pyproject.toml

[build-system]
requires = ["setuptools>=68.0", "wheel", "Cython>=3.0.0"]
build-backend = "setuptools.build_meta"

[project]
name = "myspellchecker"
version = "1.8.0"
description = "Myanmar (Burmese) text intelligence library — spell checking, grammar validation, dictionary building, and AI model training"
readme = "README.md"
# NOTE: Codebase uses Python 3.10+ syntax (PEP 604 unions like `str | None`).
requires-python = ">=3.10"
license = {file = "LICENSE"}
authors = [
    {name = "Thet Twe Aung", email = "thettweaung@gmail.com"}
]
keywords = ["myanmar", "burmese", "spell-checker", "nlp", "language-processing"]
classifiers = [
    "Development Status :: 5 - Production/Stable",
    "Intended Audience :: Developers",
    "License :: OSI Approved :: MIT License",
    "Topic :: Text Processing :: Linguistic",
    "Programming Language :: Python :: 3",
    "Programming Language :: Python :: 3.10",
    "Programming Language :: Python :: 3.11",
    "Programming Language :: Python :: 3.12",
    "Programming Language :: Python :: 3.13",
    "Programming Language :: Python :: 3.14",
]

dependencies = [
    # Core dependencies (required for spell checking)
    "numpy>=1.22.0",  # CVE-2021-41496 fixed in 1.22.0
    "python-crfsuite>=0.9.7",
    "rich>=13.0.0",  # Terminal formatting, colors, and progress bars
    "pyyaml>=6.0.0",  # YAML parsing for grammar rules (required)
    # Configuration validation (Pydantic v2)
    "pydantic>=2.0.0,<3.0.0",  # Type-safe configuration with rich validation
    # Zawgyi detection and conversion (required for Myanmar language support)
    "myanmartools>=1.2.1",  # Google's statistical Zawgyi detector (~95% accuracy)
    "python-myanmar>=1.0.0,<2.0.0",  # Zawgyi-to-Unicode conversion
]

[project.scripts]
myspellchecker = "myspellchecker.cli:main"

[project.optional-dependencies]
# Dictionary building pipeline (myspellchecker build)
build = [
    "pyarrow>=14.0.1",  # CVE-2023-47248 fixed in 14.0.1
    "duckdb>=1.0.0",  # Fast OLAP database for build pipeline aggregations
    "xxhash>=3.0.0",  # Fast non-cryptographic hashing for dedup
    "tqdm>=4.65.0",  # Progress bars for data pipeline
    "cached-path>=1.2.0",  # Resource downloading and caching
]

# AI features (Semantic Context Checking)
ai = [
    "onnxruntime>=1.17.0",
    "tokenizers>=0.13.0",
]

# Transformer-based POS tagging (Optional)
transformers = [
    "transformers>=4.38.0",
    "torch>=2.6.0",  # CVE-2025-32434 RCE fixed in 2.6.0
]

# Complete AI stack (Semantic + Transformer POS)
ai-full = ["myspellchecker[ai,transformers]"]

# Training dependencies (for custom model training)
# Includes ONNX export dependencies since training pipeline exports to ONNX
train = [
    "myspellchecker[ai,transformers]",
    "datasets>=2.14.0",
    "accelerate>=0.21.0",
    "onnx>=1.14.0",         # ONNX model format (required by onnxruntime.quantization)
    "onnxscript>=0.1.0",    # Required by torch.onnx.export
]

dev = [
    # Testing
    "pytest>=8.0.0",
    "pytest-cov>=4.0.0",
    "pytest-benchmark>=4.0.0",
    "pytest-xdist>=3.0.0",  # Parallel test execution
    "pytest-timeout>=2.2.0",  # Abort hanging tests
    "hypothesis>=6.0.0",    # Property-based testing

    # Type checking and linting
    "mypy>=1.0.0",
    "types-PyYAML>=6.0.12",
    "ruff>=0.3.0",

    # AI Training (Only needed for model development, not usage)
    "myspellchecker[train]",
]

# Note: pyyaml is now a core dependency (required for grammar rules)

[project.urls]
Homepage = "https://github.com/thettwe/myspellchecker"
Documentation = "https://docs.myspellchecker.com/"
Repository = "https://github.com/thettwe/myspellchecker"
"Bug Tracker" = "https://github.com/thettwe/myspellchecker/issues"

