Files
wehub-resource-sync c56bef871b
Sync docs with Docusaurus / sync (push) Waiting to run
Tests / Check if changed (push) Waiting to run
Tests / format (push) Blocked by required conditions
Tests / check-imports (push) Blocked by required conditions
Tests / Unit / macos-latest (push) Blocked by required conditions
Tests / Unit / ubuntu-latest (push) Blocked by required conditions
Tests / Unit / windows-latest (push) Blocked by required conditions
Tests / mypy (push) Blocked by required conditions
Tests / Integration / ubuntu-latest (push) Blocked by required conditions
Tests / Integration / macos-latest (push) Blocked by required conditions
Tests / Integration / windows-latest (push) Blocked by required conditions
Tests / notify-slack-on-failure (push) Blocked by required conditions
Tests / Mark tests as completed (push) Blocked by required conditions
Update Platform Components Table / update (push) Waiting to run
CodeQL / Analyze (python) (push) Waiting to run
Docker image release / Build base image (push) Waiting to run
chore: import upstream snapshot with attribution
2026-07-13 13:22:28 +08:00

352 lines
12 KiB
TOML

[build-system]
requires = ["hatchling>=1.8.0"]
build-backend = "hatchling.build"
[project]
name = "haystack-ai"
dynamic = ["version"]
description = "LLM framework to build customizable, production-ready LLM applications. Connect components (models, vector DBs, file converters) to pipelines or agents that can interact with your data."
readme = "README.md"
license = "Apache-2.0"
requires-python = ">=3.10"
authors = [{ name = "deepset.ai", email = "malte.pietsch@deepset.ai" }]
keywords = [
"BERT",
"QA",
"Question-Answering",
"Reader",
"Retriever",
"albert",
"language-model",
"mrc",
"roberta",
"search",
"semantic-search",
"squad",
"transfer-learning",
"transformer",
]
classifiers = [
"Development Status :: 5 - Production/Stable",
"Intended Audience :: Science/Research",
"License :: Freely Distributable",
"License :: OSI Approved :: Apache Software License",
"Operating System :: OS Independent",
"Programming Language :: Python",
"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",
"Topic :: Scientific/Engineering :: Artificial Intelligence",
]
dependencies = [
"tqdm",
"tenacity!=8.4.0",
"lazy-imports",
"openai>=1.99.2",
"pydantic",
"Jinja2",
"MarkupSafe", # already required by Jinja2 but used directly in templatize_part
"posthog!=3.12.0", # telemetry # 3.12.0 was problematic https://github.com/PostHog/posthog-python/issues/187
"pyyaml",
"more-itertools", # DocumentSplitter
"networkx", # Pipeline graphs
"typing_extensions>=4.7", # Extended typing features (NotRequired, etc.)
"httpx",
"numpy",
"python-dateutil",
"jsonschema", # JsonSchemaValidator, Tool
"docstring-parser", # ComponentTool
"filetype", # MIME type guessing for ImageContent
]
[tool.hatch.envs.default]
installer = "uv"
