playwright.yml

name: Playwright E2E Tests

on:
  push:
    branches:
      - "develop"
  pull_request:
    types: [opened, synchronize, reopened]

  # Allows workflow to be called from other workflows
  workflow_call:
    inputs:
      ref:
        required: true
        type: string

# Avoid duplicate workflows on same branch
concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}-playwright
  cancel-in-progress: true

jobs:
  playwright-e2e-tests:
    runs-on: ubuntu-latest-64-cores
    timeout-minutes: 30

    defaults:
      run:
        shell: bash

    steps:
      - name: Checkout Streamlit code
        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
        with:
          ref: ${{ inputs.ref }}
          persist-credentials: false
          submodules: "recursive"
          fetch-depth: 2
      - name: Set Python version vars
        uses: ./.github/actions/build_info
      - name: Setup virtual env
        uses: ./.github/actions/make_init
        with:
          python_version: ${{ env.PYTHON_MAX_VERSION }}
      - name: Install playwright
        uses: ./.github/actions/playwright_install
      - name: Build frontend
        run: make frontend-with-profiler
      - name: Get short SHA
        id: short_sha
        run: |
          if [[ "${{ github.event_name }}" == "pull_request" ]]; then
            echo "sha_short=$(echo ${{ github.event.pull_request.head.sha }} | cut -c1-6)" >> $GITHUB_OUTPUT
          else
            echo "sha_short=$(echo ${{ github.sha }} | cut -c1-6)" >> $GITHUB_OUTPUT
          fi
      - name: Run playwright tests
        # With psutil installed, pytest-xdist's `-n auto` uses physical cores (32 on 64-core runner)
        run: |
          cd e2e_playwright && rm -rf ./test-results
          uv run pytest --ignore ./custom_components --ignore ./load_testing --browser webkit --browser chromium --browser firefox -n auto --reruns 1 -m "not performance"
      - name: Upload test statistics
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        if: always()
        with:
          name: playwright_test_stats
          path: e2e_playwright/test-results/test-stats.json
          if-no-files-found: ignore
      - name: Upload failed test results
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        if: always()
        with:
          name: playwright_test_results_${{ steps.short_sha.outputs.sha_short }}
          path: e2e_playwright/test-results

  check-e2e-test-count:
    runs-on: ubuntu-latest
    needs: playwright-e2e-tests
    if: github.event_name == 'pull_request' && github.repository == 'streamlit/streamlit'
    continue-on-error: true

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

    env:
      NEW_TEST_THRESHOLD: 30

    steps:
      - name: Download current test stats
        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
        continue-on-error: true
        with:
          name: playwright_test_stats
          path: current_stats

      - name: Get latest develop run ID
        id: get-latest-run
        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
        with:
          script: |
            const { data: runs } = await github.rest.actions.listWorkflowRuns({
              owner: context.repo.owner,
              repo: context.repo.repo,
              workflow_id: 'playwright.yml',
              branch: 'develop',
              status: 'success',
              per_page: 1
            });

            if (runs.workflow_runs.length === 0) {
              core.warning('No successful workflow runs found on develop branch');
              return;
            }

            core.setOutput('run_id', runs.workflow_runs[0].id);

      - name: Download develop test stats
        if: steps.get-latest-run.outputs.run_id
        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
        continue-on-error: true
        with:
          repository: streamlit/streamlit
          run-id: ${{ steps.get-latest-run.outputs.run_id }}
          name: playwright_test_stats
          path: develop_stats
          github-token: ${{ secrets.GITHUB_TOKEN }}

      - name: Compare test counts and comment
        if: steps.get-latest-run.outputs.run_id
        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
        with:
          script: |
            const fs = require('fs');
            const path = require('path');
            const threshold = +process.env.NEW_TEST_THRESHOLD;

            const loadStats = (dir) => {
              try {
                const jsonPath = path.join(process.cwd(), dir, 'test-stats.json');
                if (fs.existsSync(jsonPath)) {
                  return JSON.parse(fs.readFileSync(jsonPath, 'utf8'));
                }
                console.log(`File not found: ${jsonPath}`);
                return null;
              } catch (e) {
                console.log(`Error loading stats from ${dir}:`, e.message);
                return null;
              }
            };

            const currentStats = loadStats('current_stats');
            const developStats = loadStats('develop_stats');

            if (!currentStats) {
              console.log('Missing current test stats, skipping comparison.');
              return;
            }

            if (!developStats) {
              console.log('Missing develop test stats, skipping comparison.');
              return;
            }

            const currentTotal = currentStats.summary?.total_tests || 0;
            const developTotal = developStats.summary?.total_tests || 0;
            const newTests = currentTotal - developTotal;

            console.log(`Current PR test count: ${currentTotal}`);
            console.log(`Develop test count: ${developTotal}`);
            console.log(`New tests: ${newTests}`);
            console.log(`Threshold: ${threshold}`);

            // Check if this PR is from a fork
            const isForkPR = context.payload.pull_request?.head?.repo?.fork === true;
            if (isForkPR) {
              console.log('PR is from a fork - skipping comment (insufficient permissions)');
              console.log('Test count information is available in the job logs above.');
              return;
            }

            const isSignificant = newTests > threshold;
            const commentIdentifier = '<!-- STREAMLIT-E2E-TEST-COUNT-CHECK -->';

            // Find existing comment
            const { data: comments } = await github.rest.issues.listComments({
              owner: context.repo.owner,
              repo: context.repo.repo,
              issue_number: context.issue.number,
            });
            const existingComment = comments.find(comment => comment.body.includes(commentIdentifier));

            // Only comment if tests increased significantly (> threshold)
            if (isSignificant || existingComment) {
              let header;
              let message;
              if (isSignificant) {
                header = `### 🧪 Significant E2E test count increase detected`;
                message = `This PR has **added ${newTests} E2E test cases** (threshold: ${threshold})`;
              } else {
                // Previous comment exists but change is now within range
                header = `### ✅ E2E test count change is within normal range`;
                if (newTests === 0) {
                  message = `This PR has **no change** in E2E test count`;
                } else if (newTests > 0) {
                  message = `This PR has **added ${newTests} E2E test cases**`;
                } else {
                  message = `This PR has **removed ${Math.abs(newTests)} E2E test cases**`;
                }
              }

              const lines = [
                header,
                '',
                message,
                '',
                `- Current PR: ${currentTotal} tests`,
                `- Latest develop: ${developTotal} tests`,
              ];

              if (isSignificant) {
                lines.push(
                  '',
                  '> ⚠️ **Note:** E2E tests are expensive to run. Please ensure you\'re following best practices:',
                  '> - Prefer aggregated scenario tests over many micro-tests',
                  '> - Add tests to existing files when they fit the scope',
                  '> - Test each aspect only once per browser run'
                );
              }

              const commentBody = lines.join('\n');
              const fullCommentBody = `${commentIdentifier}\n${commentBody}`;

              if (existingComment) {
                await github.rest.issues.updateComment({
                  owner: context.repo.owner,
                  repo: context.repo.repo,
                  comment_id: existingComment.id,
                  body: fullCommentBody
                });
                console.log('Updated existing E2E test count comment');
              } else {
                await github.rest.issues.createComment({
                  owner: context.repo.owner,
                  repo: context.repo.repo,
                  issue_number: context.issue.number,
                  body: fullCommentBody
                });
                console.log('Created new E2E test count comment');
              }
            } else {
              console.log('Test count change is not significant (added <= threshold), skipping comment');
            }