pyproject.toml

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

[project]
name = "fireflyframework-datascience"
version = "26.6.0"
description = "A state-of-the-art Python metaframework for AutoML that fuses GenAI (built on Firefly Agentic / Pydantic AI) with classical ML and Deep Learning — hexagonal, secure-by-default, native to the Firefly Framework."
readme = "README.md"
license = { text = "Apache-2.0" }
authors = [{ name = "Firefly Software Foundation" }]
maintainers = [{ name = "Firefly Software Foundation" }]
requires-python = ">=3.13"
classifiers = [
    "Development Status :: 3 - Alpha",
    "Framework :: Pydantic :: 2",
    "Intended Audience :: Developers",
    "Intended Audience :: Science/Research",
    "License :: OSI Approved :: Apache Software License",
    "Operating System :: OS Independent",
    "Programming Language :: Python :: 3",
    "Programming Language :: Python :: 3.13",
    "Topic :: Scientific/Engineering :: Artificial Intelligence",
    "Topic :: Software Development :: Libraries :: Python Modules",
    "Typing :: Typed",
]
keywords = ["automl", "machine-learning", "deep-learning", "genai", "agents", "data-science", "framework"]
dependencies = [
    # The single framework hard-dependency: the Firefly GenAI substrate. It is not on PyPI, so we pin a
    # direct git URL (PEP 508). This is tool-agnostic (pip and uv both honour it) and — crucially —
    # avoids a dependency-confusion vector: a bare name would let pip fall back to PyPI for an unowned
    # name. Pin @<rev> for a fully reproducible release.
    "fireflyframework-agentic @ git+https://github.com/fireflyframework/fireflyframework-agentic@main",
    "pydantic>=2.13,<3",
    "pydantic-settings>=2.14.2,<3",
    "pyyaml>=6.0.1",
    "click>=8.1.7",
    "rich>=13.7.0",
    "typing-extensions>=4.12.0",
]

[project.optional-dependencies]
# --- classical ML core ---
tabular = ["pandas>=2.2.0", "numpy>=1.26.0", "scikit-learn>=1.5.0", "xgboost>=2.1.0", "lightgbm>=4.5.0", "catboost>=1.2.7", "optuna>=4.0.0"]
tabfm = ["tabpfn>=2.0.0"]
automl = ["autogluon.tabular>=1.2.0"]
# --- deep learning ---
dl = ["torch>=2.4.0", "lightning>=2.4.0", "torchvision>=0.19.0", "accelerate>=1.0.0"]
nlp = ["transformers>=4.45.0", "datasets>=3.0.0", "peft>=0.13.0", "trl>=0.11.0", "sentencepiece>=0.2.0"]
# --- mlops ---
tracking = ["mlflow>=2.17.0"]
tracking-wandb = ["wandb>=0.18.0"]
validation = ["pandera>=0.20.0"]
explain = ["shap>=0.46.0"]
featurestore = ["feast>=0.40.0"]
serving = ["bentoml>=1.3.0"]
serving-llm = ["vllm>=0.6.0"]
lineage = ["openlineage-python>=1.23.0"]
orchestration = ["apache-airflow>=2.10.0"]
# --- data / benchmarks ---
data = ["openml>=0.15.0", "polars>=1.9.0"]
# --- genai accelerators (agentic extras; the base dep is pinned via the direct git URL above) ---
genai = [
    "fireflyframework-agentic[script-execution,embeddings,openai-embeddings,vectorstores-chroma]",
]
# convenience bundles
automl-stack = ["fireflyframework-datascience[tabular,tabfm,automl,tracking,validation,data]"]
full = ["fireflyframework-datascience[tabular,tabfm,automl,dl,nlp,tracking,validation,explain,featurestore,serving,lineage,orchestration,data,genai]"]

[project.scripts]
firefly-ds = "fireflyframework_datascience.cli.main:cli"

