table-stitcher 0.3.0__tar.gz

table_stitcher-0.3.0/.claude/settings.json

@@ -0,0 +1,76 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(python3 -c \"import docling, table_stitcher; print\\('ok'\\)\")",
+      "Bash(python3 -m pip show docling)",
+      "Bash(python3 -m pip show table_stitcher)",
+      "Bash(python3 -c \"import sys; print\\(sys.executable\\)\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -c \"import table_stitcher; print\\(table_stitcher.__file__\\); import docling; print\\('docling ok'\\)\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 /tmp/inspect_testdocs.py)",
+      "Bash(tee /tmp/inspect_run.log)",
+      "Bash(python3 -c \"import yaml; print\\(yaml.__version__\\)\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/integration/ -v)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/ -v)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 /tmp/gen_expected.py)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/ -v --no-header)",
+      "Bash(curl -L -o images.tar.gz --progress-bar \"https://huggingface.co/datasets/kensho/PubTables-v2/resolve/main/PubTables-v2_Full-Documents_test_images.tar.gz\")",
+      "Bash(curl -sL -o cross-page-pairs.tar.gz \"https://huggingface.co/datasets/kensho/PubTables-v2/resolve/main/PubTables-v2_Full-Documents_test_cross-page-table-pairs.tar.gz\")",
+      "Bash(curl -sL -o xml-annotations.tar.gz \"https://huggingface.co/datasets/kensho/PubTables-v2/resolve/main/PubTables-v2_Full-Documents_test_xml-annotations.tar.gz\")",
+      "Bash(curl -sL -o tables.tar.gz \"https://huggingface.co/datasets/kensho/PubTables-v2/resolve/main/PubTables-v2_Full-Documents_test_tables.tar.gz\")",
+      "Bash(curl -sL -o README.md \"https://huggingface.co/datasets/kensho/PubTables-v2/resolve/main/README.md\")",
+      "Bash(mkdir -p extracted)",
+      "Bash(tar -xzf ../cross-page-pairs.tar.gz)",
+      "Bash(tar -xzf ../xml-annotations.tar.gz)",
+      "Bash(tar -xzf ../tables.tar.gz)",
+      "Bash(awk -F/ '{print $4}')",
+      "Bash(python3 -c \"import json; d=json.load\\(open\\('Full Documents/test/tables/PMC10157558_tables.json'\\)\\); print\\(type\\(d\\)\\); print\\(json.dumps\\(d, indent=2\\)[:2000]\\)\")",
+      "Bash(python3 /tmp/select_multipage.py)",
+      "Bash(tar -tzf /Users/pebbleroad/Documents/table-stitcher/.temp/images.tar.gz)",
+      "Bash(python3 -c \"import img2pdf; print\\('img2pdf', img2pdf.__version__\\)\")",
+      "Bash(tar -xzf ../images.tar.gz)",
+      "Bash(python3 /tmp/bundle_pdfs.py)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 /tmp/inspect_pt2.py)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 /tmp/install_pt2_fixtures.py)",
+      "Bash(tee /tmp/install_output.log)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/)",
+      "Bash(python3 -c ' *)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/test_docling_adapter.py -v)",
+      "Bash(python3 /tmp/select_second_slice.py)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 /tmp/inspect_pt2_v2.py)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 /tmp/install_pt2_v2_fixtures.py)",
+      "Bash(tee /tmp/install_v2.log)",
+      "Bash(python3 -c \"import reportlab; print\\(reportlab.Version\\)\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m tests.integration.fixtures._synth.generate)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/test_merger.py -v)",
+      "Bash(git rm *)",
+      "Bash(python3 -m pip show docling docling-core)",
+      "Bash(python3 -c \"import pypdf; print\\(pypdf.__version__\\)\")",
+      "Bash(python3 -c \"import pikepdf; print\\(pikepdf.__version__\\)\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/integration/ -v -k \"edinet\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/test_merger.py -v -k \"tokenize\")",
+      "Bash(python3 -m tests.integration._tools.regenerate_expected tests/integration/fixtures/multilingual/corporate-history-2pg.edinet.pdf --clear-xfail --description 'Two-page corporate history table from a Japanese EDINET filing. Both pages carry identical 2-column Japanese headers \\(年⽉ / 概要\\). Exercises the repeated-header Jaccard path against pure CJK text — merges because tokenize\\(\\) emits one token per CJK character, giving Jaccard 1.0 on identical headers.')",
+      "Bash(python3 -m tests.integration._tools.regenerate_expected tests/integration/fixtures/multilingual/subsidiaries-4pg.edinet.pdf --clear-xfail --description 'Four-page subsidiaries list from a Japanese EDINET filing. Three fragments \\(widths 6/6/7\\), all with identical 6-column Japanese headers \\(名称 / 住所 / 資本⾦⼜は出資⾦ / 主要な事業の内容 / 議決権の所有割合 / 関係内容\\). The width-7 final fragment is the summary-rows continuation of the same table. All three merge on CJK-aware header Jaccard.')",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest -m integration tests/integration/test_fixtures.py::test_fixture_stitches_as_expected -k \"varicose\")",
+      "Bash(python3 -m tests.integration._tools.regenerate_expected tests/integration/fixtures/orphan-pair/varicose-veins-new-table-header-7pg.pt2.pdf)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/ tests/ -m \"integration\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest -m integration tests/)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest -m integration tests/integration/test_fixtures.py::test_fixture_stitches_as_expected -k \"substance-list or note-overflow\")",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/test_docling_adapter.py::TestInjection::test_injection_failure_restores_partial_mutations -v)",
+      "Bash(scripts/release_gate.sh)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest -m integration tests/integration/test_fixtures.py -k \"substance-list or note-overflow\")",
+      "Bash(python -c \"import sys; print\\(sys.version\\)\")",
+      "Bash(python3 -m pip show ruff)",
+      "Bash(ruff format *)",
+      "Bash(ruff check *)",
+      "Bash(python3 -m pytest -q)",
+      "Bash(python3 -m build --wheel --outdir /tmp/wheel-check)",
+      "Bash(unzip -l /tmp/wheel-check/*.whl)",
+      "Bash(python3 -m pytest tests/test_public_api.py -v)",
+      "Bash(python3 -m pytest tests/:*)",
+      "Bash(python3 -m pytest tests:*)",
+      "Bash(PYTHONPATH=/Users/pebbleroad/Documents/table-stitcher/src python3 -m pytest tests/:*)",
+      "Bash(python3 -m pytest -q:*)",
+      "Bash(python3 -m pytest -m:*)"
+    ]
+  }
+}

