sync-grammar.yml

name: Sync Tree-sitter Grammar

on:
  schedule:
    - cron: "0 6 * * 1"  # Weekly on Monday at 6 AM UTC
  workflow_dispatch:
    inputs:
      lf_version:
        description: "LF release tag (e.g. v0.11.0). Leave empty for latest."
        required: false
        type: string

jobs:
  sync:
    runs-on: ubuntu-latest
    steps:
      - name: Checkout lf.nvim
        uses: actions/checkout@v4

      - name: Resolve LF version
        id: resolve
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          if [ -n "${{ inputs.lf_version }}" ]; then
            version="${{ inputs.lf_version }}"
          else
            version=$(gh api repos/lf-lang/lingua-franca/releases \
              --jq '[.[] | select(.tag_name != "nightly" and .prerelease == false)][0].tag_name')
          fi
          echo "version=$version" >> "$GITHUB_OUTPUT"
          echo "Resolved LF version: $version"

      - name: Fetch upstream Xtext grammar
        run: |
          curl -fSL "https://raw.githubusercontent.com/lf-lang/lingua-franca/${{ steps.resolve.outputs.version }}/core/src/main/java/org/lflang/LinguaFranca.xtext" \
            -o /tmp/LinguaFranca.xtext

      - name: Check if grammar changed
        id: check
        run: |
          new_hash=$(sha256sum /tmp/LinguaFranca.xtext | cut -d' ' -f1)
          if [ -f .xtext-grammar-sha256 ]; then
            old_hash=$(cat .xtext-grammar-sha256)
          else
            old_hash=""
          fi

          if [ "$new_hash" = "$old_hash" ]; then
            echo "Grammar unchanged, skipping."
            echo "changed=false" >> "$GITHUB_OUTPUT"
          else
            echo "Grammar changed ($old_hash -> $new_hash)."
            echo "changed=true" >> "$GITHUB_OUTPUT"
            echo "$new_hash" > .xtext-grammar-sha256
            echo "${{ steps.resolve.outputs.version }}" > .lf-upstream-version
          fi

      - name: Set up Python
        if: steps.check.outputs.changed == 'true'
        uses: actions/setup-python@v5
        with:
          python-version: "3.12"

      - name: Set up Node.js
        if: steps.check.outputs.changed == 'true'
        uses: actions/setup-node@v4
        with:
          node-version: "20"

      - name: Install tree-sitter CLI
        if: steps.check.outputs.changed == 'true'
        run: npm install -g tree-sitter-cli

      - name: Generate grammar.js
        if: steps.check.outputs.changed == 'true'
        run: |
          python -m xtext2ts \
            --xtext /tmp/LinguaFranca.xtext \
            --output tree-sitter-lf/grammar.js

      - name: Generate parser
        if: steps.check.outputs.changed == 'true'
        run: cd tree-sitter-lf && tree-sitter generate

      - name: Checkout upstream test files at anchored version
        if: steps.check.outputs.changed == 'true'
        uses: actions/checkout@v4
        with:
          repository: lf-lang/lingua-franca
          ref: ${{ steps.resolve.outputs.version }}
          sparse-checkout: test
          path: /tmp/lingua-franca

      - name: Validate against upstream test files
        if: steps.check.outputs.changed == 'true'
        run: |
          cd tree-sitter-lf
          total=$(find /tmp/lingua-franca/test -name "*.lf" | wc -l)
          errors=$(find /tmp/lingua-franca/test -name "*.lf" \
            | xargs tree-sitter parse --quiet 2>&1 \
            | grep -c "ERROR" || true)

          echo "Validation against LF ${{ steps.resolve.outputs.version }}: $errors failures out of $total files"

          # Show failed files if any
          if [ "$errors" -gt 0 ]; then
            echo "Failed:"
            find /tmp/lingua-franca/test -name "*.lf" \
              | xargs tree-sitter parse --quiet 2>&1 \
              | grep "ERROR" \
              | sed 's|/tmp/lingua-franca/||'
          fi

          # All upstream tests must pass
          if [ "$errors" -gt 0 ]; then
            echo "Parse errors detected, aborting."
            exit 1
          fi

      - name: Create Pull Request
        if: steps.check.outputs.changed == 'true'
        uses: peter-evans/create-pull-request@v7
        with:
          token: ${{ secrets.GITHUB_TOKEN }}
          commit-message: "update tree-sitter grammar for LF ${{ steps.resolve.outputs.version }}"
          branch: update-grammar/${{ steps.resolve.outputs.version }}
          title: "Update tree-sitter grammar for LF ${{ steps.resolve.outputs.version }}"
          body: |
            ## Summary
            - Auto-generated tree-sitter grammar from upstream Xtext grammar
            - LF version: `${{ steps.resolve.outputs.version }}`
            - Generated by `xtext2ts` converter

            ## Test plan
            - [ ] Verify `tree-sitter generate` succeeds
            - [ ] Verify test .lf files parse without errors
            - [ ] Test syntax highlighting in Neovim
          labels: automated