From 831c218a9312d15ab57d6d81f84f18fd332a4166 Mon Sep 17 00:00:00 2001 From: Matthew Hodgson Date: Sat, 16 Aug 2014 22:21:52 +0100 Subject: [PATCH] autohyperlink messages using linky --- webclient/index.html | 1 + webclient/js/angular-sanitize.js | 577 ++++++++++++++++++++++++++++++ webclient/room/room-controller.js | 2 +- webclient/room/room.html | 4 +- 4 files changed, 581 insertions(+), 3 deletions(-) create mode 100644 webclient/js/angular-sanitize.js diff --git a/webclient/index.html b/webclient/index.html index 085ac58a8b..ee77dd2faa 100644 --- a/webclient/index.html +++ b/webclient/index.html @@ -9,6 +9,7 @@ + diff --git a/webclient/js/angular-sanitize.js b/webclient/js/angular-sanitize.js new file mode 100644 index 0000000000..d34522ac8d --- /dev/null +++ b/webclient/js/angular-sanitize.js @@ -0,0 +1,577 @@ +/** + * @license AngularJS v1.2.0 + * (c) 2010-2012 Google, Inc. http://angularjs.org + * License: MIT + */ +(function(window, angular, undefined) {'use strict'; + +var $sanitizeMinErr = angular.$$minErr('$sanitize'); + +/** + * @ngdoc overview + * @name ngSanitize + * @description + * + * # ngSanitize + * + * The `ngSanitize` module provides functionality to sanitize HTML. + * + * {@installModule sanitize} + * + *
+ * + * See {@link ngSanitize.$sanitize `$sanitize`} for usage. + */ + +/* + * HTML Parser By Misko Hevery (misko@hevery.com) + * based on: HTML Parser By John Resig (ejohn.org) + * Original code by Erik Arvidsson, Mozilla Public License + * http://erik.eae.net/simplehtmlparser/simplehtmlparser.js + * + * // Use like so: + * htmlParser(htmlString, { + * start: function(tag, attrs, unary) {}, + * end: function(tag) {}, + * chars: function(text) {}, + * comment: function(text) {} + * }); + * + */ + + +/** + * @ngdoc service + * @name ngSanitize.$sanitize + * @function + * + * @description + * The input is sanitized by parsing the html into tokens. All safe tokens (from a whitelist) are + * then serialized back to properly escaped html string. This means that no unsafe input can make + * 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. + * + * @param {string} html Html input. + * @returns {string} Sanitized html. + * + * @example + + + +
+ Snippet: + + + + + + + + + + + + + + + + + + + + + + + + + +
DirectiveHowSourceRendered
ng-bind-htmlAutomatically uses $sanitize
<div ng-bind-html="snippet">
</div>
ng-bind-htmlBypass $sanitize by explicitly trusting the dangerous value + 
<div ng-bind-html="deliberatelyTrustDangerousSnippet()">
+</div>
+
ng-bindAutomatically escapes
<div ng-bind="snippet">
</div>
+
+ + + it('should sanitize the html snippet by default', function() { + expect(using('#bind-html-with-sanitize').element('div').html()). + toBe('

an html\nclick here\nsnippet

'); + }); + + it('should inline raw snippet if bound to a trusted value', function() { + expect(using('#bind-html-with-trust').element("div").html()). + toBe("

an html\n" + + "click here\n" + + "snippet

"); + }); + + it('should escape snippet without any filter', function() { + expect(using('#bind-default').element('div').html()). + toBe("<p style=\"color:blue\">an html\n" + + "<em onmouseover=\"this.textContent='PWN3D!'\">click here</em>\n" + + "snippet</p>"); + }); + + it('should update', function() { + input('snippet').enter('new text'); + expect(using('#bind-html-with-sanitize').element('div').html()).toBe('new text'); + expect(using('#bind-html-with-trust').element('div').html()).toBe( + 'new text'); + expect(using('#bind-default').element('div').html()).toBe( + "new <b onclick=\"alert(1)\">text</b>"); + }); + + + */ +var $sanitize = function(html) { + var buf = []; + htmlParser(html, htmlSanitizeWriter(buf)); + return buf.join(''); +}; + + +// Regular Expressions for parsing tags and attributes +var START_TAG_REGEXP = + /^<\s*([\w:-]+)((?:\s+[\w:-]+(?:\s*=\s*(?:(?:"[^"]*")|(?:'[^']*')|[^>\s]+))?)