fix($sanitize): Use same whitelist mechanism as $compile does.

`$sanitize` now uses the same mechanism as `$compile` to validate uris.
By this, the validation in `$sanitize` is more general and can be
configured in the same way as the one in `$compile`.

Changes
- Creates the new private service `$$sanitizeUri`.
- Moves related specs from `compileSpec.js` into `sanitizeUriSpec.js`.
- Refactors the `linky` filter to be less dependent on `$sanitize`
  internal functions.

Fixes #3748.

5 files changed, 155 insertions, 48 deletions

diff --git a/src/AngularPublic.js b/src/AngularPublic.js
index eb97b4c5..d2c325c5 100644
--- a/src/AngularPublic.js
+++ b/src/AngularPublic.js
@@ -65,6 +65,7 @@
     $ParseProvider,
     $RootScopeProvider,
     $QProvider,
+    $$SanitizeUriProvider,
     $SceProvider,
     $SceDelegateProvider,
     $SnifferProvider,
@@ -136,6 +137,10 @@ function publishExternalAPI(angular){
 
   angularModule('ng', ['ngLocale'], ['$provide',
     function ngModule($provide) {
+      // $$sanitizeUriProvider needs to be before $compileProvider as it is used by it.
+      $provide.provider({
+        $$sanitizeUri: $$SanitizeUriProvider
+      });
       $provide.provider('$compile', $CompileProvider).
         directive({
             a: htmlAnchorDirective,
diff --git a/src/ng/compile.js b/src/ng/compile.js
index 13fb9682..54d2dc9f 100644
--- a/src/ng/compile.js
+++ b/src/ng/compile.js
@@ -493,14 +493,12 @@ var $compileMinErr = minErr('$compile');
  *
  * @description
  */
-$CompileProvider.$inject = ['$provide'];
-function $CompileProvider($provide) {
+$CompileProvider.$inject = ['$provide', '$$sanitizeUriProvider'];
+function $CompileProvider($provide, $$sanitizeUriProvider) {
   var hasDirectives = {},
       Suffix = 'Directive',
       COMMENT_DIRECTIVE_REGEXP = /^\s*directive\:\s*([\d\w\-_]+)\s+(.*)$/,
-      CLASS_DIRECTIVE_REGEXP = /(([\d\w\-_]+)(?:\:([^;]+))?;?)/,
-      aHrefSanitizationWhitelist = /^\s*(https?|ftp|mailto|tel|file):/,
-      imgSrcSanitizationWhitelist = /^\s*(https?|ftp|file):|data:image\//;
+      CLASS_DIRECTIVE_REGEXP = /(([\d\w\-_]+)(?:\:([^;]+))?;?)/;
 
   // Ref: http://developers.whatwg.org/webappapis.html#event-handler-idl-attributes
   // The assumption is that future DOM event attribute names will begin with
@@ -584,10 +582,11 @@ function $CompileProvider($provide) {
    */
   this.aHrefSanitizationWhitelist = function(regexp) {
     if (isDefined(regexp)) {
-      aHrefSanitizationWhitelist = regexp;
+      $$sanitizeUriProvider.aHrefSanitizationWhitelist(regexp);
       return this;
+    } else {
+      return $$sanitizeUriProvider.aHrefSanitizationWhitelist();
     }
-    return aHrefSanitizationWhitelist;
   };
 
 
@@ -614,18 +613,18 @@ function $CompileProvider($provide) {
    */
   this.imgSrcSanitizationWhitelist = function(regexp) {
     if (isDefined(regexp)) {
-      imgSrcSanitizationWhitelist = regexp;
+      $$sanitizeUriProvider.imgSrcSanitizationWhitelist(regexp);
       return this;
+    } else {
+      return $$sanitizeUriProvider.imgSrcSanitizationWhitelist();
     }
-    return imgSrcSanitizationWhitelist;
   };
 
