Add vnu-jar testing from Bootstrap, adapted for this repo. (#2469)

If Java isn't present, the test will be skipped. On CI Java is available, though.

.github/workflows/test.yml

@@ -16,6 +16,7 @@ jobs:
 
       - run: node --version
       - run: npm --version
+      - run: java -version
 
       - name: Install npm dependencies
         run: npm ci
@@ -28,5 +29,8 @@ jobs:
       - name: Run tests
         run: npm test
 
+      - name: Run HTML validator
+        run: npm run test:html
+
       - name: Run linkinator
         run: npm run test:linkinator

package.json

@@ -21,6 +21,7 @@
     "test:lint:md": "markdownlint \"**/*.md\" -i \"**/node_modules/**\"",
     "test:lint:stylint": "stylint layouts/css",
     "test:lint": "npm-run-all --parallel test:lint:*",
+    "test:html": "node scripts/vnu-jar.js",
     "test:unit": "tape tests/**/*.test.js | faucet"
   },
   "repository": "nodejs/nodejs.org",
@@ -78,6 +79,7 @@
     "st": "^1.2.2",
     "standard": "^14.3.1",
     "stylint": "^2.0.0",
-    "tape": "^4.11.0"
+    "tape": "^4.11.0",
+    "vnu-jar": "19.9.4"
   }
 }

scripts/vnu-jar.js

@@ -0,0 +1,54 @@
+#!/usr/bin/env node
+
+'use strict'
+
+const childProcess = require('child_process')
+const vnu = require('vnu-jar')
+
+childProcess.exec('java -version', (error, stdout, stderr) => {
+  if (error) {
+    console.error('Skipping vnu-jar test; Java is missing.')
+    return
+  }
+
+  const is32bitJava = !stderr.match(/64-Bit/)
+
+  // vnu-jar accepts multiple ignores joined with a `|`.
+  // Also note that the ignores are regular expressions.
+  const ignores = [
+    // Low priority warnings
+    'Section lacks heading.*',
+    // This seems to happen due to some Unicode characters
+    'Text run is not in Unicode Normalization Form C.',
+    // These are real errors but hard to tackle.
+    // We should fix them eventually
+    'Table column.*',
+    // These happen due to the commented out English HTML code some translations have...
+    // They should be removed at some point
+    'The document is not mappable to XML 1.0 due to two consecutive hyphens in a comment.'
+  ].join('|')
+
+  const args = [
+    '-jar',
+    vnu,
+    '--asciiquotes',
+    '--skip-non-html',
+    // Ignore the language code warnings
+    '--no-langdetect',
+    '--Werror',
+    '--errors-only',
+    `--filterpattern "${ignores}"`,
+    'build/'
+  ]
+
+  // For the 32-bit Java we need to pass `-Xss512k`
+  if (is32bitJava) {
+    args.splice(0, 0, '-Xss512k')
+  }
+
+  return childProcess.spawn('java', args, {
+    shell: true,
+    stdio: 'inherit'
+  })
+    .on('exit', process.exit)
+})