pr-duplicate-check-writer.yml

name: PR Duplicate Check Writer

on:
  workflow_run:
    workflows: ["PR Duplicate Check"]
    types: [completed]

permissions:
  actions: read
  issues: write
  pull-requests: write

concurrency:
  group: pdcw-${{ github.event.workflow_run.head_repository.id || 'unknown-repo' }}-${{ github.event.workflow_run.head_branch || 'unknown-branch' }}
  cancel-in-progress: true

jobs:
  process-safe-output:
    runs-on: ubuntu-latest
    if: github.event.workflow_run.event == 'pull_request'
    steps:
      - name: Download agent artifact
        id: download-agent
        continue-on-error: true
        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
        with:
          name: agent
          path: ${{ runner.temp }}/pr-duplicate-agent
          run-id: ${{ github.event.workflow_run.id }}
          github-token: ${{ github.token }}

      - name: Download PR context artifact
        id: download-context
        continue-on-error: true
        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
        with:
          name: pr-duplicate-check-context
          path: ${{ runner.temp }}/pr-duplicate-context
          run-id: ${{ github.event.workflow_run.id }}
          github-token: ${{ github.token }}

      - name: Validate and publish safe comment output
        if: steps.download-agent.outcome == 'success'
        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
        with:
          script: |
            const fs = require('fs');
            const path = require('path');

            const workflowRun = context.payload.workflow_run;
            const artifactRoot = path.join(process.env.RUNNER_TEMP, 'pr-duplicate-agent');
            const contextRoot = path.join(process.env.RUNNER_TEMP, 'pr-duplicate-context');

            function readJsonIfExists(filePath) {
              if (!fs.existsSync(filePath)) return null;
              const stat = fs.statSync(filePath);
              if (stat.size > 1024 * 1024) throw new Error(`${filePath} exceeds 1 MiB`);
              return JSON.parse(fs.readFileSync(filePath, 'utf8'));
            }

            const prContext =
              readJsonIfExists(path.join(contextRoot, 'pr-context.json')) ||
              readJsonIfExists(path.join(artifactRoot, 'pr-context.json'));
            if (!prContext || prContext.schema_version !== 'pr-duplicate-check-context/v1') {
              throw new Error('Missing or invalid pr-context.json artifact.');
            }
            if (!Number.isInteger(prContext.pr_number) || prContext.pr_number < 1) {
              throw new Error('Invalid PR context pr_number.');
            }
            if (!/^[0-9a-f]{40}$/i.test(String(prContext.head_sha || ''))) {
              throw new Error('Invalid PR context head_sha.');
            }
            if (workflowRun.event !== 'pull_request') {
              throw new Error('Invalid workflow_run event.');
            }
            if (String(prContext.run_id || '') !== String(workflowRun.id)) {
              throw new Error('PR context run_id did not match workflow_run.');
            }
            if (prContext.head_sha !== workflowRun.head_sha) {
              throw new Error('PR context head_sha did not match workflow_run.');
            }

            const prNumber = prContext.pr_number;
            const { data: pr } = await github.rest.pulls.get({
              owner: context.repo.owner,
              repo: context.repo.repo,
              pull_number: prNumber,
            });
            if (pr.state !== 'open') {
              core.info(`Skipping non-open PR #${prNumber}.`);
              return;
            }
            const expectedBaseRepository = `${context.repo.owner}/${context.repo.repo}`.toLowerCase();
            const runHeadRepository = String(workflowRun.head_repository?.full_name || '');
            const runHeadRepositoryParts = runHeadRepository.split('/');
            const runHeadRef = String(workflowRun.head_branch || '');
            if (String(pr.base?.repo?.full_name || '').toLowerCase() !== expectedBaseRepository) {
              throw new Error(`PR #${prNumber} base repository ${pr.base?.repo?.full_name || '<unknown>'} is not this repository.`);
            }
            if (pr.head.sha !== workflowRun.head_sha) {
              core.warning(`Skipping stale PR duplicate output for PR #${prNumber}: artifact head ${prContext.head_sha}, current head ${pr.head.sha}`);
              return;
            }
            if (
              runHeadRepositoryParts.length !== 2 ||
              !runHeadRepositoryParts[0] ||
              !runHeadRepositoryParts[1] ||
              !runHeadRef ||
              String(pr.head?.repo?.full_name || '').toLowerCase() !== runHeadRepository.toLowerCase() ||
              String(pr.head?.ref || '') !== runHeadRef
            ) {
              throw new Error(`PR #${prNumber} head did not match workflow_run.`);
            }

            const workflowRunPullRequests = Array.isArray(workflowRun.pull_requests) ? workflowRun.pull_requests : [];
            if (workflowRunPullRequests.length > 0) {
              if (!workflowRunPullRequests.some((pullRequest) => pullRequest.number === prNumber)) {
                throw new Error(`PR context number ${prNumber} was not present in workflow_run.pull_requests.`);
              }
            } else {
              const candidatePullRequests = await github.paginate(github.rest.pulls.list, {
                owner: context.repo.owner,
                repo: context.repo.repo,
                state: 'open',
                head: `${runHeadRepositoryParts[0]}:${runHeadRef}`,
                per_page: 100,
              });
              const trustedMatches = candidatePullRequests.filter((candidate) =>
                candidate.head?.sha === workflowRun.head_sha &&
                String(candidate.head?.ref || '') === runHeadRef &&
                String(candidate.head?.repo?.full_name || '').toLowerCase() === runHeadRepository.toLowerCase() &&
                String(candidate.base?.repo?.full_name || '').toLowerCase() === expectedBaseRepository
              );
              if (trustedMatches.length !== 1 || trustedMatches[0].number !== prNumber) {
                throw new Error(`PR context number ${prNumber} could not be uniquely associated with workflow_run.`);
              }
            }