table_stitcher-0.3.0/.github/ISSUE_TEMPLATE/bug_report.yml

@@ -0,0 +1,90 @@
+name: Bug report
+description: A merge produced wrong output, threw an unexpected error, or behaved differently than documented.
+labels: ["bug"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Thanks for filing a bug. The fields below are what we need to reproduce
+        the issue. The biggest unknown for a table-stitching bug is usually
+        *what the upstream parser produced* — please attach a minimal fixture
+        if you can.
+
+  - type: input
+    id: version
+    attributes:
+      label: table-stitcher version
+      description: Output of `python -c "import table_stitcher; print(table_stitcher.__version__)"`
+      placeholder: "0.2.0"
+    validations:
+      required: true
+
+  - type: input
+    id: python
+    attributes:
+      label: Python version
+      placeholder: "3.11.7"
+    validations:
+      required: true
+
+  - type: dropdown
+    id: parser
+    attributes:
+      label: Upstream parser
+      description: Which extractor produced the input tables?
+      options:
+        - Docling (built-in adapter)
+        - Custom adapter
+        - Other
+    validations:
+      required: true
+
+  - type: textarea
+    id: what-happened
+    attributes:
+      label: What happened
+      description: What did the merger produce, and what did you expect instead?
+      placeholder: |
+        Pages 4–5 of report.pdf contain one logical table split by a page break.
+        I expected `stitch_tables(doc)` to return one merged table; instead I
+        got two separate tables with the headers duplicated.
+    validations:
+      required: true
+
+  - type: textarea
+    id: repro
+    attributes:
+      label: Minimal reproduction
+      description: |
+        Smallest input + code that demonstrates the bug. If you can attach a
+        PDF fragment or a synthetic `TableMeta` list, that's ideal. See
+        [CONTRIBUTING.md](https://github.com/pebbleroad/table-stitcher/blob/main/CONTRIBUTING.md#adding-a-new-fixture)
+        for the fixture format we use internally.
+      render: python
+      placeholder: |
+        from docling.document_converter import DocumentConverter
+        from table_stitcher import stitch_tables
+
+        doc = DocumentConverter().convert("attached.pdf").document
+        result = stitch_tables(doc)
+        # observed: 2 tables; expected: 1
+    validations:
+      required: true
+
+  - type: textarea
+    id: config
+    attributes:
+      label: MultiPageConfig (if non-default)
+      description: Any custom thresholds you passed via `MultiPageConfig(...)`.
+      placeholder: "MultiPageConfig(max_page_gap=2, headerless_width_tolerance=3)"
+    validations:
+      required: false
+
+  - type: textarea
+    id: logs
+    attributes:
+      label: Logs / traceback
+      description: Output of the merger at INFO level if relevant. Wrap in triple backticks.
+      render: shell
+    validations:
+      required: false

