run-llama--liteparse/.github/workflows/ci.yml

name: CI

on:
  push:
    branches: [main]
  pull_request:
    branches: [main]

env:
  CARGO_TERM_COLOR: always
  SKIP_INTEGRATION_TESTS: yes

jobs:
  fmt:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - run: rustup toolchain install stable --profile minimal --component rustfmt
      - run: cargo fmt --all -- --check

  clippy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Install system dependencies
        run: |
          sudo apt-get update
          sudo apt-get install -y libtesseract-dev libleptonica-dev
      - uses: actions/cache@v4
        with:
          path: |
            ~/.cargo/registry
            ~/.cargo/git
            ~/.cache/pdfium-rs
            target
          key: ${{ runner.os }}-cargo-clippy-${{ hashFiles('**/Cargo.lock') }}
      - run: rustup toolchain install stable --profile minimal --component clippy
      - run: cargo clippy --all-targets

  build-and-test:
    strategy:
      matrix:
        os: [ubuntu-latest, macos-latest, macos-26-intel, windows-latest, windows-11-arm]
    runs-on: ${{ matrix.os }}
    steps:
      - uses: actions/checkout@v4

      - name: Install system dependencies (Linux)
        if: runner.os == 'Linux'
        run: |
          sudo apt-get update
          sudo apt-get install -y libtesseract-dev libleptonica-dev

      - name: Install system dependencies (macOS)
        if: runner.os == 'macOS'
        run: brew install tesseract leptonica

      - uses: actions/cache@v4
        with:
          path: |
            ~/.cargo/registry
            ~/.cargo/git
            ~/.cache/pdfium-rs
            ~/Library/Caches/pdfium-rs
            ${{ runner.os == 'Windows' && format('{0}\pdfium-rs', env.LOCALAPPDATA) || '' }}
            target
          key: ${{ matrix.os }}-cargo-build-${{ hashFiles('**/Cargo.lock') }}

      - name: Force Ninja generator (Windows)
        if: runner.os == 'Windows'
        run: echo "CMAKE_GENERATOR=Ninja" >> $GITHUB_ENV
        shell: bash

      - run: cargo build --workspace --all-targets --exclude liteparse-python --exclude liteparse-napi
      - run: cargo test --workspace --exclude liteparse-python --exclude liteparse-napi

      - name: CLI smoke test
        if: runner.os != 'Windows'
        run: |
          ./target/debug/lit --version
          ./target/debug/lit parse demo/docs/apple-10k-2024.pdf --no-ocr --max-pages 1 -q | head -5

      - name: OCR smoke test (runtime tessdata download)
        shell: bash
        run: |
          # Wipe any tessdata that tesseract-rs/build.rs downloaded during
          # `cargo build`, so the parse exercises the runtime first-use
          # download path (the code path that npm/pip prebuilt users hit).
          rm -rf "$HOME/Library/Application Support/tesseract-rs" \
                 "$HOME/.tesseract-rs" \
                 "${APPDATA:-}/tesseract-rs" 2>/dev/null || true

          LIT=./target/debug/lit
          [ -f "$LIT.exe" ] && LIT="$LIT.exe"
          "$LIT" parse demo/docs/long_tiny_text.pdf --max-pages 1 -q -o ocr-out.txt

          # If OCR ran, page 1 has substantial text beyond just the
          # "--- Page 1 ---" header (~16 bytes). A tessdata-load failure
          # surfaces as systemic OCR failure and an empty/tiny output.
          size=$(wc -c < ocr-out.txt)
          echo "OCR output size: $size bytes"
          test "$size" -gt 200

  # Test that docker builds
  build-docker-image:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Build the base Docker image
        run: docker build .
      - name: Build the full Docker image
        run: docker build . -f full.Dockerfile