Fix AddEnglishSpecifier() producing illegal HTML (#3293)

* Update gulpfile.mjs * Update regex to exclude <title> * Finalized regex * Reworked AddEnglishSpecifier to improve build performance * Fixed replacements inside html-tags * Readded english-texts ---------
corona-warn-app · Jan 27, 2023 · eeb2c20 · eeb2c20
1 parent 2f8b4bd
commit eeb2c20
Show file tree

Hide file tree

Showing 4 changed files with 117 additions and 15 deletions.
diff --git a/gulpfile.mjs b/gulpfile.mjs
@@ -11,6 +11,7 @@ import replace from 'gulp-replace';
 import sitemap from 'gulp-sitemap';
 import sourcemaps from 'gulp-sourcemaps';
 import webp from 'gulp-webp';
+import change from 'gulp-change';
 
 import autoprefixer from 'autoprefixer';
 import browser from 'browser-sync';
@@ -539,9 +540,18 @@ function deleteTmpFiles(done) {
 
 function AddEnglishSpecifier() {
   const data = JSON.parse(fs.readFileSync('src/data/english-texts.json', 'utf8'));
-  let task = gulp.src([PATHS.dist + '/**/*.html']);
-  data.texts.forEach((value) => {
-    task = task.pipe(replace(' ' + value + ' ', `<span lang="en"> ${value} </span>`));
-  });
-  return task.pipe(gulp.dest(PATHS.dist));
+  return gulp
+    .src([PATHS.dist + '/**/*.html'])
+    .pipe(
+      change(function () {
+        // split HTML to exclude the <head>-section from replacement
+        const splittedHtml = this.originalContent.split('</head>');
+        const regex = new RegExp(`(?<!\w)(${data.texts.join('|')})(?![^<]*>)`, 'g');
+        splittedHtml[1] = splittedHtml[1].replace(regex, (match) => {
+          return `<span lang="en">${match}</span>`;
+        });
+        return splittedHtml.join('</head>');
+      })
+    )
+    .pipe(gulp.dest(PATHS.dist));
 }
diff --git a/package-lock.json b/package-lock.json
diff --git a/package.json b/package.json
@@ -37,6 +37,7 @@
     "dayjs": "^1.11.7",
     "google-protobuf": "^3.21.2",
     "gulp": "^4.0.2",
+    "gulp-change": "^1.0.2",
     "gulp-clean-css": "^4.3.0",
     "gulp-cli": "^2.3.0",
     "gulp-if": "^3.0.0",

diff --git a/src/data/english-texts.json b/src/data/english-texts.json
@@ -4,6 +4,7 @@
         "API",
         "APIs",
         "App-Identifier",
+        "Apple App Store",
         "Apple Exposure Notifications API",
         "BLE",
         "BLE-Messung",
@@ -46,6 +47,7 @@
         "Google Exposure Notifications API",
         "Google Mobile Services",
         "Google Play Services",
+        "Google Play Store",
         "HTTPS",
         "In-App-Reset",
         "Interoperability Gateway",
@@ -80,6 +82,7 @@
         "RPIs",
         "SafetyNet",
         "SARS-CoV-2-Exposition",
+        "Science",
         "Science-Blog",
         "Scoping-Document",
         "Screenshots",