*)\s*(\/?)\s*>/, + END_TAG_REGEXP = /^<\s*\/\s*([\w:-]+)[^>]*>/, + ATTR_REGEXP = /([\w:-]+)(?:\s*=\s*(?:(?:"((?:[^"])*)")|(?:'((?:[^'])*)')|([^>\s]+)))?/g, + BEGIN_TAG_REGEXP = /^/g, + DOCTYPE_REGEXP = /]*?)>/i, + CDATA_REGEXP = //g, + URI_REGEXP = /^((ftp|https?):\/\/|mailto:|tel:|#)/i, + // Match everything outside of normal chars and " (quote character) + NON_ALPHANUMERIC_REGEXP = /([^\#-~| |!])/g; + + +// Good source of info about elements and attributes +// http://dev.w3.org/html5/spec/Overview.html#semantics +// http://simon.html5.org/html-elements + +// Safe Void Elements - HTML5 +// http://dev.w3.org/html5/spec/Overview.html#void-elements +var voidElements = makeMap("area,br,col,hr,img,wbr"); + +// Elements that you can, intentionally, leave open (and which close themselves) +// http://dev.w3.org/html5/spec/Overview.html#optional-tags +var optionalEndTagBlockElements = makeMap("colgroup,dd,dt,li,p,tbody,td,tfoot,th,thead,tr"), + optionalEndTagInlineElements = makeMap("rp,rt"), + optionalEndTagElements = angular.extend({}, + optionalEndTagInlineElements, + optionalEndTagBlockElements); + +// Safe Block Elements - HTML5 +var blockElements = angular.extend({}, optionalEndTagBlockElements, makeMap("address,article," + + "aside,blockquote,caption,center,del,dir,div,dl,figure,figcaption,footer,h1,h2,h3,h4,h5," + + "h6,header,hgroup,hr,ins,map,menu,nav,ol,pre,script,section,table,ul")); + +// Inline Elements - HTML5 +var inlineElements = angular.extend({}, optionalEndTagInlineElements, makeMap("a,abbr,acronym,b," + + "bdi,bdo,big,br,cite,code,del,dfn,em,font,i,img,ins,kbd,label,map,mark,q,ruby,rp,rt,s," + + "samp,small,span,strike,strong,sub,sup,time,tt,u,var")); + + +// Special Elements (can contain anything) +var specialElements = makeMap("script,style"); + +var validElements = angular.extend({}, + voidElements, + blockElements, + inlineElements, + optionalEndTagElements); + +//Attributes that have href and hence need to be sanitized +var uriAttrs = makeMap("background,cite,href,longdesc,src,usemap"); +var validAttrs = angular.extend({}, uriAttrs, makeMap( + 'abbr,align,alt,axis,bgcolor,border,cellpadding,cellspacing,class,clear,'+ + 'color,cols,colspan,compact,coords,dir,face,headers,height,hreflang,hspace,'+ + 'ismap,lang,language,nohref,nowrap,rel,rev,rows,rowspan,rules,'+ + 'scope,scrolling,shape,span,start,summary,target,title,type,'+ + 'valign,value,vspace,width')); + +function makeMap(str) { + var obj = {}, items = str.split(','), i; + for (i = 0; i < items.length; i++) obj[items[i]] = true; + return obj; +} + + +/** + * @example + * htmlParser(htmlString, { + * start: function(tag, attrs, unary) {}, + * end: function(tag) {}, + * chars: function(text) {}, + * comment: function(text) {} + * }); + * + * @param {string} html string + * @param {object} handler + */ +function htmlParser( html, handler ) { + var index, chars, match, stack = [], last = html; + stack.last = function() { return stack[ stack.length - 1 ]; }; + + while ( html ) { + chars = true; + + // Make sure we're not in a script or style element + if ( !stack.last() || !specialElements[ stack.last() ] ) { + + // Comment + if ( html.indexOf("", index) === index) { + if (handler.comment) handler.comment( html.substring( 4, index ) ); + html = html.substring( index + 3 ); + chars = false; + } + // DOCTYPE + } else if ( DOCTYPE_REGEXP.test(html) ) { + match = html.match( DOCTYPE_REGEXP ); + + if ( match ) { + html = html.replace( match[0] , ''); + chars = false; + } + // end tag + } else if ( BEGING_END_TAGE_REGEXP.test(html) ) { + match = html.match( END_TAG_REGEXP ); + + if ( match ) { + html = html.substring( match[0].length ); + match[0].replace( END_TAG_REGEXP, parseEndTag ); + chars = false; + } + + // start tag + } else if ( BEGIN_TAG_REGEXP.test(html) ) { + match = html.match( START_TAG_REGEXP ); + + if ( match ) { + html = html.substring( match[0].length ); + match[0].replace( START_TAG_REGEXP, parseStartTag ); + chars = false; + } + } + + if ( chars ) { + index = html.indexOf("<"); + + var text = index < 0 ? html : html.substring( 0, index ); + html = index < 0 ? "" : html.substring( index ); + + if (handler.chars) handler.chars( decodeEntities(text) ); + } + + } else { + html = html.replace(new RegExp("(.