-
   this.$get = [
             '$injector', '$interpolate', '$exceptionHandler', '$http', '$templateCache', '$parse',
-            '$controller', '$rootScope', '$document', '$sce', '$animate',
+            '$controller', '$rootScope', '$document', '$sce', '$animate', '$$sanitizeUri',
     function($injector,   $interpolate,   $exceptionHandler,   $http,   $templateCache,   $parse,
-             $controller,   $rootScope,   $document,   $sce,   $animate) {
+             $controller,   $rootScope,   $document,   $sce,   $animate,   $$sanitizeUri) {
 
     var Attributes = function(element, attr) {
       this.$$element = element;
@@ -730,16 +729,7 @@ function $CompileProvider($provide) {
         // sanitize a[href] and img[src] values
         if ((nodeName === 'A' && key === 'href') ||
             (nodeName === 'IMG' && key === 'src')) {
-          // NOTE: urlResolve() doesn't support IE < 8 so we don't sanitize for that case.
-          if (!msie || msie >= 8 ) {
-            normalizedVal = urlResolve(value).href;
-            if (normalizedVal !== '') {
-              if ((key === 'href' && !normalizedVal.match(aHrefSanitizationWhitelist)) ||
-                  (key === 'src' && !normalizedVal.match(imgSrcSanitizationWhitelist))) {
-                this[key] = value = 'unsafe:' + normalizedVal;
-              }
-            }
-          }
+          this[key] = value = $$sanitizeUri(value, key === 'src');
         }
 
         if (writeAttr !== false) {
diff --git a/src/ng/sanitizeUri.js b/src/ng/sanitizeUri.js
new file mode 100644
index 00000000..97325094
--- /dev/null
+++ b/src/ng/sanitizeUri.js
@@ -0,0 +1,74 @@
+'use strict';
+
+/**
+ * @description
+ * Private service to sanitize uris for links and images. Used by $compile and $sanitize.
+ */
+function $$SanitizeUriProvider() {
+  var aHrefSanitizationWhitelist = /^\s*(https?|ftp|mailto|tel|file):/,
+    imgSrcSanitizationWhitelist = /^\s*(https?|ftp|file):|data:image\//;
+
+  /**
+   * @description
+   * Retrieves or overrides the default regular expression that is used for whitelisting of safe
+   * urls during a[href] sanitization.
+   *
+   * The sanitization is a security measure aimed at prevent XSS attacks via html links.
+   *
+   * Any url about to be assigned to a[href] via data-binding is first normalized and turned into
+   * an absolute url. Afterwards, the url is matched against the `aHrefSanitizationWhitelist`
+   * regular expression. If a match is found, the original url is written into the dom. Otherwise,
+   * the absolute url is prefixed with `'unsafe:'` string and only then is it written into the DOM.
+   *
+   * @param {RegExp=} regexp New regexp to whitelist urls with.
+   * @returns {RegExp|ng.$compileProvider} Current RegExp if called without value or self for
+   *    chaining otherwise.
+   */
+  this.aHrefSanitizationWhitelist = function(regexp) {
+    if (isDefined(regexp)) {
+      aHrefSanitizationWhitelist = regexp;
+      return this;
+    }
+    return aHrefSanitizationWhitelist;
+  };
+
+
+  /**
+   * @description
+   * Retrieves or overrides the default regular expression that is used for whitelisting of safe
+   * urls during img[src] sanitization.
+   *
+   * The sanitization is a security measure aimed at prevent XSS attacks via html links.
+   *
+   * Any url about to be assigned to img[src] via data-binding is first normalized and turned into
+   * an absolute url. Afterwards, the url is matched against the `imgSrcSanitizationWhitelist`
+   * regular expression. If a match is found, the original url is written into the dom. Otherwise,
+   * the absolute url is prefixed with `'unsafe:'` string and only then is it written into the DOM.
+   *
+   * @param {RegExp=} regexp New regexp to whitelist urls with.
+   * @returns {RegExp|ng.$compileProvider} Current RegExp if called without value or self for
+   *    chaining otherwise.
+   */
+  this.imgSrcSanitizationWhitelist = function(regexp) {
+    if (isDefined(regexp)) {
+      imgSrcSanitizationWhitelist = regexp;
+      return this;
+    }
+    return imgSrcSanitizationWhitelist;
+  };
+
+  this.$get = function() {
+    return function sanitizeUri(uri, isImage) {
+      var regex = isImage ? imgSrcSanitizationWhitelist : aHrefSanitizationWhitelist;
+      var normalizedVal;
+      // NOTE: urlResolve() doesn't support IE < 8 so we don't sanitize for that case.
+      if (!msie || msie >= 8 ) {
+        normalizedVal = urlResolve(uri).href;
+        if (normalizedVal !== '' && !normalizedVal.match(regex)) {
+          return 'unsafe:'+normalizedVal;
+        }
+      }
+      return uri;
+    };
+  };
+}
diff --git a/src/ngSanitize/filter/linky.js b/src/ngSanitize/filter/linky.js
index 39494f2d..2c05d84e 100644
--- a/src/ngSanitize/filter/linky.js
+++ b/src/ngSanitize/filter/linky.js
@@ -1,6 +1,6 @@
 'use strict';
 
