Better error handling in link normalizer + more tests for edge cases

lib/helpers/normalize_link.js

@@ -5,5 +5,13 @@ var replaceEntities = require('../common/utils').replaceEntities;
 
 
 module.exports = function normalizeLink(url) {
-  return encodeURI(decodeURI(replaceEntities(url)));
+  var normalized = replaceEntities(url);
+
+  // We don't care much about result of mailformed URIs,
+  // but shoud not throw exception.
+  try {
+    normalized = decodeURI(normalized);
+  } catch (__) {}
+
+  return encodeURI(normalized);
 };

lib/parser_inline.js

@@ -3,8 +3,9 @@
 'use strict';
 
 
-var Ruler       = require('./ruler');
-var StateInline = require('./rules_inline/state_inline');
+var Ruler           = require('./ruler');
+var StateInline     = require('./rules_inline/state_inline');
+var replaceEntities = require('./common/utils').replaceEntities;
 
 ////////////////////////////////////////////////////////////////////////////////
 // Parser rules
@@ -30,7 +31,10 @@ var _rules = [
 var BAD_PROTOCOLS = [ 'vbscript', 'javascript', 'file' ];
 
 function validateLink(url) {
-  var str = decodeURI(url).trim().toLowerCase();
+  var str = url.trim().toLowerCase();
+
+  // Care about digital entities "javascript:alert(1)"
+  str = replaceEntities(str);
 
   if (str.indexOf(':') >= 0 && BAD_PROTOCOLS.indexOf(str.split(':')[0]) >= 0) {
     return false;

test/fixtures/remarkable/commonmark_extras.txt

@@ -123,3 +123,12 @@ Autolinks do not allow escaping:
 .
 <p><a href="http://example.com/%5C%5B%5C">http://example.com/\[\</a></p>
 .
+
+
+Should not throw exception on mailformed URI
+
+.
+[foo](<&#x25;test>)
+.
+<p><a href="%25test">foo</a></p>
+.

test/fixtures/remarkable/xss.txt

@@ -10,47 +10,70 @@ Should not allow some protocols in links and images
 
 .
 [xss link](javascript:alert(1))
+
+[xss link](JAVASCRIPT:alert(1))
+
+[xss link](vbscript:alert(1))
+
+[xss link](VBSCRIPT:alert(1))
+
+[xss link](file:///123)
 .
 <p>[xss link](javascript:alert(1))</p>
+<p>[xss link](JAVASCRIPT:alert(1))</p>
+<p>[xss link](vbscript:alert(1))</p>
+<p>[xss link](VBSCRIPT:alert(1))</p>
+<p>[xss link](file:///123)</p>
 .
 
+
 .
-[xss link](JAVASCRIPT:alert(1))
+[xss link](&#34;&#62;&#60;script&#62;alert&#40;&#34;xss&#34;&#41;&#60;/script&#62;)
 .
-<p>[xss link](JAVASCRIPT:alert(1))</p>
+<p><a href="%22%3E%3Cscript%3Ealert(%22xss%22)%3C/script%3E">xss link</a></p>
 .
 
 .
-[xss link](vbscript:alert(1))
+[xss link](<javascript:alert(1)>)
 .
-<p>[xss link](vbscript:alert(1))</p>
+<p>[xss link](&lt;javascript:alert(1)&gt;)</p>
 .
 
 .
-[xss link](VBSCRIPT:alert(1))
+[xss link](javascript&#x3A;alert(1))
 .
-<p>[xss link](VBSCRIPT:alert(1))</p>
+<p>[xss link](javascript:alert(1))</p>
 .
 
+
+Image parser use the same code base.
+
 .
-[xss link](file:///123)
+![xss link](javascript:alert(1))
 .
-<p>[xss link](file:///123)</p>
+<p>![xss link](javascript:alert(1))</p>
 .
 
 
+Autolinks
+
 .
-[xss link](&#34;&#62;&#60;script&#62;alert&#40;&#34;xss&#34;&#41;&#60;/script&#62;)
+<javascript&#x3A;alert(1)>
+
+<javascript:alert(1)>
 .
-<p><a href="%22%3E%3Cscript%3Ealert(%22xss%22)%3C/script%3E">xss link</a></p>
+<p>&lt;javascript:alert(1)&gt;</p>
+<p>&lt;javascript:alert(1)&gt;</p>
 .
 
 
-Image parser use the same code base.
+Linkifier
 
 .
-![xss link](javascript:alert(1))
+javascript&#x3A;alert(1)
+
+javascript:alert(1)
 .
-<p>![xss link](javascript:alert(1))</p>
+<p>javascript:alert(1)</p>
+<p>javascript:alert(1)</p>
 .
-