# Auto-configuration plugin discovery group (mirrors pyfly's pattern with our own
# namespace). Each DS module self-registers its adapters here when present.
[project.entry-points."firefly_datascience.auto_configuration"]
core = "fireflyframework_datascience.core.auto_configuration:CoreAutoConfiguration"
datasets = "fireflyframework_datascience.datasets.auto_configuration:DatasetsAutoConfiguration"
engineering = "fireflyframework_datascience.engineering.auto_configuration:EngineeringAutoConfiguration"
models = "fireflyframework_datascience.models.auto_configuration:ModelsAutoConfiguration"
evaluation = "fireflyframework_datascience.evaluation.auto_configuration:EvaluationAutoConfiguration"
explainability = "fireflyframework_datascience.explainability.auto_configuration:ExplainabilityAutoConfiguration"
features = "fireflyframework_datascience.features.auto_configuration:FeaturesAutoConfiguration"
search = "fireflyframework_datascience.search.auto_configuration:SearchAutoConfiguration"
validation = "fireflyframework_datascience.validation.auto_configuration:ValidationAutoConfiguration"
tracking = "fireflyframework_datascience.tracking.auto_configuration:TrackingAutoConfiguration"
dl = "fireflyframework_datascience.dl.auto_configuration:DLAutoConfiguration"
nlp = "fireflyframework_datascience.nlp.auto_configuration:NLPAutoConfiguration"
vision = "fireflyframework_datascience.vision.auto_configuration:VisionAutoConfiguration"
serving = "fireflyframework_datascience.serving.auto_configuration:ServingAutoConfiguration"
lineage = "fireflyframework_datascience.lineage.auto_configuration:LineageAutoConfiguration"
automl = "fireflyframework_datascience.automl.auto_configuration:AutoMLAutoConfiguration"

[dependency-groups]
dev = [
    "pytest>=8.3.0",
    "pytest-asyncio>=0.24.0",
    "pytest-cov>=5.0.0",
    "coverage[toml]>=7.6.0",
    "ruff>=0.7.0",
    "pyright>=1.1.380",
]
docs = ["mkdocs-material>=9.5.0"]

[tool.hatch.metadata]
# Allow the PEP 508 direct git URL for fireflyframework-agentic (it is not on PyPI).
allow-direct-references = true

[tool.hatch.build.targets.wheel]
packages = ["src/fireflyframework_datascience"]

[tool.ruff]
target-version = "py313"
line-length = 120
extend-exclude = ["assets"]
force-exclude = true

[tool.ruff.lint]
select = ["E", "F", "W", "I", "N", "UP", "B", "SIM", "TC", "PLC0415"]
ignore = ["E501", "TC001", "TC002", "TC003", "UP040", "UP046", "UP047", "B008", "N803", "N806", "N812"]

[tool.ruff.lint.per-file-ignores]
# The DataScience domain layer lazy-imports optional heavy dependencies (pandas, scikit-learn,
# xgboost, mlflow, …) inside methods so the core stays importable without any extra installed — the
# deferred import IS the design. The foundation (core/container/application) keeps PLC0415 enforced.
"src/fireflyframework_datascience/**/auto_configuration.py" = ["PLC0415"]
"src/fireflyframework_datascience/datasets/**" = ["PLC0415"]
"src/fireflyframework_datascience/models/**" = ["PLC0415"]
"src/fireflyframework_datascience/engineering/**" = ["PLC0415"]
"src/fireflyframework_datascience/evaluation/**" = ["PLC0415"]
"src/fireflyframework_datascience/explainability/**" = ["PLC0415"]
"src/fireflyframework_datascience/features/**" = ["PLC0415"]
"src/fireflyframework_datascience/search/**" = ["PLC0415"]
"src/fireflyframework_datascience/validation/**" = ["PLC0415"]
"src/fireflyframework_datascience/tracking/**" = ["PLC0415"]
"src/fireflyframework_datascience/automl/**" = ["PLC0415"]
"src/fireflyframework_datascience/dl/**" = ["PLC0415"]
"src/fireflyframework_datascience/nlp/**" = ["PLC0415"]
"src/fireflyframework_datascience/vision/**" = ["PLC0415"]
"src/fireflyframework_datascience/serving/**" = ["PLC0415"]
"src/fireflyframework_datascience/lineage/**" = ["PLC0415"]
"src/fireflyframework_datascience/preprocessing.py" = ["PLC0415"]
"tests/**" = ["PLC0415"]
"samples/**" = ["PLC0415"]
"benchmarks/**" = ["PLC0415"]

[tool.ruff.lint.isort]
known-first-party = ["fireflyframework_datascience"]

[tool.pyright]
pythonVersion = "3.13"
typeCheckingMode = "basic"
include = ["src/fireflyframework_datascience"]
exclude = ["tests/**", "samples/**", "benchmarks/**", "assets/**"]
extraPaths = ["src"]

[tool.pytest.ini_options]
asyncio_mode = "auto"
testpaths = ["tests"]
pythonpath = ["src", "."]
addopts = "-m 'not nightly and not integration'"
markers = [
    "nightly: long-running tests (benchmarks, real LLM/GPU/Kaggle); excluded from the PR gate.",
    "integration: real-backend tests (testcontainers / real ML libs); excluded from the fast gate.",
]

[tool.coverage.run]
source = ["src/fireflyframework_datascience"]
branch = true