dependencies = [
"pre-commit",
"ruff>=0.15.0", # RUF104 (unmatched-suppression-comment) added in 0.15.0
"toml",
"reno",
# dulwich is a reno dependency, they pin it at >=0.15.0 so pip takes ton of time to resolve the dependency tree.
# We pin it here to avoid taking too much time.
# https://opendev.org/openstack/reno/src/branch/master/requirements.txt#L7
"dulwich>=0.21.0,<1.0.0",
"haystack-pydoc-tools",
]
[tool.hatch.envs.default.scripts]
release-note = "reno new {args}"
fmt = "ruff check --fix {args}; ruff format {args}"
fmt-check = "ruff check {args} && ruff format --check {args}"
docs = "haystack-pydoc pydoc tmp_api_reference"
[tool.hatch.envs.test]
# we override dependencies from the default environment
dependencies = [
"numpy>=2", # Haystack is compatible both with numpy 1.x and 2.x, but we test with 2.x
"numba>=0.54.0", # This pin helps uv resolve the dependency tree. See https://github.com/astral-sh/uv/issues/7881
"pandas", # AzureOCRDocumentConverter, CSVDocumentCleaner, CSVDocumentSplitter,
# EvaluationRunResult, XLSXToDocument, and pipeline tests
"transformers[torch, sentencepiece]>=4.57", # SASEvaluator and TopPSampler
"sentence-transformers>=5.0.0", # SASEvaluator
"arrow>=1.3.0", # Jinja2TimeExtension
"pillow", # ImageContent
"pypdfium2", # PDFToImageContent
# Converters
"pypdf", # PyPDFToDocument
"pdfminer.six", # PDFMinerToDocument
"markdown-it-py", # MarkdownToDocument
"mdit_plain", # MarkdownToDocument
"trafilatura", # HTMLToDocument
"python-pptx", # PPTXToDocument
"python-docx", # DocxToDocument
"jq", # JSONConverter
"openpyxl", # XLSXToDocument
"tabulate", # XLSXToDocument
"python-oxmsg", # MSGToDocument
"nltk>=3.9.1", # NLTKDocumentSplitter, RecursiveDocumentSplitter
"tiktoken", # RecursiveDocumentSplitter
# Human in the Loop
"rich", # Console rendering for HITL
# Structured logging
"structlog",
# needed in link content fetcher tests
"httpx[http2]",
# Azure Utils
"azure-identity",
# Test
"pytest",
"pytest-bdd",
"pytest-cov",
"pytest-asyncio",
"pytest-rerunfailures",
"coverage",
"mypy",
"pip", # mypy needs pip to install missing stub packages
"ipython",
"colorama==0.4.6", # Pipeline checkpoints test
]
[tool.hatch.envs.test.scripts]
unit = 'pytest --cov="haystack" -m "not integration" {args:test}'
integration = 'pytest --maxfail=5 -m "integration" {args:test}'
integration-only-fast = 'pytest --maxfail=5 -m "integration and not slow" {args:test}'
integration-only-slow = 'pytest --maxfail=5 -m "integration and slow" {args:test}'
all = 'pytest {args:test}'
e2e = 'pytest {args:e2e}'
# TODO We want to eventually type the whole test folder
types = "mypy --install-types --non-interactive --cache-dir=.mypy_cache/ {args:haystack test/core/ test/marshal/ test/testing/ test/tracing/ test/tools/ test/human_in_the_loop test/evaluation test/document_stores test/dataclasses test/utils/ test/components/embedders/}"
[project.urls]
"CI: GitHub" = "https://github.com/deepset-ai/haystack/actions"
"Docs: RTD" = "https://haystack.deepset.ai/overview/intro"
"GitHub: issues" = "https://github.com/deepset-ai/haystack/issues"
"GitHub: repo" = "https://github.com/deepset-ai/haystack"
Homepage = "https://github.com/deepset-ai/haystack"
[tool.hatch.version]
path = "VERSION.txt"
pattern = "(?P<version>.+)"
[tool.hatch.metadata]
allow-direct-references = true
[tool.hatch.build.targets.sdist]
include = ["/haystack", "/VERSION.txt"]
[tool.hatch.build.targets.wheel]
packages = ["haystack"]
[tool.codespell]
ignore-words-list = "ans,astroid,nd,ned,nin,ue,rouge,ist, Claus,SME"
quiet-level = 3
skip = "./test,./e2e,*.svg,*/abbreviations/*"
[tool.pytest.ini_options]
minversion = "6.0"
addopts = "--strict-markers"
markers = [
"unit: unit tests",
"integration: integration tests",
# integration tests that are slow (e.g. model inference on CPU), unstable (e.g. call unstable external services)
# or require special setup (e.g. installing system dependencies, running Docker containers)
"slow: slow/unstable integration tests",
]
log_cli = true
asyncio_mode = "auto"
asyncio_default_fixture_loop_scope = "class"
[tool.mypy]
python_version = "3.10"
check_untyped_defs = true
disallow_incomplete_defs = true
ignore_missing_imports = true
warn_unused_ignores = true
warn_redundant_casts = true
warn_return_any = false
[[tool.mypy.overrides]]
# TODO: Fix testing typings
module = ["haystack.testing.*"]
disallow_incomplete_defs = false
[tool.ruff]
exclude = ["e2e/**", "proposals"]
line-length = 120
[tool.ruff.format]
skip-magic-trailing-comma = true
[tool.ruff.lint.isort]
split-on-trailing-comma = false
[tool.ruff.lint.mccabe]
