkdocstrings[python]",
  "mkdocs-material"
]

features = [
    "ai",
    "ai-dev"
]

[tool.hatch.envs.default.scripts]
cov = "pytest --cov-report=term-missing --cov-config=pyproject.toml --cov=edgar --cov=tests --ignore=tests/legacy {args}"
no-cov = "cov --no-cov {args}"
lint = "ruff check edgar"
smoke-filings = "python tests/batch/batch_filings.py {args}"
# Test categorization commands (sequential execution)
test-fast = "pytest -m 'fast' {args}"
test-slow = "pytest -m 'slow' {args}"
test-network = "pytest -m 'network' {args}"
test-core = "pytest -m 'not (slow or network or performance or batch)' --ignore=tests/manual --ignore=tests/perf {args}"

# Parallel test commands (selective parallelization for SEC rate limit safety)
# - Fast tests: Full parallelization (no network calls)
# - Network/slow: Sequential (respects SEC rate limits)
# - Core: Limited parallelization (mixed tests)
test-fast-parallel = "pytest -n auto -m 'fast' {args}"
test-core-parallel = "pytest -n 2 -m 'not (slow or network or performance or batch)' --ignore=tests/manual --ignore=tests/perf {args}"
test-parallel-safe = "pytest -n auto -m 'fast' {args}"

# Parallel CI test strategy (skip regression tests for faster feedback)
#
# SELECTIVE PARALLELIZATION STRATEGY:
# 1. Marker-based: -m 'not regression' excludes tests marked as regression
# 2. Path-based: --ignore=tests/issues/regression excludes entire folder
# 3. Auto-marking: conftest.py automatically marks tests as fast/network based on filename
# 4. Selective parallel: Only parallelize tests safe from SEC rate limits
# 5. Combined coverage: CI combines coverage from fast/network/slow groups for 70% threshold
#
# Note: "core" group removed from CI - auto-marking leaves 0 tests without markers

test-ci-fast = "pytest -n auto --cov --cov-report=xml -m 'fast and not regression' {args}"
test-ci-network = "pytest --cov --cov-report=xml -m 'network and not slow and not regression' {args}"
test-ci-slow = "pytest --cov --cov-report=xml -m 'slow and not regression' {args}"
test-ci-all = "pytest --cov --cov-report=xml -m 'not regression' --ignore=tests/manual --ignore=tests/perf --ignore=tests/issues/regression {args}"

# Regression tests (run separately/on-demand for comprehensive bug prevention)
test-regression = "pytest --cov --cov-report=xml -m regression {args}"

# Other test categories  
test-batch = "pytest tests/batch/ -m 'batch' {args}"
test-reproduction = "pytest tests/issues/reproductions/ -m 'reproduction' {args}"
test-full = "pytest --ignore=tests/manual --ignore=tests/perf {args}"

[tool.hatch.envs.test]
dependencies = [
  "pytest",
  "pytest-cov",
  "pytest-env",
  "pytest-asyncio"
]

[[tool.hatch.envs.test.matrix]]
python = ["39", "310", "311", "312"]

[tool.coverage.run]
branch = true
parallel = true
omit = [
  "edgar/__about__.py",
  "tests/perf/*"
]

[tool.coverage.report]
exclude_lines = [
  "no cov",
  "if __name__ == .__main__.:",
  "if TYPE_CHECKING:",
]
# Coverage threshold moved to CI combined coverage check
# Individual test groups don't meet 70% alone, only combined


[tool.ruff]
line-length = 150
exclude = [
  ".git",
  "__pycache__",
  "build",
  "dist",
  "docs/source/conf.py",
  "tests",
]
extend-include = ["*.ipynb"]

[tool.ruff.lint]
select = ["F", "E", "W", "S", "B", "G", "N", "I001", "T", "PD", "C90"]
ignore = [
  "W291",
  "E501",
  "S608",
  "PD901",
  "C901",
  "S101"
]

[tool.ruff.lint.per-file-ignores]
"__init__.py" = ["F401"]

[dependency-groups]
dev = [
    "pre-commit>=4.3.0",
    "pytest-retry>=1.7.0",
]

[tool.pytest.ini_options]
testpaths = ["tests"]
env = [
    "EDGAR_IDENTITY=Dev Gunning developer-gunning@gmail.com",
]
asyncio_mode = "auto"
asyncio_default_fixture_loop_scope = "function"
retries = 2
retry_delay = 4
cumulative_timing = false
retry_outcome = "rerun"

# Pytest markers for test categorization
markers = [
    "fast: Fast tests that run quickly (< 0.1s each)",
    "slow: Slow tests that take significant time (> 1s each)",
    "network: Tests that require internet connectivity",
    "regression: Regression tests for specific GitHub issues",
    "batch: Batch processing tests",
    "performance: Performance and benchmarking tests",
    "reproduction: Tests that reproduce specific issues",
    "integration: Integration tests",
    "data_quality: Data quality validation tests",
    "vcr: Tests using VCR cassettes for recorded HTTP responses",
]

[tool.pyright]
# Suppress false positives for C extension modules
reportAttributeAccessIssue = "none"
# PyArrow is a C extension - type stubs are incomplete
ignoreExternal = true

[tool.ty]
# Suppress unresolved-attribute errors for pyarrow.compute (C extension with incomplete stubs)
# These are false positives - the methods exist at runtime
[[tool.ty.overrides]]
include = ["edgar/_filings.py"]
rules = { unresolved-attribute = "ignore" }