name: Parallel Tests

on:
  workflow_dispatch:
  pull_request:
    branches: [dev, main, master]
  push:
    branches: [dev]

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

env:
  PYTHONIOENCODING: utf-8
  PYTHONLEGACYWINDOWSSTDIO: utf-8
  USE_COLOR: False
  UV_NO_SOURCES: "1"

jobs:
  discover-tests:
    name: Discover test files
    runs-on: ubuntu-24.04
    outputs:
      test-files: ${{ steps.set-matrix.outputs.test-files }}
      plugin-tests: ${{ steps.set-plugin-matrix.outputs.plugin-tests }}

    steps:
      - uses: actions/checkout@v4
        with:
          submodules: true
          fetch-depth: 1

      - name: Discover test files
        id: set-matrix
        run: |
          # Find all main test files
          all_tests=$(find archivebox/tests -maxdepth 1 -name "test_*.py" -type f | sort)

          # Create JSON array with test file info
          json_array="["
          first=true
          for test_file in $all_tests; do
            if [ "$first" = true ]; then
              first=false
            else
              json_array+=","
            fi

            # Extract a display name for the test
            name="main/$(basename $test_file .py | sed 's/^test_//')"
            needs_chromium=false
            if grep -Eiq '(chrom|archivewebpage|PLUGINS=.*title|--plugins=.*title|SAVE_TITLE.*[Tt]rue)' "$test_file"; then
              needs_chromium=true
            fi
            needs_sonic=false
            if grep -Eiq "shutil\\.which\\([\"']sonic|SEARCH_BACKEND_ENGINE=.*sonic|worker_sonic" "$test_file"; then
              needs_sonic=true
            fi

            json_array+="{\"path\":\"$test_file\",\"name\":\"$name\",\"needs_chromium\":$needs_chromium,\"needs_sonic\":$needs_sonic}"
          done
          json_array+="]"

          echo "test-files=$json_array" >> $GITHUB_OUTPUT
          echo "Found $(echo $all_tests | wc -w) test files"
          echo "$json_array" | jq '.'

      - name: Clone abx-plugins
        run: git clone --depth=1 https://github.com/ArchiveBox/abx-plugins.git abx-plugins

      - name: Discover plugin tests
        id: set-plugin-matrix
        run: |
          plugin_tests=$(find abx-plugins/abx_plugins/plugins -maxdepth 2 -type d -name tests | sed 's#abx-plugins/abx_plugins/plugins/##; s#/tests##' | sort)

          json_array="["
          first=true
          for plugin_name in $plugin_tests; do
            if [ "$first" = true ]; then
              first=false
            else
              json_array+=","
            fi

            needs_chromium=false
            if grep -Riq "chrom" "abx-plugins/abx_plugins/plugins/$plugin_name"; then
              needs_chromium=true
            fi
            needs_sonic=false
            if [ "$plugin_name" = "search_backend_sonic" ]; then
              needs_sonic=true
            fi

            json_array+="{\"plugin\":\"$plugin_name\",\"name\":\"plugin/$plugin_name\",\"needs_chromium\":$needs_chromium,\"needs_sonic\":$needs_sonic}"
          done
          json_array+="]"

          echo "plugin-tests=$json_array" >> $GITHUB_OUTPUT
          echo "Found $(echo $plugin_tests | wc -w) plugin test suites"
          echo "$json_array" | jq '.'

  run-tests:
    name: ${{ matrix.test.name }}
    runs-on: ubuntu-24.04
    needs: discover-tests
    env:
      PYTHONPATH: ${{ github.workspace }}/abxpkg:${{ github.workspace }}/abx-plugins:${{ github.workspace }}/abx-dl
      CHROME_HEADLESS: "true"
      CHROME_SANDBOX: "false"
      PERSONAS_DIR: /tmp/abx-personas
      CHROME_USER_DATA_DIR: /tmp/abx-personas/Default/chrome_profile

    strategy:
      fail-fast: false
      matrix:
        test: ${{ fromJson(needs.discover-tests.outputs.test-files) }}
        python: ["3.13"]

    steps:
      - uses: actions/checkout@v4
        with:
          submodules: true
          fetch-depth: 1

      - name: Clone abxpkg
        run: git clone --depth=1 https://github.com/ArchiveBox/abxpkg.git abxpkg

      - name: Clone abx-plugins
        run: git clone --depth=1 https://github.com/ArchiveBox/abx-plugins.git abx-plugins

      - name: Clone abx-dl
        run: git clone --depth=1 https://github.com/ArchiveBox/abx-dl.git abx-dl

      - name: Set up Python ${{ matrix.python }}
        uses: actions/setup-python@v5
        with:
          python-version: ${{ matrix.python }}
          architecture: x64

      - name: Install uv
        uses: astral-sh/setup-uv@v6
        with:
          version: "0.10.6"

      - name: Set up Node JS
        uses: actions/setup-node@v4
        with:
          node-version: 22

      - name: Cache uv
        uses: actions/cache@v4
        with:
          path: ~/.cache/uv
          key: ${{ runner.os }}-${{ matrix.python }}-uv-${{ hashFiles('pyproject.toml') }}
          restore-keys: |
            ${{ runner.os }}-${{ matrix.python }}-uv-