-/* global htmlSanitizeWriter: false */
+/* global sanitizeText: false */
 
 /**
  * @ngdoc filter
@@ -100,7 +100,7 @@
      </doc:scenario>
    </doc:example>
  */
-angular.module('ngSanitize').filter('linky', function() {
+angular.module('ngSanitize').filter('linky', ['$sanitize', function($sanitize) {
   var LINKY_URL_REGEXP =
         /((ftp|https?):\/\/|(mailto:)?[A-Za-z0-9._%+-]+@)\S*[^\s.;,(){}<>]/,
       MAILTO_REGEXP = /^mailto:/;
@@ -110,28 +110,40 @@ angular.module('ngSanitize').filter('linky', function() {
     var match;
     var raw = text;
     var html = [];
-    // TODO(vojta): use $sanitize instead
-    var writer = htmlSanitizeWriter(html);
     var url;
     var i;
-    var properties = {};
-    if (angular.isDefined(target)) {
-      properties.target = target;
-    }
     while ((match = raw.match(LINKY_URL_REGEXP))) {
       // We can not end in these as they are sometimes found at the end of the sentence
       url = match[0];
       // if we did not match ftp/http/mailto then assume mailto
       if (match[2] == match[3]) url = 'mailto:' + url;
       i = match.index;
-      writer.chars(raw.substr(0, i));
-      properties.href = url;
-      writer.start('a', properties);
-      writer.chars(match[0].replace(MAILTO_REGEXP, ''));
-      writer.end('a');
+      addText(raw.substr(0, i));
+      addLink(url, match[0].replace(MAILTO_REGEXP, ''));
       raw = raw.substring(i + match[0].length);
     }
-    writer.chars(raw);
-    return html.join('');
+    addText(raw);
+    return $sanitize(html.join(''));
+
+    function addText(text) {
+      if (!text) {
+        return;
+      }
+      html.push(sanitizeText(text));
+    }
+
+    function addLink(url, text) {
+      html.push('<a ');
+      if (angular.isDefined(target)) {
+        html.push('target="');
+        html.push(target);
+        html.push('" ');
+      }
+      html.push('href="');
+      html.push(url);
+      html.push('">');
+      addText(text);
+      html.push('</a>');
+    }
   };
-});
+}]);
diff --git a/src/ngSanitize/sanitize.js b/src/ngSanitize/sanitize.js
index 7bd9aae3..5d378b02 100644
--- a/src/ngSanitize/sanitize.js
+++ b/src/ngSanitize/sanitize.js
@@ -46,6 +46,8 @@ var $sanitizeMinErr = angular.$$minErr('$sanitize');
  *   it into the returned string, however, since our parser is more strict than a typical browser
  *   parser, it's possible that some obscure input, which would be recognized as valid HTML by a
  *   browser, won't make it through the sanitizer.
+ *   The whitelist is configured using the functions `aHrefSanitizationWhitelist` and
+ *   `imgSrcSanitizationWhitelist` of {@link ng.$compileProvider `$compileProvider`}.
  *
  * @param {string} html Html input.
  * @returns {string} Sanitized html.
@@ -128,11 +130,24 @@ var $sanitizeMinErr = angular.$$minErr('$sanitize');
    </doc:scenario>
    </doc:example>
  */
-var $sanitize = function(html) {
+function $SanitizeProvider() {
+  this.$get = ['$$sanitizeUri', function($$sanitizeUri) {
+    return function(html) {
+      var buf = [];
+      htmlParser(html, htmlSanitizeWriter(buf, function(uri, isImage) {
+        return !/^unsafe/.test($$sanitizeUri(uri, isImage));
+      }));
+      return buf.join('');
+    };
+  }];
+}
+
+function sanitizeText(chars) {
   var buf = [];
-    htmlParser(html, htmlSanitizeWriter(buf));
-    return buf.join('');
-};
+  var writer = htmlSanitizeWriter(buf, angular.noop);
+  writer.chars(chars);
+  return buf.join('');
+}
 