table_stitcher-0.3.0/.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,8 @@
+blank_issues_enabled: false
+contact_links:
+  - name: Security vulnerability
+    url: https://github.com/pebbleroad/table-stitcher/security/policy
+    about: Please report security issues privately — see SECURITY.md, do not file a public issue.
+  - name: Question or discussion
+    url: https://github.com/pebbleroad/table-stitcher/discussions
+    about: For usage questions or open-ended discussion, use Discussions instead of Issues.

table_stitcher-0.3.0/.github/ISSUE_TEMPLATE/feature_request.yml

@@ -0,0 +1,49 @@
+name: Feature request
+description: Propose a new merge signal, config knob, adapter, or API surface.
+labels: ["enhancement"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        The library deliberately keeps merge signals **structural** (column
+        counts, header similarity, layout proximity) and avoids
+        language-specific or vocabulary-based rules. Proposals that fit that
+        model land faster — see [CONTRIBUTING.md](https://github.com/pebbleroad/table-stitcher/blob/main/CONTRIBUTING.md)
+        for context.
+
+  - type: textarea
+    id: problem
+    attributes:
+      label: What problem does this solve?
+      description: A real document or pipeline where the current behavior falls short.
+    validations:
+      required: true
+
+  - type: textarea
+    id: proposal
+    attributes:
+      label: Proposed change
+      description: |
+        Sketch the API or behavior change. New `MultiPageConfig` field?
+        New adapter? New merge phase? Pseudocode is welcome.
+    validations:
+      required: true
+
+  - type: textarea
+    id: alternatives
+    attributes:
+      label: Alternatives considered
+      description: |
+        What you tried already (different config, custom adapter, post-processing) and why it wasn't enough.
+    validations:
+      required: false
+
+  - type: checkboxes
+    id: structural
+    attributes:
+      label: Compatibility with the structural-only constraint
+      options:
+        - label: This proposal does **not** rely on language-specific dictionaries, vocabulary lists, or trained models.
+          required: false
+        - label: I understand the maintainers may decline language-aware proposals to keep the core multilingual.
+          required: false

table_stitcher-0.3.0/.github/dependabot.yml

@@ -0,0 +1,27 @@
+version: 2
+updates:
+  # Python deps in pyproject.toml — bumps `pandas`, `docling`, dev tools.
+  # Weekly cadence avoids PR spam; security advisories are picked up out-of-band.
+  - package-ecosystem: pip
+    directory: "/"
+    schedule:
+      interval: weekly
+    open-pull-requests-limit: 5
+    groups:
+      # Group dev-only deps so a routine ruff/pytest bump doesn't
+      # produce three separate PRs in one week.
+      dev-tooling:
+        patterns:
+          - "ruff"
+          - "pytest*"
+          - "build"
+          - "twine"
+          - "pre-commit"
+
+  # GitHub Actions versions in .github/workflows/*.yml — keeps action
+  # references current (e.g. actions/checkout@v4 → @v5 when stable).
+  - package-ecosystem: github-actions
+    directory: "/"
+    schedule:
+      interval: weekly
+    open-pull-requests-limit: 5

table_stitcher-0.3.0/.github/pull_request_template.md

@@ -0,0 +1,15 @@
+## What changed
+
+<!-- One or two sentences. The "why" matters more than the "what" — the diff shows the what. -->
+
+## Why
+
+<!-- The motivation. A linked issue, a fixture that was failing, a downstream user request. -->
+
+## Checklist
+
+- [ ] `pytest tests/` is green locally (or xfails are intentional, with reasons in the YAML)
+- [ ] `ruff check .` and `ruff format --check .` pass (pre-commit handles this for you)
+- [ ] New behavior has a test (unit or fixture)
+- [ ] User-visible changes are noted in [CHANGELOG.md](https://github.com/pebbleroad/table-stitcher/blob/main/CHANGELOG.md)
+- [ ] No language-specific or vocabulary assumptions added — merge signals stay structural