*)<\\s*\\/\\s*" + stack.last() + "[^>]*>", 'i'), + function(all, text){ + text = text.replace(COMMENT_REGEXP, "$1").replace(CDATA_REGEXP, "$1"); + + if (handler.chars) handler.chars( decodeEntities(text) ); + + return ""; + }); + + parseEndTag( "", stack.last() ); + } + + if ( html == last ) { + throw $sanitizeMinErr('badparse', "The sanitizer was unable to parse the following block " + + "of html: {0}", html); + } + last = html; + } + + // Clean up any remaining tags + parseEndTag(); + + function parseStartTag( tag, tagName, rest, unary ) { + tagName = angular.lowercase(tagName); + if ( blockElements[ tagName ] ) { + while ( stack.last() && inlineElements[ stack.last() ] ) { + parseEndTag( "", stack.last() ); + } + } + + if ( optionalEndTagElements[ tagName ] && stack.last() == tagName ) { + parseEndTag( "", tagName ); + } + + unary = voidElements[ tagName ] || !!unary; + + if ( !unary ) + stack.push( tagName ); + + var attrs = {}; + + rest.replace(ATTR_REGEXP, + function(match, name, doubleQuotedValue, singleQuotedValue, unquotedValue) { + var value = doubleQuotedValue + || singleQuotedValue + || unquotedValue + || ''; + + attrs[name] = decodeEntities(value); + }); + if (handler.start) handler.start( tagName, attrs, unary ); + } + + function parseEndTag( tag, tagName ) { + var pos = 0, i; + tagName = angular.lowercase(tagName); + if ( tagName ) + // Find the closest opened tag of the same type + for ( pos = stack.length - 1; pos >= 0; pos-- ) + if ( stack[ pos ] == tagName ) + break; + + if ( pos >= 0 ) { + // Close all the open elements, up the stack + for ( i = stack.length - 1; i >= pos; i-- ) + if (handler.end) handler.end( stack[ i ] ); + + // Remove the open elements from the stack + stack.length = pos; + } + } +} + +/** + * decodes all entities into regular string + * @param value + * @returns {string} A string with decoded entities. + */ +var hiddenPre=document.createElement("pre"); +function decodeEntities(value) { + hiddenPre.innerHTML=value.replace(//g, '>'); +} + +/** + * create an HTML/XML writer which writes to buffer + * @param {Array} buf use buf.jain('') to get out sanitized html string + * @returns {object} in the form of { + * start: function(tag, attrs, unary) {}, + * end: function(tag) {}, + * chars: function(text) {}, + * comment: function(text) {} + * } + */ +function htmlSanitizeWriter(buf){ + var ignore = false; + var out = angular.bind(buf, buf.push); + return { + start: function(tag, attrs, unary){ + tag = angular.lowercase(tag); + if (!ignore && specialElements[tag]) { + ignore = tag; + } + if (!ignore && validElements[tag] === true) { + out('<'); + out(tag); + angular.forEach(attrs, function(value, key){ + var lkey=angular.lowercase(key); + if (validAttrs[lkey]===true && (uriAttrs[lkey]!==true || value.match(URI_REGEXP))) { + out(' '); + out(key); + out('="'); + out(encodeEntities(value)); + out('"'); + } + }); + out(unary ? '/>' : '>'); + } + }, + end: function(tag){ + tag = angular.lowercase(tag); + if (!ignore && validElements[tag] === true) { + out(''); + } + if (tag == ignore) { + ignore = false; + } + }, + chars: function(chars){ + if (!ignore) { + out(encodeEntities(chars)); + } + } + }; +} + + +// define ngSanitize module and register $sanitize service +angular.module('ngSanitize', []).value('$sanitize', $sanitize); + +/* global htmlSanitizeWriter: false */ + +/** + * @ngdoc filter + * @name ngSanitize.filter:linky + * @function + * + * @description + * Finds links in text input and turns them into html links. Supports http/https/ftp/mailto and + * plain email address links. + * + * Requires the {@link ngSanitize `ngSanitize`} module to be installed. + * + * @param {string} text Input text. + * @param {string} target Window (_blank|_self|_parent|_top) or named frame to open links in. + * @returns {string} Html-linkified text. + * + * @usage + + * + * @example + + + +
