code-review-sweep.yml

name: Code Review Sweep

on:
  schedule:
    # Every 15 minutes
    - cron: "*/15 * * * *"
  workflow_dispatch:

permissions:
  contents: read

# Prevent overlapping sweeps
concurrency:
  group: code-review-sweep
  cancel-in-progress: false

jobs:
  # ---------------------------------------------------------------------------
  # Job 1: Determine which modules to review
  # ---------------------------------------------------------------------------
  dispatch:
    # Only run on official repo, not forks
    if: github.repository == 'open-telemetry/opentelemetry-java-instrumentation'
    runs-on: ubuntu-latest
    outputs:
      modules: ${{ steps.build-matrix.outputs.modules }}
      has_work: ${{ steps.build-matrix.outputs.has_work }}
    steps:
      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          fetch-depth: 1

      - name: Fetch progress branch
        run: git fetch origin otelbot/code-review-progress || true

      - name: Build review matrix
        id: build-matrix
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          # Read progress from the dedicated orphan branch (if it exists)
          progress=$(git show origin/otelbot/code-review-progress:reviewed.txt 2>/dev/null || true)
          if [[ -n "$progress" ]]; then
            export REVIEW_PROGRESS="$progress"
          fi
          python .github/scripts/code-review/build-review-matrix.py

  # ---------------------------------------------------------------------------
  # Job 2: Walk modules sequentially on a single branch, stopping once the
  # accumulated change set reaches FILE_THRESHOLD modified files. One PR per run.
  # ---------------------------------------------------------------------------
  review:
    needs: dispatch
    if: needs.dispatch.outputs.has_work == 'true'
    runs-on: ubuntu-latest
    environment: protected
    permissions:
      contents: write  # for git push
    env:
      MODULES_JSON: ${{ needs.dispatch.outputs.modules }}
      MODEL: "gpt-5.4"
      # Stop processing further modules once at least this many files have been
      # modified (vs origin/main) at the end of a module.
      FILE_THRESHOLD: 10
      COPILOT_ROOT: /tmp/copilot
      FRAGMENTS_DIR: /tmp/pr-body-fragments
      PROCESSED_MODULES: /tmp/processed-modules.txt
      PR_BODY: /tmp/pr-body.md
    steps:
      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2

      - name: Fetch progress branch
        run: git fetch origin otelbot/code-review-progress || true

      - name: Free disk space
        run: .github/scripts/gha-free-disk-space.sh

      - name: Set up JDK for running Gradle
        uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
        with:
          distribution: temurin
          java-version-file: .java-version

      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
          cache-read-only: true

      - name: Install Copilot CLI
        run: |
          curl -fsSL https://gh.io/copilot-install | bash
          echo "$HOME/.local/bin" >> "$GITHUB_PATH"

      - name: Use CLA approved bot
        run: .github/scripts/use-cla-approved-bot.sh

      - name: Check out review branch
        id: branch
        run: |
          branch="otelbot/code-review-sweep-${GITHUB_RUN_ID}"
          git checkout -B "$branch" origin/main
          echo "name=$branch" >> "$GITHUB_OUTPUT"

      - name: Run Copilot review loop
        id: review-loop
        env:
          COPILOT_GITHUB_TOKEN: ${{ secrets.COPILOT_GITHUB_TOKEN }}
          COPILOT_REVIEW_PROMPT_TEMPLATE: >-
            Review all files under __MODULE_DIR__. Apply safe repository-guideline fixes directly.
            Return ONLY a valid JSON object as your final answer with this exact schema:
            {"summary": string, "changes": [{"path": string, "category": string, "change": string, "reason": string, "line_hint": number|null}], "unresolved": [{"path": string, "reason": string}]}
            Include one changes entry for every file you changed.
            Use concise factual reasons that cite the review guideline or repository rule behind each change.
            In `summary`, `change`, and `reason`, use Markdown inline code backticks around code-like constructs when helpful,
            including annotations, class names, method names, field names, file names, Gradle tasks, commands, flags, and config keys.
            If no safe fixes were applied, still return valid JSON with an empty changes array and a brief summary.
            Do not write markdown and do not wrap the JSON in code fences.
        run: python .github/scripts/code-review/sweep-loop.py

      - name: Upload review diagnostics artifact
        if: always()
        id: upload-review-diagnostics
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: code-review-diagnostics-${{ github.run_id }}
          path: |
            /tmp/copilot/**
            /tmp/processed-modules.txt
          if-no-files-found: ignore

      - name: Assemble PR body
        if: steps.review-loop.outputs.commits_on_branch != '0'
        env:
          ARTIFACT_URL: ${{ steps.upload-review-diagnostics.outputs.artifact-url }}
        run: |
          set -euo pipefail
          {
            echo "Automated code review sweep walked the following modules in order"
            echo "and stopped after accumulating at least ${FILE_THRESHOLD} modified files:"
            echo
            while IFS= read -r m; do
              echo "- \`$m\`"
            done < "$PROCESSED_MODULES"
            echo
            echo "---"
            echo
            for f in "$FRAGMENTS_DIR"/*.md; do
              [[ -f "$f" ]] || continue
              cat "$f"
              echo
            done
            echo "---"
            echo
            echo "[Download code review diagnostics]($ARTIFACT_URL)"
            echo
          } > "$PR_BODY"

      - name: Commit summary
        if: steps.review-loop.outputs.commits_on_branch != '0'
        id: commit
        run: |
          branch="${{ steps.branch.outputs.name }}"
          git push -f origin "$branch"
          echo "pushed=true" >> "$GITHUB_OUTPUT"

      - uses: actions/create-github-app-token@1b10c78c7865c340bc4f6099eb2f838309f1e8c3 # v3.1.1
        id: otelbot-token
        if: steps.commit.outputs.pushed == 'true'
        with:
          app-id: ${{ vars.OTELBOT_APP_ID }}
          private-key: ${{ secrets.OTELBOT_PRIVATE_KEY }}

      - name: Create PR
        if: steps.commit.outputs.pushed == 'true'
        env:
          GH_TOKEN: ${{ steps.otelbot-token.outputs.token }}
        run: |
          branch="${{ steps.branch.outputs.name }}"
          title="Code review sweep (run ${GITHUB_RUN_ID})"
          gh pr create \
            --title "$title" \
            --body-file "$PR_BODY" \
            --base main \
            --head "$branch" \
            --label "automated code review"

      - name: Ensure progress branch exists
        if: steps.review-loop.outputs.processed_count != '0'
        run: |
          if ! git rev-parse --verify origin/otelbot/code-review-progress >/dev/null 2>&1; then
            git checkout --orphan otelbot/code-review-progress
            git reset --hard
            git commit --allow-empty -m "Initialize progress tracking"
            git push origin HEAD:otelbot/code-review-progress || true
          fi

      - name: Check out progress branch
        if: steps.review-loop.outputs.processed_count != '0'
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
          ref: otelbot/code-review-progress
          path: progress

      - name: Mark processed modules as reviewed
        if: steps.review-loop.outputs.processed_count != '0'
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          set -euo pipefail
          cd progress

          git config user.name otelbot
          git config user.email 197425009+otelbot@users.noreply.github.com

          cat "$PROCESSED_MODULES" >> reviewed.txt

          git add reviewed.txt
          git commit -m "Mark $(wc -l < "$PROCESSED_MODULES" | tr -d ' ') module(s) as reviewed"
          git push origin HEAD:otelbot/code-review-progress