max-complexity = 20 # Default is 10 for Ruff
[tool.ruff.lint.pylint]
allow-magic-value-types = ["float", "int", "str"]
max-args = 13 # Default is 5
max-branches = 21 # Default is 12
max-public-methods = 20 # Default is 20
max-returns = 7 # Default is 6
max-statements = 56 # Default is 50
[tool.ruff.lint]
select = [
"A001", # builtin-variable-shadowing
"A002", # builtin-argument-shadowing
"A003", # builtin-attribute-shadowing
"ANN", # flake8-annotations
"ARG", # flake8-unused-arguments
"ASYNC", # flake8-async
"B", # flake8-bugbear
"BLE", # flake8-blind-except
"C4", # flake8-comprehensions
"C90", # mccabe complexity
"COM818", # trailing-comma-on-bare-tuple
"D102", # Missing docstring in public method
"D103", # Missing docstring in public function
"D205", # 1 blank line required between summary line and description
"D209", # Closing triple quotes go to new line
"D213", # summary lines must be positioned on the second physical line of the docstring
"D417", # undocumented-parameter
"D419", # undocumented-returns
"E", # pycodestyle errors
"EXE", # flake8-executable
"F", # pyflakes
"G001", # logging-format-interpolation
"G002", # logging-percent-format
"G004", # logging-f-string
"I", # isort
"ISC001", # implicit-string-concatenation
"INT", # flake8-gettext
"PERF", # perflint
"PIE", # flake8-pie
"PL", # pylint
"RET", # flake8-return
"RUF104", # unmatched-suppression-comment
"S102", # flake8-bandit (exec-builtin)
"S307", # flake8-bandit (eval)
"SIM", # flake8-simplify
"SLOT", # flake8-slots
"T10", # flake8-debugger
"TID252", # relative-imports
"TRY", # tryceratops
"UP", # pyupgrade
"W", # pycodestyle warnings
"YTT", # flake8-2020
]
ignore = [
"ANN401", # Allow `Any` - used legitimately for dynamic types and SDK boundaries
"B008", # function-call-as-argument-default
"B904", # raise-without-from-inside-except
"BLE001", # blind-except
"E721", # type-comparison
"E722", # bare-except
"PERF203", # try-except-in-loop
"PERF401", # manual-list-comprehension
"PIE790", # unnecessary-pass
# we re-export symbols in __init__ for correct type checking
# https://typing.python.org/en/latest/spec/distributing.html#import-conventions
"PLC0414", # useless-import-alias
"PLC0415", # import-outside-top-level
"PLR1714", # repeated-equality-comparison
"PLW0603", # global-statement
"PLW1514", # unspecified-encoding
"PLW2901", # redefined-loop-name
"RET505", # superfluous-else-return
"SIM108", # if-else-block-instead-of-ternary
"SIM109", # multiple-comparisons-with-in
"SIM118", # in-dict-keys
"TRY002", # raise-vanilla-class
"TRY003", # raise-vanilla-args
"TRY201", # verbose-raise
"TRY300", # try-consider-else
"UP008", # super-with-arguments
"UP032", # use-f-string
"UP037", # quoted-annotation
]
[tool.ruff.lint.per-file-ignores]
"test/**" = [
"ANN", # flake8-annotations
"ARG", # flake8-unused-arguments
"B018", # useless-expression
"D", # pydocstyle (docstring checks)
"E711", # none-comparison
"E731", # lambda-assignment
"PLC0206", # dict-index-missing-items
"SIM105", # suppressible-exception
"SIM117", # multiple-with-statements
]
# in the following files, we still allow old type hints to handle and test them
# UP006: generics like typing.List, typing.Dict
# UP007: Union[X, Y] syntax
# UP035: Deprecated import
# UP045: Optional[X] syntax
"haystack/testing/**" = ["ANN"]
"haystack/components/agents/state/state_utils.py" = ["UP007"]
"haystack/core/super_component/utils.py" = ["UP007"]
"haystack/core/type_utils.py" = ["UP007"]
"haystack/tools/parameters_schema_utils.py" = ["UP007"]
"test/components/agents/test_state_class.py" = ["UP006", "UP007", "UP035", "UP045"]
"test/components/converters/test_output_adapter.py" = ["UP006", "UP035"]
"test/components/joiners/test_list_joiner.py" = ["UP006", "UP035"]
"test/core/pipeline/test_type_syntax_compatibility.py" = ["UP007", "UP045"]
"test/core/super_component/test_utils.py" = ["UP007", "UP045"]
"test/core/test_type_utils.py" = ["UP006", "UP007", "UP035", "UP045"]
"test/tools/test_parameters_schema_utils.py" = ["UP007"]
"test/utils/test_type_serialization.py" = ["UP006", "UP007", "UP035", "UP045"]
[tool.uv]
# Exclude package versions published within the last 24 hours to protect against supply chain
# attacks via compromised dependencies. uv resolves this relative to the current clock at
# install/lock time, so no manual date updates are needed.
# First-party packages are exempted so freshly published releases are always resolvable.
exclude-newer = "24 hours"
exclude-newer-package = { haystack-pydoc-tools = "0 days" }
[tool.coverage.run]
omit = ["haystack/testing/*"]
relative_files = true