analyze.yml

name: Submission-Triggered Analyze (via PR)

# 与 auto-analyze.yml 的关系：
# - auto-analyze.yml 服务「定时调度 / 维护者 /analyze 评论」，结果直接 push main
# - 本 workflow 服务「外部用户站点 submit 提交」，结果走 PR 审核后合并
# 两条流并存，触发条件不重叠

on:
  issues:
    types: [opened, labeled]

permissions:
  contents: write
  issues: write
  pull-requests: write

# 按 issue 编号分组：同一 issue 的重复触发（opened+labeled 双发、或手动重打标签）合并成一次，
# 取消旧 run 只留最新；不同 issue 各跑各的（各自 analyze/<slug> 分支，互不干扰）。
concurrency:
  group: analyze-issue-${{ github.event.issue.number }}
  cancel-in-progress: true

jobs:
  analyze:
    # 按事件类型精确 gate，避免 opened 与 labeled 对同一 issue 双触发：
    # - opened：带 analyze-request 标签，或标题以 [Analyze] 开头（无标签兜底）
    # - labeled：只认新加的那枚标签恰为 analyze-request（加别的标签不打扰运行中的分析）
    if: >-
      (github.event.action == 'opened'
        && (contains(github.event.issue.labels.*.name, 'analyze-request')
            || startsWith(github.event.issue.title, '[Analyze]')))
      || (github.event.action == 'labeled'
        && github.event.label.name == 'analyze-request')
    runs-on: ubuntu-latest
    timeout-minutes: 60
    env:
      GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
      MINIMAX_API_KEY: ${{ secrets.MINIMAX_API_KEY }}
      MINIMAX_BASE_URL: ${{ secrets.MINIMAX_BASE_URL }}
      ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
      ISSUE_NUMBER: ${{ github.event.issue.number }}

    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Dump issue body to file
        env:
          BODY: ${{ github.event.issue.body }}
        run: |
          mkdir -p /tmp/runner
          printf '%s' "$BODY" > /tmp/runner/issue_body.md

      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
          cache: "pip"
          cache-dependency-path: requirements-ci.txt

      - name: Install Python deps
        run: pip install -r requirements-ci.txt

      - name: Validate submission
        id: validate
        run: python3 src/scripts/validate_submission.py /tmp/runner/issue_body.md

      - name: Reject & close on validation failure
        if: failure() && steps.validate.outcome == 'failure'
        run: |
          MSG="${{ steps.validate.outputs.comment || '❌ 校验失败' }}"
          gh issue comment "$ISSUE_NUMBER" --body "$MSG"
          gh issue edit "$ISSUE_NUMBER" --add-label "analyze-failed" --remove-label "analyze-request" || true
          gh issue close "$ISSUE_NUMBER"
          exit 1

      - name: Comment in-progress
        run: gh issue comment "$ISSUE_NUMBER" --body "🔧 校验通过，开始分析 \`${{ steps.validate.outputs.repo_url }}\`，约 10-30 分钟。"

      - uses: actions/setup-node@v4
        with: { node-version: "20" }

      # 缓存用户级 tokei 二进制；key 含 setup_ci_env.sh hash → TOKEI_VERSION 改即失效
      - name: Cache tokei binary
        uses: actions/cache@v4
        with:
          path: ~/.local/bin/tokei
          key: tokei-${{ runner.os }}-${{ hashFiles('src/scripts/setup_ci_env.sh') }}

      # 缓存用户级 npm 全局目录（含 @anthropic-ai/claude-code）；同上 key 策略
      - name: Cache npm global (claude-code CLI)
        uses: actions/cache@v4
        with:
          path: ~/.npm-global
          key: npm-global-${{ runner.os }}-${{ hashFiles('src/scripts/setup_ci_env.sh') }}

      - name: Bootstrap CI env (gh / jq / tokei / claude / vendored skills)
        run: bash src/scripts/setup_ci_env.sh

      - name: Run /repo-miner (via run_skill.sh, dual-backend)
        env:
          REPO_URL: ${{ steps.validate.outputs.repo_url }}
          REPORT_PATH: ${{ steps.validate.outputs.report_path }}
        run: |
          PROMPT="/repo-miner $REPO_URL ，中间过程文件输出到 tmp/ ，最终分析报告输出到 $REPORT_PATH"
          bash src/scripts/run_skill.sh "$PROMPT" --timeout 1800 --log "tmp/logs/repo-miner-$ISSUE_NUMBER.log"

      - name: Verify report file
        id: verify
        env:
          REPORT_PATH: ${{ steps.validate.outputs.report_path }}
        run: |
          if [ ! -s "$REPORT_PATH" ]; then
            echo "::error::分析跑完但未产出 $REPORT_PATH"
            exit 1
          fi
          echo "lines=$(wc -l < "$REPORT_PATH")" >> "$GITHUB_OUTPUT"

      - name: Prefix report title（统一加「GitHub推荐：」前缀）
        # 确定性后处理：给新产报告的 H1 加前缀（幂等）。
        # 必须在 build_reports_index 之前，reports.json 索引的 title 才会带上前缀。
        # 只改本次新产报告，不动历史。
        env:
          REPORT_PATH: ${{ steps.validate.outputs.report_path }}
        run: python3 src/scripts/prefix_report_title.py "$REPORT_PATH"

      - name: Refresh site metadata (DB-backed)
        # 阶段 1-4：reports + tags + starred + trending + publish_history 主写 DB
        # JSON/YAML 由 export-json dump；publish_history.jsonl 是 publish 历史的 SoR
        run: |
          python3 src/scripts/init_db.py init
          python3 src/scripts/init_db.py seed-publish
          python3 src/scripts/build_reports_index.py
          python3 src/scripts/extract_tags.py
          python3 src/scripts/parse_starred.py
          python3 src/scripts/seed_trending.py
          python3 src/scripts/init_db.py export-json

      - name: Open PR
        if: success()
        uses: peter-evans/create-pull-request@v6
        with:
          # org 策略禁用了 GITHUB_TOKEN 建 PR，这里改用 PAT（fine-grained：Contents+Pull requests=write）
          # 缺失时回退到 GITHUB_TOKEN（仍会被 org 策略挡住，但不至于让表达式报错）
          token: ${{ secrets.PAT_PR || secrets.GITHUB_TOKEN }}
          branch: analyze/${{ steps.validate.outputs.slug }}
          delete-branch: true
          base: main
          title: "Add report: ${{ steps.validate.outputs.owner }}/${{ steps.validate.outputs.repo }}"
          body: |
            Auto-generated from issue #${{ github.event.issue.number }}.

            - Repo: ${{ steps.validate.outputs.repo_url }}
            - Report: `${{ steps.validate.outputs.report_path }}`
            - Lines: ${{ steps.verify.outputs.lines }}

            Closes #${{ github.event.issue.number }}
          commit-message: "feat(report): ${{ steps.validate.outputs.owner }}/${{ steps.validate.outputs.repo }}"
          add-paths: |
            src/analysis_report/
            src/data/

      - name: Comment success
        if: success()
        run: gh issue comment "$ISSUE_NUMBER" --body "✅ 分析完成，PR 已开，合并后 5 分钟内上线站点。"

      - name: Comment failure
        if: failure() && steps.validate.outcome == 'success'
        run: |
          gh issue comment "$ISSUE_NUMBER" --body "❌ 分析失败。日志：${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
          gh issue edit "$ISSUE_NUMBER" --add-label "analyze-failed" || true

      - name: Upload logs
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: submission-logs-${{ github.event.issue.number }}
          path: |
            tmp/logs/
            tmp/*.md
          retention-days: 14
          if-no-files-found: warn