pyproject.toml

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

[project]
name = "bugbug"
dynamic = ["version"]
description = "ML tools for Mozilla projects"
license = "MPL-2.0"
requires-python = ">=3.14"
authors = [{ name = "Marco Castelluccio", email = "mcastelluccio@mozilla.com" }]
classifiers = [
    "Programming Language :: Python :: 3.14",
    "Programming Language :: Python :: 3 :: Only",
    "License :: OSI Approved :: Mozilla Public License 2.0 (MPL 2.0)",
]
dependencies = [
    "amqp~=5.3.1",
    "async-lru~=2.3.0",
    "beautifulsoup4>=4.14.3,<4.16.0",
    "boto3>=1.42.78,<1.44.0",
    "httpx~=0.28.1",
    "imbalanced-learn~=0.14.1",
    "langchain~=1.3.15",
    "langchain-anthropic>=1.3.4,<1.6.0",
    "langchain-classic~=1.0.3",
    "langchain-community~=0.4.1",
    "langchain-google-genai>=4.2.1,<4.4.0",
    "langchain-mistralai~=1.1.2",
    "langchain-openai>=1.1.14,<1.5.0",
    "langgraph>=1.2.5",
    "libmozdata~=0.2.12",
    "lmdb>=1.8.1,<2.4.0",
    "lxml-html-clean~=0.4.4",
    "markdown2~=2.5.5",
    "matplotlib>=3.10.8,<3.12.0",
    "mercurial~=7.2",
    "microannotate~=0.0.24",
    "mozci~=2.4.8",
    "numpy~=2.4.3",
    "orjson>=3.11.7,<3.14.0",
    "ortools~=9.15.6755",
    "pandas~=3.0.1",
    "psutil~=7.2.2",
    "pyOpenSSL>=0.14",
    "python-dateutil~=2.9.0.post0",
    "python-hglib~=2.6.2",
    "qdrant-client~=1.15.1",
    "ratelimit~=2.2.1",
    "requests>=2.33,<2.35",
    "requests-html~=0.10.0",
    "rs-parsepatch~=0.4.6",
    "searchfox>=0.20.3",
    "scikit-learn~=1.7.2",
    "scipy>=1.17.1,<1.19.0",
    "sendgrid~=6.12.5",
    "shap[plots]>=0.51,<0.53",
    "tabulate~=0.10.0",
    "taskcluster>=97.1,<103.1",
    "tenacity~=9.1.4",
    "tqdm>=4.67.3,<4.71.0",
    "unidiff~=0.7.5",
    "weave>=0.53.4",
    "whatthepatch~=1.0.7",
    "xgboost>=3.2,<3.5",
    "zstandard~=0.25.0",
]

[project.optional-dependencies]
nlp = [
    "spacy==3.8.16",
]
nn = []
perf-regression-predictor = [
    "torch==2.13.0",
    "transformers==5.15.0",
]

[dependency-groups]
test = [
    "coverage==7.16.2",
    "hypothesis==6.168.5",
    "igraph==1.0.0",
    "jsonschema==4.26.0",
    "pre-commit==4.6.2",
    "pytest==9.1.1",
    "pytest-asyncio==1.4.0",
    "pytest-cov==7.1.0",
    "pytest-responses==0.6.0",
    "responses==0.26.3",
]
spawn-pipeline = [
    "json-e==4.8.4",
    "pyyaml==6.0.3",
    "requests==2.34.2",
    "taskcluster==103.0.1",
]

[project.scripts]
bugbug-data-commits = "scripts.commit_retriever:main"
bugbug-data-bugzilla = "scripts.bug_retriever:main"
bugbug-data-test-scheduling-history = "scripts.test_scheduling_history_retriever:main"
bugbug-data-revisions = "scripts.revision_retriever:main"
bugbug-train = "scripts.trainer:main"
bugbug-check = "scripts.check:main"
bugbug-maintenance-effectiveness-indicator = "scripts.maintenance_effectiveness_indicator:main"
bugbug-microannotate-generate = "scripts.microannotate_generator:main"
bugbug-classify-commit = "scripts.commit_classifier:main"
bugbug-classify-bug = "scripts.bug_classifier:main"
bugbug-classify-comment = "scripts.comment_classifier:main"
bugbug-regressor-finder = "scripts.regressor_finder:main"
bugbug-retrieve-training-metrics = "scripts.retrieve_training_metrics:main"
bugbug-analyze-training-metrics = "scripts.analyze_training_metrics:main"
bugbug-check-all-metrics = "scripts.check_all_metrics:main"
bugbug-past-bugs-by-unit = "scripts.past_bugs_by_unit:main"
bugbug-testing-policy-stats = "scripts.testing_policy_stats:main"
bugbug-generate-landings-risk-report = "scripts.generate_landings_risk_report:main"
bugbug-shadow-scheduler-stats = "scripts.shadow_scheduler_stats:main"
bugbug-data-github = "scripts.github_issue_retriever:main"
bugbug-fixed-comments = "scripts.inline_comments_data_collection:main"
bugbug-ci-failures-retriever = "scripts.retrieve_ci_failures:main"
bugbug-try-pushes-retriever = "scripts.retrieve_try_pushes:main"
bugbug-validate-review-context = "bugbug.tools.code_review.review_context_schema:main"
bugbug-predict-perf-regression = "scripts.perf_regression_predictor:main"

[tool.hatch.version]
path = "VERSION"
pattern = "(?P<version>.+)"

[tool.hatch.build.targets.sdist]
include = ["/bugbug", "/scripts", "/VERSION"]

[tool.hatch.build.targets.wheel]
packages = ["bugbug", "scripts"]

[tool.uv]
# xgboost pulls in NCCL (~300 MB) for multi-GPU training, which we never use.
override-dependencies = ["nvidia-nccl-cu13; sys_platform == 'never'"]

[tool.uv.workspace]
members = ["http_service", "agents/*", "libs/*", "services/*", "tools/*"]
exclude = [
    # JavaScript package (no pyproject.toml).
    "services/hackbot-ui",
    # Not migrated to the workspace yet.
    "services/buildrepair",
]

[tool.uv.sources]
hackbot-runtime = { workspace = true }
agent-tools = { workspace = true }
torch = [
    { index = "pytorch-cpu", marker = "sys_platform == 'linux' or sys_platform == 'win32'" },
]

[[tool.uv.index]]
name = "pytorch-cpu"
url = "https://download.pytorch.org/whl/cpu"
explicit = true

[tool.ruff]
extend-exclude = ["data"]

[tool.ruff.lint]
select = ["E4", "E7", "E9", "F", "I", "T10", "CPY", "G", "D", "B", "SIM", "FAST", "ASYNC"]
# Ignore warnings about missing docstrings.
ignore = ["D100", "D101", "D102", "D103", "D104", "D105", "D107",
    # B905: zip() without explicit strict= needs case-by-case review to pick the right value (usually strict=True); tracked in #6916.
    "B905"
]

[tool.ruff.lint.per-file-ignores]
"http_service/**" = ["D"]
# Agents currently run sequentially, so blocking calls in async functions are harmless there; tracked in #7008.
"agents/**" = ["ASYNC"]
"libs/agent-tools/**" = ["ASYNC"]
"libs/hackbot-runtime/**" = ["ASYNC"]

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

[tool.ruff.lint.isort]
known-first-party = ["bugbug_http"]

[tool.codespell]
ignore-words-list = ["aFile", "thirdparty", "checkin"]