pyproject.toml

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

[project]
name = "llm_topic_modelling"
version = "0.14.0"
description = "Generate thematic summaries from open text in tabular data files with a large language model."
requires-python = ">=3.10"
readme = "README.md"
authors = [
    { name = "Sean Pedrick-Case", email = "spedrickcase@lambeth.gov.uk" },
]
maintainers = [
    { name = "Sean Pedrick-Case", email = "spedrickcase@lambeth.gov.uk" },
]
license = "MIT"
keywords = [
    "topic-modelling",
    "topic-modeling",
    "llm",
    "large-language-models",
    "thematic-analysis",
    "text-analysis",
    "nlp",
    "natural-language-processing",
    "text-summarization",
    "text-summarisation",
    "thematic-summaries",
    "gradio",
    "data-analysis",
    "tabular-data",
    "excel",
    "csv",
    "open-text",
    "text-mining"
]
classifiers = [
    "Development Status :: 4 - Beta",
    "Intended Audience :: Developers",
    "Intended Audience :: Science/Research",
    "Intended Audience :: Information Technology",
    "Topic :: Scientific/Engineering :: Artificial Intelligence",
    "Topic :: Text Processing :: Linguistic",
    "Topic :: Text Processing :: Markup",
    "Topic :: Scientific/Engineering :: Information Analysis",
    "Programming Language :: Python :: 3",
    "Programming Language :: Python :: 3.10",
    "Programming Language :: Python :: 3.11",
    "Programming Language :: Python :: 3.12",
    "Programming Language :: Python :: 3.13",
]

dependencies = [    
    "gradio<=6.20.0",
    "transformers<=5.12.0",
    "spaces",
    "boto3<=1.43.35",
    "pandas<=2.3.3",
    "pyarrow<=23.0.1",
    "openpyxl<=3.1.5",
    "markdown<=3.10.2",
    "tabulate<=0.9.0",
    "lxml<=6.1.1",
    "google-genai<=1.73.0",
    "openai<=2.31.0",
    "html5lib<=1.1",
    "beautifulsoup4<=4.14.3",
    "rapidfuzz<=3.14.5",
    "python-dotenv<=1.2.2",
    "pyexcel<=0.7.4"
]

[project.optional-dependencies]
dev = ["pytest"]
test = ["pytest", "pytest-cov"]

# Extra dependencies for VLM models
# For torch you should use --index-url https://download.pytorch.org/whl/cu129. Additionally installs the unsloth package
torch = [
    "torch<=2.9.1",
    "torchvision",
    "accelerate",
    "bitsandbytes",
    "unsloth<=2026.6.9",
    "unsloth_zoo<=2026.6.9",
    "timm",
    "xformers"
]

# Support for llama-cpp-python is experimental - I have not directly tested the latest versions of llamacpp against the app
llamacpp = [
    "llama-cpp-python>=0.3.31",
]

# Run Gradio as an mcp server
mcp = [
    "gradio[mcp]<=6.20.0"
]

[project.urls]
Homepage = "https://github.com/seanpedrick-case/llm_topic_modelling"
repository = "https://github.com/seanpedrick-case/llm_topic_modelling"

[tool.setuptools]
packages = ["tools"]
py-modules = ["app"]

# Configuration for Ruff linter:
[tool.ruff]
line-length = 88

[tool.ruff.lint]
select = ["E", "F", "I"]
ignore = [
    "E501",  # line-too-long (handled with Black)
    "E402",  # module-import-not-at-top-of-file (sometimes needed for conditional imports)
]

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

# Configuration for a Black formatter:
[tool.black]
line-length = 88
target-version = ['py310']

# Configuration for pytest:
[tool.pytest.ini_options]
filterwarnings = [
    "ignore::DeprecationWarning:click.parser",
    "ignore::DeprecationWarning:weasel.util.config",
    "ignore::DeprecationWarning:builtin type",
    "ignore::DeprecationWarning:websockets.legacy",
    "ignore::DeprecationWarning:websockets.server",
    "ignore::DeprecationWarning:spacy.cli._util",
    "ignore::DeprecationWarning:weasel.util.config",
    "ignore::DeprecationWarning:importlib._bootstrap",
]
testpaths = ["test"]
python_files = ["test_*.py", "*_test.py"]
python_classes = ["Test*"]
python_functions = ["test_*"]
addopts = [
    "-v",
    "--tb=short",
    "--strict-markers",
    "--disable-warnings",
]