[project]
name = "iscc-sct"
version = "0.2.2"
description = "ISCC - Semantic Code Text"
authors = [{ name = "Titusz", email = "tp@piscc.io" }]
requires-python = ">=3.11"
readme = "README.md"
license = "Apache-2.0"
license-files = ["LICENSE"]
keywords = [
    "iscc",
    "text similarity",
    "cross lingual",
    "semantic similarity",
]
classifiers = [
    "Development Status :: 4 - Beta",
    "Intended Audience :: Developers",
    "Intended Audience :: Science/Research",
    "Natural Language :: English",
    "Operating System :: OS Independent",
    "Programming Language :: Python :: 3.11",
    "Programming Language :: Python :: 3.12",
    "Programming Language :: Python :: 3.13",
    "Programming Language :: Python :: 3.14",
    "Topic :: Text Processing",
    "Topic :: Text Processing :: General",
    "Topic :: Text Processing :: Indexing",
    "Topic :: Text Processing :: Linguistic",
    "Topic :: Scientific/Engineering :: Artificial Intelligence",
    "Topic :: Multimedia :: Graphics",
    "Topic :: Scientific/Engineering :: Image Recognition",
    "Topic :: Scientific/Engineering :: Information Analysis",
    "Topic :: Software Development :: Libraries :: Python Modules",
    "Topic :: Software Development :: Libraries",
    "Topic :: System :: Archiving",
    "Topic :: System :: Clustering",
    "Topic :: System :: Distributed Computing",
]
dependencies = [
    # >=0.32.0 ships the upstream "avoid sizing whole distant split sections" fix and releases
    # the GIL during native chunking. It mitigates - but does not eliminate - the issue #24
    # super-linear cost on separator-free spans, so the chunking guard below is retained.
    "semantic-text-splitter>=0.32.0",
    "loguru",
    "blake3",
    "platformdirs",
    "tokenizers",
    "pydantic-settings",
    "charset-normalizer",
    "numpy",
    "pybase64",
    "certifi",
    "pyyaml",
    "pydantic",
]

# The onnxruntime variant wheels unpack into the same site-packages directory and clobber each
# other, so exactly one of the cpu/gpu extras must be installed (see issue #23).
[project.optional-dependencies]
cpu = ["onnxruntime"]
gpu = ["onnxruntime-gpu"]
demo = ["gradio"]

[project.urls]
Homepage = "https://iscc.codes"
Repository = "https://github.com/iscc/iscc-sct"
Documentation = "https://github.com/iscc/iscc-sct"
Changelog = "https://github.com/iscc/iscc-sct/blob/main/CHANGELOG.md"
"Bug Tracker" = "https://github.com/iscc/iscc-sct/issues"
Twitter = "https://twitter.com/iscc_foundation"
Donate = "https://iscc.foundation/support"

[project.scripts]
iscc-sct = "iscc_sct.cli:main"

[dependency-groups]
test = [
    "pytest",
    "coverage",
    "pytest-cov",
    "onnxruntime",  # dev/test environments use CPU inference
]
dev = [
    "poethepoet",
    "ruff",
    "mdformat-mkdocs[recommended]",
    "mdformat-gfm-alerts",
    "prek>=0.4.4",
    "zensical>=0.0.21",
    "mkdocstrings-python>=1.16.0",
]

[tool.uv]
default-groups = "all"
conflicts = [
    [{ extra = "cpu" }, { extra = "gpu" }],
    [{ group = "test" }, { extra = "gpu" }],
]
# The HF Space installs requirements.txt alongside gradio[oauth,mcp], whose mcp extra pins
# pydantic<=2.12.5 — keep the locked pydantic co-installable with the Space runtime.
constraint-dependencies = ["pydantic<=2.12.5"]

[tool.uv.build-backend]
module-root = ""

[build-system]
requires = ["uv_build>=0.11.20,<0.12.0"]
build-backend = "uv_build"

[tool.ruff]
line-length = 100

[tool.ruff.lint]
# F401: imports referenced only in PEP 484 type comments are invisible to ruff
ignore = ["F401"]

[tool.ruff.lint.per-file-ignores]
# F403: star imports intentionally re-export the public API
"iscc_sct/__init__.py" = ["F403"]

[tool.ruff.format]
line-ending = "lf"

[tool.coverage.run]
omit = ["iscc_sct/dev.py", "tests/", "iscc_sct/demo.py"]
patch = ["subprocess"]

[tool.poe.tasks]
format-code = { cmd = "ruff format", help = "Code style formating with ruff" }
format-markdown = { cmd = "mdformat --wrap 100 --end-of-line lf --exclude docs/reference/api.md README.md CHANGELOG.md docs", help = "Markdown formating with mdformat" }
format-yml = { script = "iscc_sct.dev:format_yml", help = "Format YML files"}
convert-lf = { script = "iscc_sct.dev:convert_lf", help = "Convert line endings to LF"}
# Erase any stale .coverage.* parallel data files left by an aborted run before measuring, so the
# 100% gate never inherits orphaned subprocess data (patch=["subprocess"]) and report a spurious
# <100% (issue: leftover .coverage.GOLEM.pid* files + failing coverage gate).
test.sequence = [
    { cmd = "coverage erase" },
    { cmd = "pytest --cov=iscc_sct --cov-fail-under=100" },
]
test.help = "Run tests with coverage"
docs-serve = { cmd = "zensical serve", help = "Serve the documentation site locally" }
docs-build.sequence = [
    { cmd = "zensical build" },
    { cmd = "python scripts/gen_llms_full.py" },
]
docs-build.help = "Build the docs site and generate llms-full.txt"
precommit = { cmd = "uv run prek run", help = "Run prek hooks on changed files" }
export-requirements = { cmd = "uv export --no-default-groups --extra cpu --no-emit-project --no-hashes --no-annotate --no-header --output-file requirements.txt", help = "Export requirements.txt from uv.lock for HF Space" }
update-dependencies = { cmd = "uv sync --upgrade --extra demo", help = "Update dependencies" }
all = ["format-code", "format-markdown", "format-yml", "convert-lf", "export-requirements", "test"]
update = ["update-dependencies", "all"]