Report files reset during the translation batch (#23159)

* Add script to report reset files

* Has the workflow report the reset files

* Fix script name

* Filter report by language

* Update .github/workflows/create-translation-batch-pr.yml

* Use build in array methods

Author: mjacobus

.github/workflows/create-translation-batch-pr.yml

@@ -136,17 +136,25 @@ jobs:
           node script/i18n/lint-translation-files.js --check rendering
           git add translations/${{ matrix.language }} && git commit -m "Run script/i18n/lint-translation-files.js --check rendering" || echo "Nothing to commit"
 
+      - name: Check in CSV report
+        run: |
+          mkdir -p log
+          csvFile=log/${{ matrix.language_code }}-resets.csv
+          script/i18n/report-reset-files.js --report-type=csv --language=${{ matrix.language_code }} --log-file=/tmp/batch.log > $csvFile
+          git add -f $csvFile && git commit -m "Check in ${{ matrix.language }} CSV report" || echo "Nothing to commit"
+
       - name: Create Pull Request
         env:
           GITHUB_TOKEN: ${{ secrets.DOCUBOT_REPO_PAT }}
         # We'll try to create the pull request based on the changes we pushed up in the branch.
         # If there are actually no differences between the branch and `main`, we'll delete it.
         run: |
+          script/i18n/report-reset-files.js --report-type=pull-request-body --language=${{ matrix.language_code }} --log-file=/tmp/batch.log > /tmp/pr-body.txt
           git push origin ${{ steps.set-branch.outputs.BRANCH_NAME }}
-          gh pr create -t "New translation batch for ${{ matrix.language }}" \
+          gh pr create --title "New translation batch for ${{ matrix.language }}" \
             --base=main \
             --head=${{ steps.set-branch.outputs.BRANCH_NAME }} \
-            -b "New batch for ${{ matrix.language }} created by [this workflow]($GITHUB_SERVER_URL/$GITHUB_REPOSITORY/actions/runs/$GITHUB_RUN_ID)" ||  git push origin :${{ steps.set-branch.outputs.BRANCH_NAME }}
+            --body-file /tmp/pr-body.txt || git push origin :${{ steps.set-branch.outputs.BRANCH_NAME }}
 
       # When the maximum execution time is reached for this job, Actions cancels the workflow run.
       # This emits a notification for the first responder to triage.

script/i18n/report-reset-files.js

@@ -0,0 +1,85 @@
+#!/usr/bin/env node
+
+import program from 'commander'
+import fs from 'fs'
+import languages from '../../lib/languages.js'
+
+const defaultWorkflowUrl = [
+  process.env.GITHUB_SERVER_URL,
+  process.env.GITHUB_REPOSITORY,
+  'actions/runs',
+  process.env.GITHUB_RUN_ID,
+].join('/')
+
+const reportTypes = {
+  'pull-request-body': pullRequestBodyReport,
+  csv: csvReport,
+}
+
+program
+  .description('Reads a translation batch log and generates a report')
+  .requiredOption('--language <language>', 'The language to compare')
+  .requiredOption('--log-file <log-file>', 'The batch log file')
+  .requiredOption(
+    '--report-type <report-type>',
+    'The batch log file, I.E: ' + Object.keys(reportTypes).join(', ')
+  )
+  .option('--workflow-url <workflow-url>', 'The workflow url', defaultWorkflowUrl)
+  .parse(process.argv)
+
+const options = program.opts()
+const language = languages[options.language]
+const { logFile, workflowUrl, reportType } = options
+
+if (!Object.keys(reportTypes).includes(reportType)) {
+  throw new Error(`Invalid report type: ${reportType}`)
+}
+
+const logFileContents = fs.readFileSync(logFile, 'utf8')
+
+const revertLines = logFileContents
+  .split('\n')
+  .filter((line) => line.match(/^-> reverted to English/))
+  .filter((line) => line.match(language.dir))
+
+const reportEntries = revertLines.sort().map((line) => {
+  const [, file, reason] = line.match(/^-> reverted to English: (.*) Reason: (.*)$/)
+  return { file, reason }
+})
+
+function pullRequestBodyReport() {
+  const body = [
+    `New translation batch for ${language.name}. Product of [this workflow](${workflowUrl}).`,
+    '\n',
+    `## ${reportEntries.length} files reverted.`,
+  ]
+
+  const filesByReason = {}
+
+  reportEntries.forEach(({ file, reason }) => {
+    filesByReason[reason] = filesByReason[reason] || []
+    filesByReason[reason].push(file)
+  })
+
+  Object.keys(filesByReason)
+    .sort()
+    .forEach((reason) => {
+      const files = filesByReason[reason]
+      body.push(`\n### ${reason}`)
+      body.push(`\n${files.length} files:\n`)
+      const checkBoxes = files.map((file) => `- [ ] ${file}`)
+      body.push(checkBoxes)
+    })
+
+  return body.join('\n')
+}
+
+function csvReport() {
+  const lines = reportEntries.map(({ file, reason }) => {
+    return [file, reason].join(',')
+  })
+
+  return ['file,reason', lines].flat().join('\n')
+}
+
+console.log(reportTypes[reportType]())