      - uses: awalsh128/cache-apt-pkgs-action@latest
        with:
          packages: git ripgrep build-essential python3-dev python3-setuptools libssl-dev libldap2-dev libsasl2-dev zlib1g-dev libatomic1 python3-minimal gnupg2 curl wget python3-ldap python3-msgpack python3-mutagen python3-regex python3-pycryptodome procps
          version: 1.1

      - name: Install dependencies with uv
        run: |
          uv venv
          uv pip install --group dev -e ./abxpkg -e ./abx-plugins -e ./abx-dl -e ".[all]"

      - name: Install Chrome through ArchiveBox
        if: ${{ matrix.test.needs_chromium }}
        env:
          MIN_CHROMIUM_MAJOR: "149"
          LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_INSTALL_TIMEOUT: "900"
          ABX_CI_DATA_DIR: ${{ runner.temp }}/archivebox-ci-data
        run: |
          set -euo pipefail
          mkdir -p "$LIB_DIR" "$ABX_CI_DATA_DIR"
          echo "LIB_DIR=$LIB_DIR" >> "$GITHUB_ENV"
          echo "ABXPKG_LIB_DIR=$ABXPKG_LIB_DIR" >> "$GITHUB_ENV"

          uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox install chrome

          candidate="$(uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox shell -c 'from archivebox.machine.models import Binary; binary = Binary.objects.filter(name="chromium", status="installed").order_by("-modified_at").first(); print(binary.abspath if binary else "")' | tail -n 1)"
          if [ ! -x "$candidate" ]; then
            echo "ArchiveBox did not install an executable Chromium binary: ${candidate:-not found}" >&2
            exit 1
          fi
          version="$("$candidate" --version || true)"
          major="$(printf '%s\n' "$version" | sed -E 's/.* ([0-9]+)\..*/\1/' | head -1)"
          case "$major" in
            ''|*[!0-9]*) major=0 ;;
          esac
          if [ "$major" -lt "$MIN_CHROMIUM_MAJOR" ]; then
            echo "ArchiveBox installed Chromium is too old: ${version:-unknown}" >&2
            exit 1
          fi
          echo "Using ArchiveBox-installed Chromium: $candidate ($version)"
          echo "CHROME_BINARY=$candidate" >> "$GITHUB_ENV"

      - name: Install Sonic through ArchiveBox
        if: ${{ matrix.test.needs_sonic }}
        env:
          LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_INSTALL_TIMEOUT: "900"
          ABX_CI_DATA_DIR: ${{ runner.temp }}/archivebox-ci-data
        run: |
          set -euo pipefail
          mkdir -p "$LIB_DIR" "$ABX_CI_DATA_DIR"
          echo "LIB_DIR=$LIB_DIR" >> "$GITHUB_ENV"
          echo "ABXPKG_LIB_DIR=$ABXPKG_LIB_DIR" >> "$GITHUB_ENV"

          uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox install search_backend_sonic

          candidate="$(uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox shell -c 'from archivebox.machine.models import Binary; binary = Binary.objects.filter(name="sonic", status="installed").order_by("-modified_at").first(); print(binary.abspath if binary else "")' | tail -n 1)"
          if [ ! -x "$candidate" ]; then
            echo "ArchiveBox did not install an executable Sonic binary: ${candidate:-not found}" >&2
            exit 1
          fi
          version="$("$candidate" --version || true)"
          echo "Using ArchiveBox-installed Sonic: $candidate (${version:-version unknown})"
          echo "SONIC_BINARY=$candidate" >> "$GITHUB_ENV"
          echo "$(dirname "$candidate")" >> "$GITHUB_PATH"

      - name: Run test - ${{ matrix.test.name }}
        run: |
          mkdir -p tests/out
          set +e
          uv run --no-sync --no-sources pytest -xvs "${{ matrix.test.path }}" --basetemp=tests/out --ignore=archivebox/pkgs
          status=$?
          set -e
          if [ "$status" -eq 5 ]; then
            echo "No tests collected from ${{ matrix.test.path }}; treating as an empty test module."
            exit 0
          fi
          exit "$status"

  plugin-tests:
    name: ${{ matrix.plugin.name }}
    runs-on: ubuntu-24.04
    needs: discover-tests
    env:
      PYTHONPATH: ${{ github.workspace }}/abxpkg:${{ github.workspace }}/abx-plugins:${{ github.workspace }}/abx-dl
      CHROME_HEADLESS: "true"
      CHROME_SANDBOX: "false"
      PERSONAS_DIR: /tmp/abx-personas
      CHROME_USER_DATA_DIR: /tmp/abx-personas/Default/chrome_profile

    strategy:
      fail-fast: false
      matrix:
        plugin: ${{ fromJson(needs.discover-tests.outputs.plugin-tests) }}

    steps:
      - uses: actions/checkout@v4
        with:
          submodules: true
          fetch-depth: 1

      - name: Clone abxpkg
        run: git clone --depth=1 https://github.com/ArchiveBox/abxpkg.git abxpkg