table_stitcher-0.3.0/.github/workflows/ci.yml

@@ -0,0 +1,119 @@
+name: ci
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  lint:
+    # Style + static-analysis gate. Single Python — ruff produces identical
+    # results across versions, so a matrix here just burns minutes.
+    name: lint
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+          cache: pip
+      - name: install ruff
+        # Pin via pyproject's dev extra so local + CI run the same version.
+        # No need to install the full dev tree (docling, reportlab) just to lint.
+        run: |
+          python -m pip install --upgrade pip
+          pip install "ruff>=0.6"
+      - name: ruff check
+        run: ruff check .
+      - name: ruff format --check
+        run: ruff format --check .
+
+  test:
+    # Behavior gate. Matrix across all supported Python versions.
+    name: test (py${{ matrix.python }})
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python: ["3.9", "3.10", "3.11", "3.12"]
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python }}
+          cache: pip
+      - name: install
+        run: |
+          python -m pip install --upgrade pip
+          pip install -e ".[dev]"
+      - name: unit tests
+        # Default addopts in pyproject excludes the integration marker, so
+        # this run is already unit-only. -v adds per-case outcomes.
+        run: pytest tests/ -v
+
+  integration:
+    # Snapshot lane: tests load committed *.docling.json fixtures and run
+    # table-stitcher against them. No PDF parsing, no OCR, no model downloads
+    # — deterministic across platforms regardless of which OCR engine docling
+    # would auto-select. The full live-parse pipeline is exercised separately
+    # by .github/workflows/upstream-smoke.yml on a nightly schedule.
+    name: integration
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: pip
+      - name: install
+        run: |
+          python -m pip install --upgrade pip
+          pip install -e ".[dev]"
+      - name: integration tests (snapshot lane)
+        # Integration tests are gated behind the `integration` marker in
+        # pyproject; opt in explicitly here. The snapshot lane is the default
+        # — no --live-parse flag.
+        run: pytest -m integration tests/ -v
+
+  build:
+    # Packaging gate. Builds sdist + wheel, validates PyPI metadata, and
+    # smoke-tests the *installed wheel* in a clean venv outside the checkout.
+    # That last step is the one that catches "passes tests, breaks on
+    # `pip install`" — the most common post-release failure.
+    name: build
+    runs-on: ubuntu-latest
+    needs: [lint, test]
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+          cache: pip
+      - name: install build tooling
+        run: |
+          python -m pip install --upgrade pip
+          pip install build twine
+      - name: build sdist + wheel
+        run: python -m build
+      - name: twine check
+        run: twine check dist/*
+      - name: smoke-test installed wheel
+        # Install the built wheel into a fresh venv *outside* the checkout,
+        # then import + call a public symbol. Proves the wheel is functional,
+        # not just buildable.
+        run: |
+          python -m venv /tmp/smoke-venv
+          /tmp/smoke-venv/bin/pip install --upgrade pip
+          /tmp/smoke-venv/bin/pip install dist/*.whl
+          cd /tmp  # leave the checkout so we can't accidentally import from src/
+          /tmp/smoke-venv/bin/python - <<'PY'
+          import pathlib
+          import table_stitcher
+
+          pkg = pathlib.Path(table_stitcher.__file__).resolve()
+          assert "site-packages" in str(pkg), pkg
+          assert table_stitcher.__version__, "missing __version__"
+          assert callable(table_stitcher.stitch_tables)
+          print(f"smoke ok: {table_stitcher.__version__} from {pkg}")
+          PY

