pyproject.toml

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

[project]
name = "bead"
version = "0.2.1"
description = "Lexicon and Template Collection Construction Pipeline for Acceptability and Inference Judgment Data"
authors = [{name = "Aaron Steven White", email = "aaron.white@rochester.edu"}]
readme = "README.md"
license = {text = "MIT"}
keywords = ["linguistics", "nlp", "experiments", "active-learning", "psycholinguistics"]
classifiers = [
    "Development Status :: 4 - Beta",
    "Intended Audience :: Science/Research",
    "License :: OSI Approved :: MIT License",
    "Operating System :: OS Independent",
    "Programming Language :: Python :: 3",
    "Programming Language :: Python :: 3.13",
    "Topic :: Scientific/Engineering :: Artificial Intelligence",
    "Topic :: Text Processing :: Linguistic",
    "Typing :: Typed",
]
requires-python = ">=3.13"
dependencies = [
    "pydantic>=2.0.0",
    "pyyaml>=6.0.0",
    "jinja2>=3.0.0",
    "uuid-utils>=0.7.0",
    "pandas>=2.0.0",
    "polars>=0.19.0",
    "langcodes>=3.3.0",
    "glazing>=0.2.0",
    "unimorph>=0.0.4",
    "torch>=2.0.0",
    "transformers>=4.35.0",
    "sentence-transformers>=2.0.0",
    "scikit-learn>=1.3.0",
    "statsmodels>=0.14.6",
    "scipy>=1.11.0",
    "click>=8.0.0",
    "rich>=13.0.0",
    "lark>=1.0.0",
    "psutil>=5.9.0",
    "typer>=0.9.0",
    "prompt-toolkit>=3.0.0",
    "krippendorff>=0.6.0",
    "peft>=0.6.0",
    "evaluate>=0.4.0",
    "accelerate>=0.25.0",
    "datasets>=2.14.0",
]

[project.urls]
Homepage = "https://github.com/FACTSlab/bead"
Documentation = "https://bead.readthedocs.io"
Repository = "https://github.com/FACTSlab/bead"
Issues = "https://github.com/FACTSlab/bead/issues"
Changelog = "https://github.com/FACTSlab/bead/blob/main/CHANGELOG.md"

[project.optional-dependencies]
dev = [
    "pytest>=7.4.0",
    "pytest-cov>=4.1.0",
    "pytest-mock>=3.11.0",
    "pytest-examples>=0.0.10",
    "pytest-codeblocks>=0.17.0",
    "pytest-env>=1.0.0",
    "ruff>=0.1.0",
    "pyright>=1.1.0",
    "pandas-stubs>=2.0.0",
]
api = [
    "openai>=1.0.0",
    "anthropic>=0.8.0",
    "google-generativeai>=0.3.0",
]
training = [
    "pytorch-lightning>=2.0.0",
    "tensorboard>=2.13.0",
]
stats = [
    "statsmodels>=0.14.0",
]
ui = [
    "textual>=0.40.0",
]
behavioral-analysis = [
    "slopit>=0.1.0",
]
tokenization = [
    "spacy>=3.7",
    "stanza>=1.8",
]

[project.scripts]
bead = "bead.cli.main:cli"

[tool.pytest.ini_options]
testpaths = ["tests", "docs/user-guide/cli"]
norecursedirs = ["tests/fixtures"]
addopts = ["-ra", "--strict-markers", "--cov=bead", "--cov-report=term-missing"]
markers = [
    "slow_model_training: marks tests that train ML models (deselect with '-m \"not slow_model_training\"')",
]


[tool.ruff]
line-length = 88
target-version = "py313"
exclude = ["gallery"]

[tool.ruff.lint]
select = ["E", "F", "I", "N", "D", "UP", "ANN", "B", "A", "C4", "PLC"]
ignore = ["ANN401"]

[tool.ruff.lint.per-file-ignores]
"tests/**" = ["ANN"]

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

[tool.pyright]
include = ["bead"]
typeCheckingMode = "strict"
pythonVersion = "3.13"
reportMissingTypeStubs = false
useLibraryCodeForTypes = true
# Disable reportUnknown* checks - industry standard for ML libraries with incomplete type stubs
# See: https://github.com/microsoft/pyright/discussions/6243
reportUnknownMemberType = false
reportUnknownVariableType = false
reportUnknownArgumentType = false
reportUnknownParameterType = false
reportUnknownLambdaType = false
exclude = [
    "tests/**",
    "gallery/**",
    "bead/active_learning/**",
    "bead/resources/adapters/unimorph.py",
    "bead/resources/adapters/glazing.py",
    "bead/templates/adapters/huggingface.py",
    "bead/templates/strategies.py",
    "bead/items/adapters/huggingface.py",
    "bead/items/adapters/sentence_transformers.py",
    "bead/items/adapters/openai.py",
    "bead/items/adapters/anthropic.py",
    "bead/items/adapters/google.py",
    "bead/items/adapters/togetherai.py",
    # jsPsych deployment uses complex JSON serialization with list invariance issues
    "bead/deployment/jspsych/**",
    "bead/simulation/**",
    # CLI modules use dynamic JSON config (JsonValue) extensively
    "bead/cli/**",
    # DSL uses lark which has no type stubs
    "bead/dsl/**",
    # Data collection modules use external APIs with incomplete stubs
    "bead/data_collection/**",
    # Resources modules use pandas/polars with incomplete type stubs
    "bead/resources/lexicon.py",
    "bead/resources/template_collection.py",
    "bead/resources/loaders.py",
    # Evaluation uses krippendorff which has no type stubs
    "bead/evaluation/**",
    # Behavioral uses slopit which has no type stubs
    "bead/behavioral/**",
    # Participants modules use polars with incomplete type stubs
    "bead/participants/collection.py",
    "bead/participants/merging.py",
    # Items cache has numpy ndarray type issues
    "bead/items/cache.py",
]