repo sync

.github/workflows/check-all-english-links.yml

@@ -18,16 +18,11 @@ jobs:
       run: npm run build
     - name: Run script
       run: script/check-english-links.js > broken_links.md
-    - name: Check if any broken links
+    - if: ${{ failure() }} 
+      name: Get title for issue
       id: check
-      run: |
-        if [ "$(grep '0 broken links found' broken_links.md)" ]; then
-          echo ::set-output name=continue::no
-        else
-          echo "::set-output name=continue::yes"
-          echo "::set-output name=title::$(head -1 broken_links.md)"
-        fi
-    - if: ${{ steps.check.outputs.continue == 'yes' }}
+      run: echo "::set-output name=title::$(head -1 broken_links.md)"
+    - if: ${{ failure() }}
       name: Create issue from file
       uses: peter-evans/create-issue-from-file@35e304e2a12caac08c568247a2cb46ecd0c3ecc5
       with:

script/content-migrations/README.md

@@ -0,0 +1,7 @@
+# Content migration scripts
+
+This directory stores scripts that modify content and/or data files. Because
+writers are updating content all the time, scripts in here require more
+cross-team coordination and planning before they are run. Make sure to consider
+whether a script added here also needs to be run on translation files or if we
+can wait for the changes to come in organically via Crowdin.

script/content-migrations/update-developer-site-links.js

@@ -0,0 +1,73 @@
+#!/usr/bin/env node
+
+const fs = require('fs')
+const path = require('path')
+const walk = require('walk-sync')
+const frontmatter = require('@github-docs/frontmatter')
+const loadPages = require('../../lib/pages')
+const patterns = require('../../lib/patterns')
+const loadRedirects = require('../../lib/redirects/precompile')
+const allVersions = Object.keys(require('../../lib/all-versions'))
+
+// get all content and data files
+const files = ['content', 'data'].map(dir => {
+  return walk(path.join(process.cwd(), dir), { includeBasePath: true, directories: false })
+    .filter(file => file.endsWith('.md') && !file.endsWith('README.md'))
+}).flat()
+
+// match [foo](/v3) and [bar](/v4) Markdown links
+const linkRegex = new RegExp('\\(/v[34].*?\\)', 'g')
+
+main()
+
+async function main () {
+  // we need to load the pages so we can get the redirects
+  const englishPages = (await loadPages()).filter(p => p.languageCode === 'en')
+  const redirects = await loadRedirects(englishPages)
+
+  for (const file of files) {
+    const { data, content } = frontmatter(fs.readFileSync(file, 'utf8'))
+
+    const links = content.match(linkRegex)
+    if (!links) continue
+
+    // remove parentheses: (/v3) -> /v3
+    // also remove trailing slash before closing parens if there is one
+    const devLinks = links
+      .map(link => link.replace('(', '').replace(/\/?\)/, ''))
+
+    let newContent = content
+
+    for (const devLink of devLinks) {
+      const [link, fragment] = devLink.split(/\/?#/)
+
+      let redirect = redirects[link]
+
+      if (!redirect) {
+        console.log(`no redirect found for ${devLink} in ${file}`)
+        continue
+      }
+
+      // do some cleanup
+      redirect = redirect
+        // remove language code segment
+        .replace(patterns.getLanguageCode, '')
+        // remove version segment
+        .replace(new RegExp(`/(${allVersions.join('|')})`), '')
+
+      // re-add the fragment
+      const newLink = fragment
+        ? redirect + '#' + fragment
+        : redirect
+
+      // first remove any trailing slashes from the old link,
+      // then replace with the new link
+      newContent = newContent
+        .replace(`${devLink}/`, devLink)
+        .replace(devLink, newLink)
+    }
+
+    fs.writeFileSync(file, frontmatter.stringify(newContent, data, { lineWidth: 10000 }))
+  }
+  console.log('Done!')
+}