pyproject.toml

[build-system]
requires = ["setuptools>=61.0", "wheel"]
build-backend = "setuptools.build_meta"

[project]
name = "llm-bot-pipeline"
version = "0.1.0"
description = "Server-side query fan-out session reporting for LLM bot traffic analysis"
readme = "README.md"
requires-python = ">=3.10"
license = {text = "AGPL-3.0"}
authors = [
    {name = "Data Team"}
]
dependencies = [
    "cloudflare>=3.0.0",
    "defusedxml>=0.7.1",
    "pandas>=2.0.0",
    "numpy>=1.24.0",
    "openpyxl>=3.1.0",
    "python-dateutil>=2.8.2",
    "pyyaml>=6.0.0",
    "httpx>=0.24.0",
    "requests>=2.28.0",
]

[project.optional-dependencies]
gcp = [
    "google-cloud-bigquery>=3.20.0",
    "google-api-core>=2.18.0",
    "google-cloud-monitoring>=2.0.0",
    "google-cloud-secret-manager>=2.16.0",
]
ml = [
    "scikit-learn>=1.3.0",
    "scipy>=1.11.0",
    "sentence-transformers>=2.2.0",
]
monitoring = [
    "apscheduler>=3.10.0",
    "prometheus-client>=0.19.0",
]
dev = [
    "black>=24.0.0",
    "isort>=5.13.0",
    "bandit>=1.7.0",
    "mypy>=1.0.0",
    "pre-commit>=3.0.0",
    "pytest>=8.0.0",
    "pytest-benchmark>=4.0.0",
    "pytest-cov>=4.1.0",
    "pytest-timeout>=2.0.0",
    "freezegun>=1.2.0",
]
viz = [
    "matplotlib>=3.7.0",
]

[tool.coverage.run]
source = ["src/llm_bot_pipeline"]
branch = true
omit = [
    "*/tests/*",
    "*/__pycache__/*",
    # Omitted: heavy external-SDK dependencies with no unit-testable seams
    "*/monitoring/*",
    "*/infrastructure/*",
    "*/cloudflare/*",
]

[tool.coverage.report]
fail_under = 65
show_missing = true
exclude_lines = [
    "pragma: no cover",
    "def __repr__",
    "raise NotImplementedError",
    "if TYPE_CHECKING:",
    "if __name__ == .__main__.:",
]

[tool.setuptools.packages.find]
where = ["src"]

[tool.black]
line-length = 88
target-version = ["py310"]
include = '\.pyi?$'

[tool.isort]
profile = "black"
line_length = 88
known_first_party = ["llm_bot_pipeline"]

[tool.bandit]
exclude_dirs = ["tests"]
skips = [
    "B104",  # 0.0.0.0 bind - intentional for Cloud Run
    "B105",  # "pass" - status string, not password
    "B110",  # try/except pass - intentional in some scripts
    "B310",  # urllib.urlopen - webhook URLs from config
    "B311",  # random - sample data generation, not crypto
    "B404",  # subprocess - gcloud, sops required
    "B405",  # xml.etree - defusedxml used for untrusted input
    "B603",  # subprocess - list args, no shell
    "B607",  # partial path - gcloud/sops in PATH
    "B608",  # SQL f-strings - BigQuery/SQLite table names from config
]

[tool.mypy]
python_version = "3.10"
warn_return_any = true
warn_unused_configs = true
ignore_missing_imports = true
exclude = [
    "tests/",
    "scripts/",
]

# Relax checks for modules with many third-party or complex type issues
[[tool.mypy.overrides]]
module = [
    "llm_bot_pipeline.config.sops_loader",
    "llm_bot_pipeline.sitemap.parser",
    "llm_bot_pipeline.infrastructure.cloud_run_setup",
    "llm_bot_pipeline.infrastructure.bigquery_setup",
    "llm_bot_pipeline.infrastructure",
    "llm_bot_pipeline.monitoring.metrics_export",
    "llm_bot_pipeline.ingestion.providers.gcp_cdn.adapter",
    "llm_bot_pipeline.ingestion.providers.fastly.adapter",
    "llm_bot_pipeline.ingestion.providers.aws_alb.adapter",
    "llm_bot_pipeline.ingestion.providers.akamai.adapter",
    "llm_bot_pipeline.cloudflare.logpush_job",
    "llm_bot_pipeline.cloudflare.logpull",
]
disable_error_code = [
    "import-untyped",
    "attr-defined",
    "no-any-return",
    "arg-type",
    "union-attr",
    "assignment",
    "return-value",
    "return",
    "index",
    "dict-item",
    "misc",
    "exit-return",
]

# Additional modules with type issues to address incrementally
[[tool.mypy.overrides]]
module = [
    "llm_bot_pipeline.utils.bot_classifier",
    "llm_bot_pipeline.monitoring.scheduler",
    "llm_bot_pipeline.monitoring.data_quality",
    "llm_bot_pipeline.monitoring.pipeline_metrics",
    "llm_bot_pipeline.ingestion.parsers.w3c_parser",
    "llm_bot_pipeline.ingestion.parsers.csv_parser",
    "llm_bot_pipeline.storage.sqlite_backend",
    "llm_bot_pipeline.storage.bigquery_backend",
    "llm_bot_pipeline.research.temporal_analysis",
    "llm_bot_pipeline.research.semantic_embeddings",
    "llm_bot_pipeline.research.experiment_runner",
    "llm_bot_pipeline.config.secret_manager",
    "llm_bot_pipeline.health_server",
    "llm_bot_pipeline.reporting.session_storage_writer",
    "llm_bot_pipeline.reporting.local_dashboard_queries",
    "llm_bot_pipeline.reporting.local_aggregations",
    "llm_bot_pipeline.pipeline.local_pipeline",
    "llm_bot_pipeline.pipeline.router",
]
disable_error_code = [
    "return-value",
    "attr-defined",
    "union-attr",
    "assignment",
    "arg-type",
    "no-any-return",
    "var-annotated",
    "misc",
    "index",
    "dict-item",
    "return",
    "exit-return",
]

[tool.pytest.ini_options]
testpaths = ["tests"]
python_files = ["test_*.py"]
python_functions = ["test_*"]
timeout = 60
markers = [
    "ingestion: E2E ingestion stage test",
    "transform: E2E transform stage test",
    "aggregation: E2E aggregation stage test",
    "query: E2E query stage test",
    "bigquery: requires BigQuery backend (skipped in public CI)",
]