+ Snippet: + + + + + + + + + + + + + + + + + + + + + +
FilterSourceRendered
linky filter + 
<div ng-bind-html="snippet | linky">
</div>
+ 		+
+
linky target + 
<div ng-bind-html="snippetWithTarget | linky:'_blank'">
</div>
+ 		+
+
no filter
<div ng-bind="snippet">
</div>
+ + + it('should linkify the snippet with urls', function() { + expect(using('#linky-filter').binding('snippet | linky')). + toBe('Pretty text with some links: ' + + 'http://angularjs.org/, ' + + 'us@somewhere.org, ' + + 'another@somewhere.org, ' + + 'and one more: ftp://127.0.0.1/.'); + }); + + it ('should not linkify snippet without the linky filter', function() { + expect(using('#escaped-html').binding('snippet')). + toBe("Pretty text with some links:\n" + + "http://angularjs.org/,\n" + + "mailto:us@somewhere.org,\n" + + "another@somewhere.org,\n" + + "and one more: ftp://127.0.0.1/."); + }); + + it('should update', function() { + input('snippet').enter('new http://link.'); + expect(using('#linky-filter').binding('snippet | linky')). + toBe('new http://link.'); + expect(using('#escaped-html').binding('snippet')).toBe('new http://link.'); + }); + + it('should work with the target property', function() { + expect(using('#linky-target').binding("snippetWithTarget | linky:'_blank'")). + toBe('http://angularjs.org/'); + }); + + + */ +angular.module('ngSanitize').filter('linky', function() { + var LINKY_URL_REGEXP = + /((ftp|https?):\/\/|(mailto:)?[A-Za-z0-9._%+-]+@)\S*[^\s.;,(){}<>]/, + MAILTO_REGEXP = /^mailto:/; + + return function(text, target) { + if (!text) return text; + var match; + var raw = text; + var html = []; + // TODO(vojta): use $sanitize instead + var writer = htmlSanitizeWriter(html); + var url; + var i; + var properties = {}; + if (angular.isDefined(target)) { + properties.target = target; + } + while ((match = raw.match(LINKY_URL_REGEXP))) { + // We can not end in these as they are sometimes found at the end of the sentence + url = match[0]; + // if we did not match ftp/http/mailto then assume mailto + if (match[2] == match[3]) url = 'mailto:' + url; + i = match.index; + writer.chars(raw.substr(0, i)); + properties.href = url; + writer.start('a', properties); + writer.chars(match[0].replace(MAILTO_REGEXP, '')); + writer.end('a'); + raw = raw.substring(i + match[0].length); + } + writer.chars(raw); + return html.join(''); + }; +}); + + +})(window, window.angular); diff --git a/webclient/room/room-controller.js b/webclient/room/room-controller.js index fa50236571..86f1379f75 100644 --- a/webclient/room/room-controller.js +++ b/webclient/room/room-controller.js @@ -14,7 +14,7 @@ See the License for the specific language governing permissions and limitations under the License. */ -angular.module('RoomController', []) +angular.module('RoomController', ['ngSanitize']) .controller('RoomController', ['$scope', '$http', '$timeout', '$routeParams', '$location', 'matrixService', 'eventStreamService', 'eventHandlerService', function($scope, $http, $timeout, $routeParams, $location, matrixService, eventStreamService, eventHandlerService) { diff --git a/webclient/room/room.html b/webclient/room/room.html index 5712ce9b4f..2726188b4b 100644 --- a/webclient/room/room.html +++ b/webclient/room/room.html @@ -34,8 +34,8 @@
- {{ msg.content.msgtype === "m.emote" ? ("* " + (members[msg.user_id].displayname || msg.user_id) + " " + msg.content.body) : "" }} - {{ msg.content.msgtype === "m.text" ? msg.content.body : "" }} + + {{ msg.content.body }}