pyproject.toml

[project]
name = "excoder"
version = "1.0.0"
description = "ExCoder: retrofitting Java code to raise exceptions, using LLMs guided by static and dynamic analysis (ISSRE 2026)"
readme = "README.md"
requires-python = ">=3.11"

# Only what the four pipelines in scripts/ actually import.  The versions are
# the ones the paper's results were produced with; loosen them at your own risk,
# since numpy/pandas in particular affect how the statistics are aggregated.
dependencies = [
    # dataset handling, CLIs and utilities
    "seutil>=0.8.12",
    "jsonargparse>=4.36.0",
    "tqdm>=4.67.1",
    "requests>=2.32.3",
    "typing-extensions>=4.12.2",
    # Java parsing / analysis
    "javalang~=0.13.0",
    "tree-sitter==0.24.0",
    "tree-sitter-java==0.23.0",
    "xmltodict>=0.14.2",
    # numerics, statistics and figures
    "numpy~=1.25.0",
    "pandas~=2.1.0",
    "scipy>=1.15.1",
    "matplotlib>=3.10.0",
    "matplotlib-venn>=1.1.2",
    # LLM backends: llama.cpp is driven through the external `llama-server`
    # binary, so only the tokenizer and the weight download are Python-side.
    "transformers>=4.50.0",
    "huggingface-hub>=0.28.1",
    "azure-ai-inference>=1.0.0b9",
]

[project.optional-dependencies]
dev = [
    "ruff>=0.6.0",
]

[build-system]
requires = ["setuptools>=68"]
build-backend = "setuptools.build_meta"

[tool.setuptools]
package-dir = { "" = "python" }

[tool.setuptools.packages.find]
where = ["python"]
include = ["throwgen*", "etestgen*"]

[tool.pyright]
include = ["./python"]
typeCheckingMode = "basic"
reportUnusedCallResult = "none"

[tool.ruff]
line-length = 88

[tool.ruff.lint]
select = [
    "E",   # pycodestyle
    "F",   # pyflakes
    "UP",  # pyupgrade
    "B",   # flake8-bugbear
    "SIM", # flake8-simplify
    "I",   # isort
]
ignore = ["I001"]