            function collectCandidate(value, targetCandidates) {
              if (!value || typeof value !== 'object') return;
              const tool = value.tool || value.name || value.tool_name || value.type || value.action;
              const args = value.arguments || value.args || value.input || value.params || value.data || value;
              if (tool === 'add_comment') {
                targetCandidates.push(args);
              }
            }

            function collectCandidatesFromJson(value) {
              const collected = [];
              if (value) {
                if (Array.isArray(value.items)) {
                  for (const item of value.items) collectCandidate(item, collected);
                } else if (Array.isArray(value)) {
                  for (const item of value) collectCandidate(item, collected);
                } else {
                  collectCandidate(value, collected);
                }
              }
              return collected;
            }

            const agentOutput = readJsonIfExists(path.join(artifactRoot, 'agent_output.json'));
            let candidates = collectCandidatesFromJson(agentOutput);

            if (candidates.length === 0) {
              const safeOutputsPath = path.join(artifactRoot, 'safeoutputs.jsonl');
              if (fs.existsSync(safeOutputsPath)) {
                const stat = fs.statSync(safeOutputsPath);
                if (stat.size > 1024 * 1024) throw new Error('safeoutputs.jsonl exceeds 1 MiB');
                const lines = fs.readFileSync(safeOutputsPath, 'utf8').split(/\r?\n/).filter(Boolean).slice(0, 100);
                const safeOutputCandidates = [];
                for (const line of lines) collectCandidate(JSON.parse(line), safeOutputCandidates);
                candidates = safeOutputCandidates;
              }
            }

            const stripControlCharacters = (value) =>
              String(value).replace(/[\u0000-\u0008\u000B\u000C\u000E-\u001F\u007F]/g, '');
            const neutralizeMentions = (value) =>
              value.replace(/@([A-Za-z0-9][A-Za-z0-9-]{0,38})/g, '@\u200B$1');
            const sanitizeCommentBody = (value, maxLength) => {
              const sanitized = neutralizeMentions(stripControlCharacters(value));
              return sanitized.length > maxLength ? sanitized.slice(0, maxLength) : sanitized;
            };

            const validComments = [];
            for (const candidate of candidates) {
              if (!candidate || typeof candidate !== 'object') continue;
              const body = candidate.body;
              if (typeof body !== 'string' || body.trim() === '' || body.length > 65000) continue;
              if (candidate.item_number !== undefined && Number(candidate.item_number) !== prNumber) continue;
              if (candidate.repo !== undefined && String(candidate.repo) !== `${context.repo.owner}/${context.repo.repo}`) continue;
              validComments.push(body);
            }

            if (validComments.length === 0) {
              core.info('No valid add_comment safe output was found.');
              return;
            }
            if (validComments.length > 1) {
              throw new Error(`Expected at most one add_comment safe output, found ${validComments.length}.`);
            }

            const marker = '<!-- pr-duplicate-check -->';
            const runUrl = workflowRun.html_url;
            const footer = `<!-- synchronized from read-only PR Duplicate Check run ${workflowRun.id}: ${runUrl} -->`;
            const maxSafeOutputBodyLength = 65000 - marker.length - footer.length - 4;
            const safeOutputBody = sanitizeCommentBody(validComments[0], Math.max(0, maxSafeOutputBodyLength));
            if (safeOutputBody.trim() === '') {
              core.info('No non-empty safe comment body remained after sanitization.');
              return;
            }
            const body = [
              marker,
              safeOutputBody,
              '',
              footer,
            ].join('\n');

            const comments = await github.paginate(github.rest.issues.listComments, {
              owner: context.repo.owner,
              repo: context.repo.repo,
              issue_number: prNumber,
              per_page: 100,
            });
            const matchingComments = comments.filter((comment) =>
              comment.user?.login === 'github-actions[bot]' && String(comment.body || '').includes(marker)
            );

            const [canonical, ...duplicates] = matchingComments;
            if (canonical) {
              await github.rest.issues.updateComment({
                owner: context.repo.owner,
                repo: context.repo.repo,
                comment_id: canonical.id,
                body,
              });
            } else {
              await github.rest.issues.createComment({
                owner: context.repo.owner,
                repo: context.repo.repo,
                issue_number: prNumber,
                body,
              });
            }

            for (const duplicate of duplicates) {
              await github.rest.issues.deleteComment({
                owner: context.repo.owner,
                repo: context.repo.repo,
                comment_id: duplicate.id,
              }).catch(() => {});
            }

      - name: Note missing artifact
        if: steps.download-agent.outcome != 'success'
        run: echo "No PR Duplicate Check agent artifact was available; nothing to synchronize."