pyproject.toml

[project]
name = "proteinfoundation"
version = "1.1.0"
description = "Proteina-Complexa"
requires-python = ">=3.12"
dependencies = [
    # Core utilities
    "python-dotenv==1.0.1",
    "einops==0.6",
    "dm-tree==0.1.8",
    "loguru==0.7.2",
    "numpy>=1.23.5",
    "tqdm==4.66.4",
    "joblib==1.4.2",
    "rich==14.0.0",
    "omegaconf",
    "pydantic",
    "rich-click",
    "multipledispatch",
    "wget==3.2",
    
    # ML/Deep Learning
    "lightning>=2.5.0,<2.6", #2.6.1 caused loading weight issue
    "hydra-core==1.3.1",
    "ml-collections==0.1.1",
    "jaxtyping",
    "loralib",
    "transformers>=5.3.0", #"transformers==4.48.3",
    
    # Data processing
    "pandas>=2.3.0",
    "scipy==1.12.0",
    "h5py",
    "xarray",
    "modin[ray]",
    "deepdiff>=8.6.1",
    
    # Visualization
    "plotly",
    "seaborn",
    "scikit-learn",
    
    # Bioinformatics
    "biopandas==0.5.1",
    "biopython",
    "biotite>=0.41.0", # it gets later reset to 1.4.0 in the build_uv_env.sh script but it causes a conflict if set here
    "cpdb-protein==0.2.0",
    "mdtraj==1.10.2",
    "ProDy==2.6.1", #error with 2.4.1 
    
    # Experiment tracking
    "wandb==0.23.1",
    
]

[project.optional-dependencies]
dev = [
    "ruff>=0.15.0",
    "ipdb",
]

[project.scripts]
complexa = "proteinfoundation.cli.cli_runner:main"
complexa-download = "proteinfoundation.cli.cli_runner:download_main"
complexa-target = "proteinfoundation.cli.target_cli:main"

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

# Allow direct references (git URLs) in dependencies
[tool.hatch.metadata]
allow-direct-references = true

[tool.uv]
extra-index-url = [
    "https://download.pytorch.org/whl/cu126",
]
# Package configuration - using src/ layout
# Main package lives in src/proteinfoundation/
# Without installation, add src/ to PYTHONPATH for imports to work
[tool.hatch.build.targets.wheel]
packages = [
    "src/proteinfoundation",  # Main package in src/ layout
    "community_models",
    "script_utils",
]

# Source mappings:
# - "src" = "" strips src/ prefix so src/proteinfoundation becomes proteinfoundation
# - community_models subpackages exposed as top-level imports
[tool.hatch.build.targets.wheel.sources]
"src" = ""
"community_models/openfold" = "openfold"
"community_models/colabdesign" = "colabdesign"
"community_models/ProteinMPNN" = "ProteinMPNN"
"community_models/LigandMPNN" = "LigandMPNN"

[tool.ruff]
src = ["src"]
line-length = 120
indent-width = 4
target-version = "py312"
exclude = [
    ".bzr",
    ".direnv",
    ".eggs",
    ".git",
    ".git-rewrite",
    ".hg",
    ".ipynb_checkpoints",
    ".mypy_cache",
    ".nox",
    ".pants.d",
    ".pyenv",
    ".pytest_cache",
    ".pytype",
    ".ruff_cache",
    ".svn",
    ".tox",
    ".venv",
    ".vscode",
    "__pypackages__",
    "_build",
    "buck-out",
    "build",
    "dist",
    "node_modules",
    "site-packages",
    "env",
    ".venv",
    "venv",
    "data",
    "dev.py"
]

[tool.ruff.format]
# Like Black, use double quotes for strings.
quote-style = "double"

# Like Black, indent with spaces, rather than tabs.
indent-style = "space"

# Like Black, respect magic trailing commas.
skip-magic-trailing-comma = false

# Like Black, automatically detect the appropriate line ending.
line-ending = "auto"

# Enable auto-formatting of code examples in docstrings. Markdown,
# reStructuredText code/literal blocks and doctests are all supported.
docstring-code-format = true
docstring-code-line-length = 88

[tool.ruff.lint]
# Allow unused variables when underscore-prefixed.
dummy-variable-rgx = "^(_+|(_+[a-zA-Z0-9_]*[a-zA-Z0-9]+?))$"

# Enable a comprehensive set of rules for better code quality
select = [
    "E",   # pycodestyle errors
    "F",   # pyflakes
    "W",   # pycodestyle warnings
    "I",   # isort
    "N",   # pep8-naming
    #"D",   # pydocstyle
    "UP",  # pyupgrade -- use the latest supported python syntax
    "B",   # flake8-bugbear
    "C4",  # flake8-comprehensions
    "SIM", # flake8-simplify
    "NPY", # fix deprecated numpy functions / types
    "RUF", # Ruff-specific rules
    "T10", # catch accidental debugger breakpoints in code
    "ANN"
]

# Ignore specific rules that might conflict with our style or are too strict
ignore = [
    # --- Annotations (too noisy for ML research code) ---
    "ANN001",  # missing type annotation for function arg
    "ANN002",  # missing annotation for *args
    "ANN003",  # missing annotation for **kwargs
    "ANN201",  # missing return type for public function
    "ANN202",  # missing return type for private function
    "ANN204",  # missing return type for __init__
    "ANN205",  # missing return type for static method
    "ANN206",  # missing return type for class method
    "ANN401",  # disallow Any type

    # --- Naming (ML math conventions: X, T, N, SO3, SE3) ---
    "N801",    # invalid class name
    "N802",    # uppercase function name
    "N803",    # uppercase argument name
    "N805",    # invalid first argument name for method
    "N806",    # uppercase variable in function
    "N812",    # lowercase imported as non-lowercase
    "N816",    # mixed-case variable in global scope

    # --- Import ordering / structure ---
    "E402",    # module import not at top — common for conditional/sys.path imports
    "F722",    # forward annotation syntax error — false positives from jaxtyping

    # --- Style preferences (not bugs, just alternatives) ---
    "B006",    # mutable default arguments — common in ML config patterns
    "B007",    # unused loop control variable — use _ convention gradually
    "B023",    # function uses loop variable — common in closures
    "D100",    # missing docstring in public module
    "D203",    # 1 blank line required before class docstring
    "D212",    # multi-line docstring summary should start at first line
    "E501",    # line too long — handled by formatter
    "E731",    # lambda assignment — style preference
    "E741",    # ambiguous variable name (l, I, O) — standard in math/ML
    "NPY002",  # legacy np.random — widespread in codebase
    "SIM102",  # collapsible if — can reduce readability
    "SIM105",  # use contextlib.suppress — style preference
    "SIM108",  # ternary instead of if-else — reduces readability
    "SIM118",  # use `in dict` instead of `in dict.keys()` — style

    # --- Ruff-specific (noisy for existing codebase) ---
    "RUF002",  # ambiguous unicode character in docstring
    "RUF005",  # collection literal concatenation — `a + [b]` is clear
    "RUF013",  # implicit optional — `x: int = None` should be Optional[int]
    "RUF059",  # unused unpacked variable

    # --- Comprehension style (not bugs) ---
    "C414",    # unnecessary double cast or process
    "C416",    # unnecessary comprehension
    "C417",    # unnecessary map
]

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

[tool.ruff.lint.per-file-ignores]
"__init__.py" = ["F401"]  # Ignore unused imports in __init__.py files