fix(lib): improve head and body regex in domparser.js

Because the head and body regexes test against the closing tag, this causes html with unclosed head or body to not be parsed correctly. For example, given the following: ```js parse('<html><body>'); ``` The expected output is: ``` [ { type: 'tag', name: 'html', attribs: {}, children: [ { type: 'tag', name: 'body', attribs: {}, children: [], next: null, prev: null, parent: [Circular] } ], next: null, prev: null, parent: null } ] ``` But the actual output is: ``` [ { "next": null, "prev": null, "parent": null, "name": "html", "attribs": {}, "type": "tag", "children": [] } ] ``` The fix is to update the regex to use the opening tag instead of the closing tag. Add test case. Fixes #18
remarkablemark · Nov 4, 2019 · 457bb58 · 457bb58
1 parent c2665b4
commit 457bb58
Show file tree

Hide file tree

Showing 2 changed files with 20 additions and 4 deletions.
diff --git a/lib/domparser.js b/lib/domparser.js
@@ -5,8 +5,8 @@ var HTML_TAG_NAME = 'html';
 var BODY_TAG_NAME = 'body';
 var HEAD_TAG_NAME = 'head';
 var FIRST_TAG_REGEX = /<([a-zA-Z]+[0-9]?)/; // e.g., <h1>
-var HEAD_REGEX = /<\/head>/i;
-var BODY_REGEX = /<\/body>/i;
+var HEAD_REGEX = /<head.*>/i;
+var BODY_REGEX = /<body.*>/i;
 // http://www.w3.org/TR/html/syntax.html#void-elements
 var VOID_ELEMENTS_REGEX = /<(area|base|br|col|embed|hr|img|input|keygen|link|menuitem|meta|param|source|track|wbr)(.*?)\/?>/gi;
 
@@ -36,7 +36,7 @@ if (typeof window.DOMParser === 'function') {
    */
   parseFromString = function domStringParser(html, tagName) {
     if (tagName) {
-      html = ['<', tagName, '>', html, '</', tagName, '>'].join('');
+      html = '<' + tagName + '>' + html + '</' + tagName + '>';
     }
     // because IE9 only supports MIME type 'text/xml', void elements need to be self-closed
     if (isIE9) {
@@ -117,6 +117,7 @@ function domparser(html) {
   // try to match first tag
   var tagName;
   var match = html.match(FIRST_TAG_REGEX);
+
   if (match && match[1]) {
     tagName = match[1].toLowerCase();
   }
@@ -131,7 +132,7 @@ function domparser(html) {
         doc = parseFromString(html);
 
         // the created document may come with filler head/body elements,
-        // so ake sure to remove them if they don't actually exist
+        // so make sure to remove them if they don't actually exist
         if (!HEAD_REGEX.test(html)) {
           element = doc.getElementsByTagName(HEAD_TAG_NAME)[0];
           if (element) element.parentNode.removeChild(element);

diff --git a/test/cases/html.js b/test/cases/html.js
@@ -31,6 +31,21 @@ module.exports = [
     data: '<html><body></body></html>',
     skip: isPhantomJS
   },
+  {
+    name: 'unclosed html and head tags',
+    data: '<html><head>',
+    skip: isPhantomJS
+  },
+  {
+    name: 'unclosed html and body tags',
+    data: '<html><body>',
+    skip: isPhantomJS
+  },
+  {
+    name: 'unclosed html, head, and body tags',
+    data: '<html><head><body>',
+    skip: isPhantomJS
+  },
 
   {
     name: 'empty head',