Strip null (\x00) characters from the URLs in LinkAnnotations (issue 6832)

Apparently some PDF files can have annotations with `URI` entries ending with `null` characters, thus breaking the links.
To handle this edge-case of bad PDFs, this patch moves the already existing utility function from `ui_utils.js` into `util.js`, in order to fix those URLs.

Fixes 6832.

src/display/annotation_layer.js

@@ -33,6 +33,7 @@ var AnnotationType = sharedUtil.AnnotationType;
 var Util = sharedUtil.Util;
 var isExternalLinkTargetSet = sharedUtil.isExternalLinkTargetSet;
 var LinkTargetStringMap = sharedUtil.LinkTargetStringMap;
+var removeNullCharacters = sharedUtil.removeNullCharacters;
 var warn = sharedUtil.warn;
 var CustomStyle = displayDOMUtils.CustomStyle;
 
@@ -232,7 +233,8 @@ var LinkAnnotationElement = (function LinkAnnotationElementClosure() {
       this.container.className = 'linkAnnotation';
 
       var link = document.createElement('a');
-      link.href = link.title = this.data.url || '';
+      link.href = link.title = (this.data.url ?
+                                removeNullCharacters(this.data.url) : '');
 
       if (this.data.url && isExternalLinkTargetSet()) {
         link.target = LinkTargetStringMap[PDFJS.externalLinkTarget];

src/shared/util.js

@@ -504,6 +504,16 @@ var XRefParseException = (function XRefParseExceptionClosure() {
   return XRefParseException;
 })();
 
+var NullCharactersRegExp = /\x00/g;
+
+function removeNullCharacters(str) {
+  if (typeof str !== 'string') {
+    warn('The argument for removeNullCharacters must be a string.');
+    return str;
+  }
+  return str.replace(NullCharactersRegExp, '');
+}
+PDFJS.removeNullCharacters = removeNullCharacters;
 
 function bytesToString(bytes) {
   assert(bytes !== null && typeof bytes === 'object' &&
@@ -1690,6 +1700,7 @@ exports.log2 = log2;
 exports.readInt8 = readInt8;
 exports.readUint16 = readUint16;
 exports.readUint32 = readUint32;
+exports.removeNullCharacters = removeNullCharacters;
 exports.shadow = shadow;
 exports.string32 = string32;
 exports.stringToBytes = stringToBytes;

test/unit/ui_utils_spec.js

@@ -1,20 +1,8 @@
-/* globals expect, it, describe, binarySearchFirstItem, removeNullCharacters */
+/* globals expect, it, describe, binarySearchFirstItem */
 
 'use strict';
 
 describe('ui_utils', function() {
-  describe('removeNullCharacters', function() {
-    it('should not modify string without null characters', function() {
-      var str = 'string without null chars';
-      expect(removeNullCharacters(str)).toEqual('string without null chars');
-    });
-
-    it('should modify string with null characters', function() {
-      var str = 'string\x00With\x00Null\x00Chars';
-      expect(removeNullCharacters(str)).toEqual('stringWithNullChars');
-    });
-  });
-
   describe('binary search', function() {
     function isTrue(boolean) {
       return boolean;

test/unit/util_spec.js

@@ -1,5 +1,6 @@
 /* globals expect, it, describe, combineUrl, Dict, isDict, Name, PDFJS,
-           stringToPDFString, isExternalLinkTargetSet, LinkTarget */
+           stringToPDFString, isExternalLinkTargetSet, LinkTarget,
+           removeNullCharacters */
 
 'use strict';
 
@@ -125,4 +126,16 @@ describe('util', function() {
     // Reset the state.
     PDFJS.externalLinkTarget = previousExternalLinkTarget;
   });
+
+  describe('removeNullCharacters', function() {
+    it('should not modify string without null characters', function() {
+      var str = 'string without null chars';
+      expect(removeNullCharacters(str)).toEqual('string without null chars');
+    });
+
+    it('should modify string with null characters', function() {
+      var str = 'string\x00With\x00Null\x00Chars';
+      expect(removeNullCharacters(str)).toEqual('stringWithNullChars');
+    });
+  });
 });

web/pdf_attachment_view.js

@@ -12,7 +12,7 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-/* globals getFileName, removeNullCharacters */
+/* globals getFileName, PDFJS */
 
 'use strict';
 
@@ -89,7 +89,7 @@ var PDFAttachmentView = (function PDFAttachmentViewClosure() {
         div.className = 'attachmentsItem';
         var button = document.createElement('button');
         this._bindLink(button, item.content, filename);
-        button.textContent = removeNullCharacters(filename);
+        button.textContent = PDFJS.removeNullCharacters(filename);
         div.appendChild(button);
         this.container.appendChild(div);
       }

web/pdf_outline_view.js

@@ -12,7 +12,7 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-/* globals removeNullCharacters */
+/* globals PDFJS */
 
 'use strict';
 
@@ -137,7 +137,7 @@ var PDFOutlineView = (function PDFOutlineViewClosure() {
           div.className = 'outlineItem';
           var element = document.createElement('a');
           this._bindLink(element, item);
-          element.textContent = removeNullCharacters(item.title);
+          element.textContent = PDFJS.removeNullCharacters(item.title);
           div.appendChild(element);
 
           if (item.items.length > 0) {

web/ui_utils.js

@@ -23,12 +23,6 @@ var MAX_AUTO_SCALE = 1.25;
 var SCROLLBAR_PADDING = 40;
 var VERTICAL_PADDING = 5;
 
-var NullCharactersRegExp = /\x00/g;
-
-function removeNullCharacters(str) {
-  return str.replace(NullCharactersRegExp, '');
-}
-
 function getFileName(url) {
   var anchor = url.indexOf('#');
   var query = url.indexOf('?');