 
 // Regular Expressions for parsing tags and attributes
@@ -145,7 +160,6 @@ var START_TAG_REGEXP =
   COMMENT_REGEXP = /<!--(.*?)-->/g,
   DOCTYPE_REGEXP = /<!DOCTYPE([^>]*?)>/i,
   CDATA_REGEXP = /<!\[CDATA\[(.*?)]]>/g,
-  URI_REGEXP = /^((ftp|https?):\/\/|mailto:|tel:|#)/i,
   // Match everything outside of normal chars and " (quote character)
   NON_ALPHANUMERIC_REGEXP = /([^\#-~| |!])/g;
 
@@ -353,8 +367,18 @@ function htmlParser( html, handler ) {
  */
 var hiddenPre=document.createElement("pre");
 function decodeEntities(value) {
-  hiddenPre.innerHTML=value.replace(/</g,"&lt;");
-  return hiddenPre.innerText || hiddenPre.textContent || '';
+  if (!value) {
+    return '';
+  }
+  // Note: IE8 does not preserve spaces at the start/end of innerHTML
+  var spaceRe = /^(\s*)([\s\S]*?)(\s*)$/;
+  var parts = spaceRe.exec(value);
+  parts[0] = '';
+  if (parts[2]) {
+    hiddenPre.innerHTML=parts[2].replace(/</g,"&lt;");
+    parts[2] = hiddenPre.innerText || hiddenPre.textContent;
+  }
+  return parts.join('');
 }
 
 /**
@@ -384,7 +408,7 @@ function encodeEntities(value) {
  *     comment: function(text) {}
  * }
  */
-function htmlSanitizeWriter(buf){
+function htmlSanitizeWriter(buf, uriValidator){
   var ignore = false;
   var out = angular.bind(buf, buf.push);
   return {
@@ -398,7 +422,9 @@ function htmlSanitizeWriter(buf){
         out(tag);
         angular.forEach(attrs, function(value, key){
           var lkey=angular.lowercase(key);
-          if (validAttrs[lkey]===true && (uriAttrs[lkey]!==true || value.match(URI_REGEXP))) {
+          var isImage = (tag === 'img' && lkey === 'src') || (lkey === 'background');
+          if (validAttrs[lkey] === true &&
+            (uriAttrs[lkey] !== true || uriValidator(value, isImage))) {
             out(' ');
             out(key);
             out('="');
@@ -430,4 +456,4 @@ function htmlSanitizeWriter(buf){
 
 
 // define ngSanitize module and register $sanitize service
-angular.module('ngSanitize', []).value('$sanitize', $sanitize);
+angular.module('ngSanitize', []).provider('$sanitize', $SanitizeProvider);