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

[tool.setuptools.package-data]
myspellchecker = [
    "py.typed",
    "**/*.pyx",
    "**/*.pxd",
    "data/*.onnx",
    "data/*.json",
    "data/models/*.mmap",
    "rules/*.yaml",
    "schemas/*.json",
]

[tool.pytest.ini_options]
# testpaths intentionally omitted — specify paths explicitly on CLI:
#   pytest tests/              → full suite
#   pytest tests/test_foo.py   → single file
#   pytest tests/ -m unit      → unit tests only
python_files = "test_*.py"
python_classes = "Test*"
python_functions = "test_*"
# Coverage is opt-in: pytest --cov=myspellchecker --cov-report=term-missing --cov-fail-under=75
addopts = [
    "--strict-markers",
    "-m", "not slow",       # Skip slow tests by default; run with: pytest -m slow
]
norecursedirs = ["scripts", ".git", "venv", "*.egg-info", "__pycache__", ".tox", "dist", "build"]
markers = [
    "slow: marks tests as slow",
    "integration: marks tests as integration tests",
    "unit: marks tests as unit tests",
    "e2e: marks tests as end-to-end tests",
    "benchmark: marks tests as benchmarks/performance checks",
]
filterwarnings = [
    "ignore:builtin type SwigPyPacked:DeprecationWarning",
    "ignore:builtin type SwigPyObject:DeprecationWarning",
]

[tool.coverage.run]
source = ["src/myspellchecker"]
omit = [
    "*/tests/*",
    "*/__pycache__/*",
    # Hard-to-test files requiring optional dependencies (torch, transformers)
    "src/myspellchecker/training/trainer.py",
    "src/myspellchecker/algorithms/pos_tagger_transformer.py",
    "src/myspellchecker/text/ner_model.py",
]

[tool.coverage.report]
exclude_lines = [
    "pragma: no cover",
    "raise NotImplementedError",
    "if TYPE_CHECKING:",
    "if __name__ == .__main__.:",
]
omit = [
    "*/tests/*",
    "*/__pycache__/*",
    "src/myspellchecker/training/trainer.py",
    "src/myspellchecker/algorithms/pos_tagger_transformer.py",
    "src/myspellchecker/text/ner_model.py",
]

[tool.ruff]
line-length = 100
target-version = "py310"
extend-exclude = ["scripts", "plan"]

[tool.ruff.lint]
# Enable Pyflakes (F), pycodestyle (E, W), isort (I), and bugbear (B)
select = ["E", "F", "W", "I", "B"]
ignore = []

# Allow unused variables when underscore-prefixed.
dummy-variable-rgx = "^(_+|(_+[a-zA-Z0-9_]*[a-zA-Z0-9]+?))$"

[tool.ruff.lint.per-file-ignores]
# Tests with pytest.importorskip() at module level need imports after the skip guard
"tests/test_frequency_builder.py" = ["E402"]
"tests/test_data_pipeline.py" = ["E402"]
"tests/test_db_build_and_usage.py" = ["E402"]
"tests/test_incremental_build.py" = ["E402"]
"tests/test_pos_backward_compat.py" = ["E402"]
"tests/test_disk_space.py" = ["E402"]

[tool.ruff.format]
quote-style = "double"
indent-style = "space"
skip-magic-trailing-comma = false
line-ending = "auto"

[tool.mypy]
python_version = "3.10"
warn_return_any = true
warn_unused_configs = true
disallow_untyped_defs = false
disallow_incomplete_defs = false
check_untyped_defs = true
ignore_missing_imports = true
exclude = [
    "^tests/",
    "^benchmarks/",
    "^venv",
]

[[tool.mypy.overrides]]
module = [
    "pycrfsuite.*",
    "myanmartools.*",
    "duckdb.*",
    "pyarrow.*",
    "onnxruntime.*",
    "transformers.*",
    "torch.*",
    "cached_path.*",
]
ignore_missing_imports = true