vgi-scikit-bio/pyproject.toml at main · Query-farm/vgi-scikit-bio · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
[project]
name = "vgi-scikit-bio"
# Single-sourced from vgi_scikit_bio/__init__.py (see [tool.hatch.version]) so the
# published wheel, the version the worker advertises over VGI
# (implementation_version / data_version), and the GitHub Release tag are always
# the same number. Bump __version__ there before tagging a release.
dynamic = ["version"]
description = "A VGI (Vector Gateway Interface) worker exposing scikit-bio sequence analysis, diversity, ordination, and phylogenetics to DuckDB/SQL"
readme = "README.md"
license = "MIT"
license-files = ["LICENSE"]
authors = [{ name = "Rusty Conover", email = "rusty@query.farm" }]
maintainers = [{ name = "Query Farm LLC", email = "hello@query.farm" }]
keywords = ["vgi", "duckdb", "arrow", "scikit-bio", "bioinformatics", "sql"]
classifiers = [
    "Development Status :: 4 - Beta",
    "Intended Audience :: Developers",
    "Intended Audience :: Science/Research",
    "Programming Language :: Python :: 3",
    "Programming Language :: Python :: 3 :: Only",
    "Programming Language :: Python :: 3.13",
    "Topic :: Database",
    "Topic :: Database :: Database Engines/Servers",
    "Topic :: Scientific/Engineering :: Bio-Informatics",
    "Operating System :: OS Independent",
]
requires-python = ">=3.13"
dependencies = [
    "vgi-python[http]>=0.17.0",
    "scikit-bio>=0.6",
    "numpy",
    "pandas",
    "scipy",
]

[project.optional-dependencies]
# HTTP-serving extras. Sentry telemetry and OAuth auth only apply to the HTTP
# server (under stdio, DuckDB owns the pipe), so they live here rather than in
# the base deps: the published Docker image installs `.[serve]`, while a plain
# `pip install vgi-scikit-bio` for an on-host stdio worker stays lean.
serve = [
    "vgi-python[http,oauth]>=0.17.0",
    "vgi-rpc[sentry]>=0.26.0",
    "authlib",
]

# Console scripts. `vgi-scikit-bio` runs the stdio transport (DuckDB spawns it);
# `vgi-scikit-bio-http` runs the HTTP server. Once installed you can launch either
# directly, e.g. `uvx vgi-scikit-bio`.
[project.scripts]
vgi-scikit-bio = "vgi_scikit_bio.worker:main"
vgi-scikit-bio-http = "vgi_scikit_bio.worker:main_http"

[project.urls]
Homepage = "https://query.farm"
Repository = "https://github.com/Query-farm/vgi-scikit-bio"
Issues = "https://github.com/Query-farm/vgi-scikit-bio/issues"

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

# Read the package version from the `__version__` literal in the package itself,
# so it is the single source of truth (see [project].dynamic above).
[tool.hatch.version]
path = "vgi_scikit_bio/__init__.py"

# The installable worker is the `vgi_scikit_bio` package (entry points live in
# vgi_scikit_bio.worker). The repo-root scikit_bio_worker.py / serve.py are
# dev/container shims and intentionally not shipped in the wheel.
[tool.hatch.build.targets.wheel]
packages = ["vgi_scikit_bio"]

[tool.hatch.build.targets.sdist]
include = [
    "vgi_scikit_bio/",
    "scikit_bio_worker.py",
    "serve.py",
    "conftest.py",
    "tests/",
    "test/",
    "ci/",
    "README.md",
    "CLAUDE.md",
    "LICENSE",
]

[dependency-groups]
dev = [
    "pytest",
    "pyarrow",
    "ruff",
    "pydoclint",
    "mypy",
]

[tool.pytest.ini_options]
addopts = "-q"
testpaths = ["tests"]

[tool.ruff]
line-length = 120
target-version = "py313"

[tool.ruff.lint]
select = ["E", "F", "I", "UP", "B", "C4", "SIM", "D"]
ignore = [
    "E741",  # ambiguous variable name (x, y are conventional)
    "B008",  # function call in argument default (Arg(...) defaults are intentional)
    "N801",  # class-name casing (TEST classmethod stand-ins are intentional)
]

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

[tool.ruff.lint.per-file-ignores]
# Tests are not part of the documented public surface: skip docstring rules.
"tests/*" = ["E402", "D"]

[tool.ruff.format]
quote-style = "double"

[tool.pydoclint]
# Docstring consistency gate (complements ruff's `D` rules, which only check
# docstring *shape*, not whether documented args/attributes match the code).
# Run via `uv run pydoclint vgi_scikit_bio/ scikit_bio_worker.py serve.py`; also
# enforced inside the pytest suite by tests/test_docstrings.py.
style = "google"
# Types live in signatures, not docstrings (Google style).
arg_type_hints_in_docstring = false
check_return_types = false
check_yield_types = false
# Validate that an `Attributes:` section matches the actual dataclass fields.
check_class_attributes = true
# Raises are documented indirectly; don't gate on DOC5xx.
skip_checking_raises = true
# Document constructor params in the class docstring, not __init__.
allow_init_docstring = true

[tool.mypy]
python_version = "3.13"
strict = true
warn_return_any = true
warn_unused_ignores = true

[[tool.mypy.overrides]]
# The scientific stack ships no (or partial) type information; don't fail on
# missing stubs for these third-party imports.
module = ["skbio.*", "scipy.*", "sklearn.*", "pandas.*", "numpy.*"]
ignore_missing_imports = true