kucherenko__jscpd/.github/workflows/rust.yml

name: Rust

on:
  push:
    branches: [main, master, "heist/*"]
  pull_request:
    branches: [main, master, "heist/*"]

permissions:
  contents: read

jobs:
  build:
    permissions:
      contents: read
      checks: write # dorny/test-reporter posts check runs
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest, macos-latest, windows-latest]

    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

      - uses: dtolnay/rust-toolchain@4360b52568e2003a75bf9bc1d59f33a8e3fc893c # stable
        with:
          toolchain: "1.97.1"
          components: clippy, rustfmt

      - uses: Swatinem/rust-cache@6323deb102c322ba6fcbdcafc7e3dddab59af2b6 # v2
        with:
          workspaces: "rust -> rust/target"
          cache-on-failure: true

      - name: Install cargo-nextest
        uses: taiki-e/install-action@d438492cf8a250514fa2d34b30bc3c0dc37c65ff # v2.87.8
        with:
          tool: nextest

      - name: Build
        working-directory: rust
        run: cargo build --workspace

      - name: Test
        working-directory: rust
        run: cargo nextest run --workspace --profile ci

      - name: Upload test results
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: nextest-results-${{ matrix.os }}
          path: rust/target/nextest/ci/nextest.xml
          if-no-files-found: warn

      - name: Test Report
        if: always()
        continue-on-error: true
        uses: dorny/test-reporter@a43b3a5f7366b97d083190328d2c652e1a8b6aa2 # v3.0.0
        with:
          name: Rust Tests (${{ matrix.os }})
          path: rust/target/nextest/ci/nextest.xml
          reporter: java-junit
          fail-on-error: false

      - name: Clippy
        working-directory: rust
        run: cargo clippy --workspace -- -D warnings

      - name: Format check
        working-directory: rust
        run: cargo fmt --check

      # Bash-only steps below run on Linux; Windows defaults to pwsh.
      - name: Tokenizer purity check
        if: runner.os == 'Linux'
        run: |
          cd rust
          TOKENIZER_DEPS=$(cargo metadata --format-version 1 | \
            python3 -c "
          import sys, json
          meta = json.load(sys.stdin)
          tokenizer_id = next(p['id'] for p in meta['packages'] if p['name'] == 'cpd-tokenizer')
          dep_names = set()
          queue = [tokenizer_id]
          id_to_pkg = {p['id']: p for p in meta['packages']}
          resolve_map = {n['id']: [d['pkg'] for d in n['deps']] for n in meta['resolve']['nodes']}
          visited = set()
          while queue:
              pid = queue.pop()
              if pid in visited: continue
              visited.add(pid)
              if pid in id_to_pkg:
                  dep_names.add(id_to_pkg[pid]['name'])
              for dep_id in resolve_map.get(pid, []):
                  queue.append(dep_id)
          print('\n'.join(dep_names))
          ")
          echo "cpd-tokenizer transitive deps: $TOKENIZER_DEPS"
          for crate in tokio async-std walkdir ignore jwalk; do
            if echo "$TOKENIZER_DEPS" | grep -q "^$crate$"; then
              echo "ERROR: cpd-tokenizer depends on I/O crate: $crate"
              exit 1
            fi
          done
          echo "Tokenizer purity check: PASS"

      - name: publish=false check
        if: runner.os == 'Linux'
        run: |
          cd rust
          for toml in crates/*/Cargo.toml; do
            if ! grep -q 'publish = false' "$toml"; then
              echo "ERROR: $toml missing 'publish = false'"
              exit 1
            fi
          done
          echo "publish=false check: PASS"

      - name: cargo-deny
        if: matrix.os == 'ubuntu-latest'
        working-directory: rust
        run: cargo install cargo-deny --locked && cargo deny check

  msrv:
    name: MSRV
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

      - name: Read declared MSRV
        id: msrv
        run: |
          version=$(sed -nE 's/^rust-version = "([0-9.]+)"$/\1/p' rust/Cargo.toml | head -1)
          if [ -z "$version" ]; then
            echo "ERROR: no rust-version found in rust/Cargo.toml"
            exit 1
          fi
          echo "Declared MSRV: $version"
          echo "version=$version" >> "$GITHUB_OUTPUT"

      - uses: dtolnay/rust-toolchain@4360b52568e2003a75bf9bc1d59f33a8e3fc893c # stable
        with:
          toolchain: ${{ steps.msrv.outputs.version }}

      - uses: Swatinem/rust-cache@6323deb102c322ba6fcbdcafc7e3dddab59af2b6 # v2
        with:
          workspaces: "rust -> rust/target"
          key: msrv-${{ steps.msrv.outputs.version }}
          cache-on-failure: true

      # `+toolchain` outranks rust-toolchain.toml, which pins a newer compiler
      # for day-to-day development. Without it this job would silently test the
      # pinned version instead of the MSRV.
      - name: cargo check at MSRV
        working-directory: rust
        run: cargo +${{ steps.msrv.outputs.version }} check --workspace --all-targets --locked

  # End-to-end run of the release binary over the multi-format corpus in
  # fixtures/ (550 files across 130+ formats). Unit tests cover the parts;
  # this catches a reporter or walker regression that only shows on the
  # real corpus.
  smoke:
    name: smoke test (fixtures corpus)
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

      - uses: dtolnay/rust-toolchain@4360b52568e2003a75bf9bc1d59f33a8e3fc893c # stable
        with:
          toolchain: "1.97.1"

      - uses: Swatinem/rust-cache@6323deb102c322ba6fcbdcafc7e3dddab59af2b6 # v2
        with:
          workspaces: "rust -> rust/target"
          key: smoke-release
          cache-on-failure: true

      - name: Build release binary
        working-directory: rust
        run: cargo build --release -p jscpd --locked

      - name: Scan fixtures (console + json)
        working-directory: rust
        run: ./target/release/jscpd ../fixtures --reporters console,json --output ../smoke-report --min-tokens 50

      - name: Assert JSON report
        run: |
          report=smoke-report/jscpd-report.json
          test -s "$report"
          clones=$(jq '.statistics.total.clones' "$report")
          sources=$(jq '.statistics.total.sources' "$report")
          formats=$(jq '.statistics.formats | keys | length' "$report")
          duplicates=$(jq '.duplicates | length' "$report")
          echo "clones=$clones sources=$sources formats=$formats duplicates=$duplicates"
          [ "$clones" -gt 0 ] || { echo "expected statistics.total.clones > 0"; exit 1; }
          [ "$duplicates" -eq "$clones" ] || { echo "duplicates[] length ($duplicates) != statistics.total.clones ($clones)"; exit 1; }
          [ "$sources" -ge 300 ] || { echo "expected at least 300 scanned files, got $sources"; exit 1; }
          [ "$formats" -ge 100 ] || { echo "expected at least 100 detected formats, got $formats"; exit 1; }

      - name: SARIF reporter
        working-directory: rust
        run: |
          ./target/release/jscpd ../fixtures --reporters sarif,silent --output ../smoke-report --min-tokens 50
          jq -e '.runs[0].results | length > 0' ../smoke-report/jscpd-report.sarif

      - name: HTML reporter
        working-directory: rust
        run: |
          ./target/release/jscpd ../fixtures --reporters html,silent --output ../smoke-report --min-tokens 50
          test -s ../smoke-report/jscpd-report.html