table_stitcher-0.3.0/.github/workflows/release.yml

@@ -0,0 +1,52 @@
+name: release
+
+on:
+  push:
+    tags:
+      - "v*"
+
+# Trusted Publishing requires `id-token: write`. No PyPI API token needed —
+# PyPI verifies the OIDC claim against the project's trusted-publisher config.
+# One-time setup on PyPI: Account → Publishing → Add a pending publisher with
+# repo `pebbleroad/table-stitcher`, workflow `release.yml`, environment `pypi`.
+permissions:
+  contents: read
+  id-token: write
+
+jobs:
+  publish:
+    name: publish to PyPI
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/p/table-stitcher
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+          cache: pip
+      - name: install build tooling
+        run: |
+          python -m pip install --upgrade pip
+          pip install -e ".[dev]"
+      - name: tag matches package version
+        # Refuse to publish if the git tag and pyproject version disagree —
+        # mismatch here is the classic source of "v0.2.0 on PyPI is actually 0.2.1".
+        run: |
+          tag="${GITHUB_REF##*/}"
+          pkg_version=$(python -c "import tomllib, pathlib; print(tomllib.loads(pathlib.Path('pyproject.toml').read_text())['project']['version'])")
+          if [ "$tag" != "v${pkg_version}" ]; then
+            echo "tag $tag does not match pyproject version v$pkg_version" >&2
+            exit 1
+          fi
+      - name: release gate (build + wheel smoke + tests)
+        # Re-runs the same gate that contributors use locally. Keeps "what
+        # gets published" and "what was validated" identical.
+        env:
+          RELEASE_GATE_ONLINE: "1"
+        run: ./scripts/release_gate.sh
+      - name: publish
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          packages-dir: dist/

table_stitcher-0.3.0/.github/workflows/upstream-smoke.yml

@@ -0,0 +1,46 @@
+name: upstream-smoke
+
+# Live-parse lane: re-runs docling against the fixture PDFs to catch upstream
+# regressions (docling, OCR engines, model updates) without blocking PRs.
+#
+# Comparisons run in lenient mode — only structural fields (members, pages,
+# shape) are checked, since cell text varies by OCR engine. The job is
+# allowed to fail; treat persistent red here as a signal to investigate, not
+# a merge blocker.
+
+on:
+  schedule:
+    # Daily at 06:17 UTC. Off-hour to avoid contention with weekday CI bursts.
+    - cron: "17 6 * * *"
+  workflow_dispatch:
+
+jobs:
+  live-parse:
+    name: live-parse (${{ matrix.os }})
+    # macos-latest matches the OCR engine (ocrmac / Apple Vision) used to
+    # generate the committed snapshots, so structural drift here points at a
+    # genuine docling change rather than at OCR-engine divergence.
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [macos-latest]
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: pip
+      - name: cache docling models
+        uses: actions/cache@v4
+        with:
+          path: ~/.cache/huggingface
+          key: docling-models-${{ runner.os }}-v1
+      - name: install
+        run: |
+          python -m pip install --upgrade pip
+          pip install -e ".[dev]"
+      - name: live-parse integration tests (lenient)
+        # Allowed to fail — we never want this to block a PR.
+        continue-on-error: true
+        run: pytest -m integration --live-parse tests/ -v

table_stitcher-0.3.0/.gitignore

@@ -0,0 +1,45 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.egg-info/
+*.egg
+dist/
+build/
+.eggs/
+
+# Virtual environments
+.venv/
+venv/
+env/
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# OS
+.DS_Store
+Thumbs.db
+
+# Testing
+.pytest_cache/
+.coverage
+htmlcov/
+
+# Build
+*.whl
+*.tar.gz
+
+# Test output
+*.enriched.*
+*.stitched.*
+
+# Dataset cache (PubTables-v2 etc.)
+.temp/
+
+# Claude Code per-user overrides (settings.json itself is shared)
+.claude/settings.local.json
+.claude/scheduled_tasks.lock
+

table_stitcher-0.3.0/.pre-commit-config.yaml

@@ -0,0 +1,31 @@
+# Mirror of the CI lint job. Contributors run `pre-commit install` once,
+# then every commit auto-lints + auto-formats. CI catches anything they
+# bypassed (--no-verify) or skipped.
+#
+# We use `language: system` so pre-commit invokes the ruff that
+# `pip install -e ".[dev]"` already installed. Pinning a separate ruff
+# version in this file would drift from the one CI installs and silently
+# disagree on rules and formatter output. Single source of truth: pyproject.
+repos:
+  - repo: local
+    hooks:
+      - id: ruff-check
+        name: ruff check
+        entry: ruff check --fix
+        language: system
+        types: [python]
+      - id: ruff-format
+        name: ruff format
+        entry: ruff format
+        language: system
+        types: [python]
+
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v5.0.0
+    hooks:
+      - id: trailing-whitespace
+      - id: end-of-file-fixer
+      - id: check-yaml
+      - id: check-toml
+      - id: check-added-large-files
+        args: [--maxkb=2000]  # PDF fixtures are legitimately large