      - name: Clone abx-plugins
        run: git clone --depth=1 https://github.com/ArchiveBox/abx-plugins.git abx-plugins

      - name: Clone abx-dl
        run: git clone --depth=1 https://github.com/ArchiveBox/abx-dl.git abx-dl

      - name: Set up Python 3.13
        uses: actions/setup-python@v5
        with:
          python-version: "3.13"
          architecture: x64

      - name: Install uv
        uses: astral-sh/setup-uv@v6
        with:
          version: "0.10.6"

      - name: Set up Node JS
        uses: actions/setup-node@v4
        with:
          node-version: 22

      - name: Cache uv
        uses: actions/cache@v4
        with:
          path: ~/.cache/uv
          key: ${{ runner.os }}-3.13-uv-${{ hashFiles('pyproject.toml') }}
          restore-keys: |
            ${{ runner.os }}-3.13-uv-

      - uses: awalsh128/cache-apt-pkgs-action@latest
        with:
          packages: git ripgrep build-essential python3-dev python3-setuptools libssl-dev libldap2-dev libsasl2-dev zlib1g-dev libatomic1 python3-minimal gnupg2 curl wget python3-ldap python3-msgpack python3-mutagen python3-regex python3-pycryptodome procps
          version: 1.1

      - name: Install dependencies with uv
        run: |
          uv venv
          uv pip install --group dev -e ./abxpkg -e ./abx-plugins -e ./abx-dl -e ".[all]"

      - name: Install Chrome through ArchiveBox
        if: ${{ matrix.plugin.needs_chromium }}
        env:
          MIN_CHROMIUM_MAJOR: "149"
          LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_INSTALL_TIMEOUT: "900"
          ABX_CI_DATA_DIR: ${{ runner.temp }}/archivebox-ci-data
        run: |
          set -euo pipefail
          mkdir -p "$LIB_DIR" "$ABX_CI_DATA_DIR"
          echo "LIB_DIR=$LIB_DIR" >> "$GITHUB_ENV"
          echo "ABXPKG_LIB_DIR=$ABXPKG_LIB_DIR" >> "$GITHUB_ENV"

          uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox install chrome

          candidate="$(uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox shell -c 'from archivebox.machine.models import Binary; binary = Binary.objects.filter(name="chromium", status="installed").order_by("-modified_at").first(); print(binary.abspath if binary else "")' | tail -n 1)"
          if [ ! -x "$candidate" ]; then
            echo "ArchiveBox did not install an executable Chromium binary: ${candidate:-not found}" >&2
            exit 1
          fi
          version="$("$candidate" --version || true)"
          major="$(printf '%s\n' "$version" | sed -E 's/.* ([0-9]+)\..*/\1/' | head -1)"
          case "$major" in
            ''|*[!0-9]*) major=0 ;;
          esac
          if [ "$major" -lt "$MIN_CHROMIUM_MAJOR" ]; then
            echo "ArchiveBox installed Chromium is too old: ${version:-unknown}" >&2
            exit 1
          fi
          echo "Using ArchiveBox-installed Chromium: $candidate ($version)"
          echo "CHROME_BINARY=$candidate" >> "$GITHUB_ENV"

      - name: Install Sonic through ArchiveBox
        if: ${{ matrix.plugin.needs_sonic }}
        env:
          LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_LIB_DIR: ${{ runner.temp }}/abx-lib
          ABXPKG_INSTALL_TIMEOUT: "900"
          ABX_CI_DATA_DIR: ${{ runner.temp }}/archivebox-ci-data
        run: |
          set -euo pipefail
          mkdir -p "$LIB_DIR" "$ABX_CI_DATA_DIR"
          echo "LIB_DIR=$LIB_DIR" >> "$GITHUB_ENV"
          echo "ABXPKG_LIB_DIR=$ABXPKG_LIB_DIR" >> "$GITHUB_ENV"

          uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox install search_backend_sonic

          candidate="$(uv run --project "$GITHUB_WORKSPACE" --directory "$ABX_CI_DATA_DIR" --no-sync --no-sources archivebox shell -c 'from archivebox.machine.models import Binary; binary = Binary.objects.filter(name="sonic", status="installed").order_by("-modified_at").first(); print(binary.abspath if binary else "")' | tail -n 1)"
          if [ ! -x "$candidate" ]; then
            echo "ArchiveBox did not install an executable Sonic binary: ${candidate:-not found}" >&2
            exit 1
          fi
          version="$("$candidate" --version || true)"
          echo "Using ArchiveBox-installed Sonic: $candidate (${version:-version unknown})"
          echo "SONIC_BINARY=$candidate" >> "$GITHUB_ENV"
          echo "$(dirname "$candidate")" >> "$GITHUB_PATH"

      - name: Run plugin tests - ${{ matrix.plugin.name }}
        env:
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
          TWOCAPTCHA_API_KEY: ${{ secrets.TWOCAPTCHA_API_KEY }}
          API_KEY_2CAPTCHA: ${{ secrets.TWOCAPTCHA_API_KEY }}
        run: |
          uv run --no-sync --no-sources bash ./bin/test_plugins.sh "${{ matrix.plugin.plugin }}" --no-coverage