Some Notes On Zetterberg's Static Search Widget

readme.md

This is pretty damn cool:

http://www.webtigerteam.com/websearch/websearch.js

I wonder if it would be demonstrably better if case-hardened?

websearch.js

(function() {
	/*
		A search widget for static web sites.
		
		Copyright (C) 2017 Johan Zetterberg
		
		This program is free software; you can redistribute it and/or modify
		it under the terms of the GNU General Public License as published by
		the Free Software Foundation; either version 3 of the License, or
		(at your option) any later version.
		
		This program is distributed in the hope that it will be useful,
		but WITHOUT ANY WARRANTY; without even the implied warranty of
		MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
		GNU General Public License for more details.
		
	*/
	
	"use strict";
	
	//alert("Hello world 5!");
	
	//alert(Object.keys);
	
	// Make sure the script doesn't break for browser's that don't support "console"
	if(!window.console) console = {};
	if(!console.log) console.log = function() {};
	if(!console.warn) console.warn = function() {};
	if(!console.time) console.time = function() {};
	if(!console.timeEnd) console.timeEnd = function() {};
	
	// Log JavaScript errors (note: some browsers doesn't support window.onerror)
	window.onerror = function(message, source, lineno, colno, error) {
		//alert(message + " line=" + lineno);
		var errimg = new Image(), w = window, d = document, e = d.documentElement, g = d.getElementsByTagName('body')[0], x = w.innerWidth || e.clientWidth || g.clientWidth, y = w.innerHeight|| e.clientHeight|| g.clientHeight;
		errimg.src = "/js-error?message=" + escape(message) + "&source=" + escape(source) + "&lineno=" + escape(lineno) + "&colno=" + escape(colno) + "&error=" + escape(error) + "&referer=" + escape(document.referrer) + "&screenRes=" + escape(x + "x" + y) + "&userAgent=" + escape(navigator.userAgent) + "&appVersion=" + escape(navigator.appVersion) + "&platform=" + escape(navigator.platform);
	};
	
	
	window.addEventListener("load", windowLoaded);
	// note: Some (old) browsers doesn't support window.addEventListener!
	
	var websearch_input = document.getElementById("websearch_input");
	
	if(websearch_input) websearch_input.disabled = true;
	
	function windowLoaded() {
		
		//alert("websearch loading ...");
		
		var resultDiv = document.getElementById("websearch_results");
		
		if(!websearch_input) websearch_input = document.getElementById("websearch_input");
		
		
		var websearch_searchButton = document.getElementById("websearch_searchButton"); // optional
		var websearch_clearButton = document.getElementById("websearch_clearButton"); // optional
		
		if(!websearch_input) throw new Error("No input found for websearch widget. Create a input with type=text and id=websearch_input");
		
		var pages = {}; // url: {textStrippedFromHtml, headerTitle}
		var fethingPages = false;
		var textRead = false;
		var searchTimer;
		var gotPages = false;
		
		/*
			Look for RSS links, example:
			<link rel="alternate" type="application/rss+xml" href="../rss_en.xml" title="RSS Feed">
		*/
		var links = document.getElementsByTagName("link");
		var rssFilesFound = 0;
		for(var i=0, type, href; i<links.length; i++) {
			type = links[i].getAttribute("type");
			href = links[i].getAttribute("href");
			if(type && type.match(/rss/i)) {
				rssFilesFound++;
				addRssFile(href);
			}
		}
		if(rssFilesFound==0) throw new Error("No RSS file(s) found!");
		else console.log(rssFilesFound + " RSS files found");
		
		websearch_input.disabled = false;
		
		if(websearch_searchButton) websearch_searchButton.addEventListener("click", function() {
			//alert("Search button clicked");
			doSearch();
		});
		
		
		websearch_input.addEventListener("focus", inputFocus);
		websearch_input.addEventListener("change", inputChange);
		websearch_input.addEventListener("keypress", inputKeyPress);
		
		if(websearch_clearButton) websearch_clearButton.addEventListener("click", cleanUp);
		
		function inputFocus() {
			console.log("inputFocus");
			if(!fethingPages && gotPages) fetchPages();
		}
		
		function inputChange() {
			console.log("inputChange");
			if(websearch_input.value == "") cleanup();
		}
		
		function inputKeyPress(keyPressEvent) {
			console.log("inputKeyPress");
			keyPressEvent = keyPressEvent || window.event;
			var keyCode = keyPressEvent.keyCode || keyPressEvent.which;
			if (keyCode == 13) doSearch()  // Enter pressed
		}
		
		function cleanup() {
			if(resultDiv) while(resultDiv.firstChild) resultDiv.removeChild(resultDiv.firstChild);
		}
		
		function doSearch() {
			console.log("Commence search ...");
			var timeoutTime = 300;
			var searchText = websearch_input.value;
			var pagesCount = Object.keys(pages).length;
			
			if(searchText == "") return console.warn("No text to search!"); 
			
			if(resultDiv) while(resultDiv.firstChild) resultDiv.removeChild(resultDiv.firstChild);
			else {
				resultDiv = document.createElement("div");
				resultDiv.setAttribute("id", "websearch_results"); // So it can be found by getElementById
				resultDiv.setAttribute("class", "websearch_results"); // For styling in CSS
				
				var referenceNode = websearch_searchButton || websearch_input;
				insertAfter(resultDiv, referenceNode);
			}
			
			var loadingText = document.createElement("p");
			loadingText.setAttribute("class", "websearch_loading");
			resultDiv.appendChild(loadingText);
			
			if(!gotPages) {
				console.log("Pages not yet retrieved from RSS ...");
				loadingText.innerText = STR("reading_rss");
			}
			else if(gotPages && pagesCount === 0) {
				throw new Error("No pages to search! Make sure RSS link exist!");
			}
			else if(!fethingPages) {
				console.log("Pages not yet fetched, waiting for them to be fetched ...");
				loadingText.innerText = STR("requesting_pages");
				fetchPages();
				clearTimeout(searchTimer);
				searchTimer = setTimeout(doSearch, timeoutTime);
			}
			else if(fethingPages && !textRead) {
				console.log("Waiting for the pages to fetch ...");
				loadingText.innerText = STR("loading_pages");
				clearTimeout(searchTimer);
				searchTimer = setTimeout(doSearch, timeoutTime);
			}
			else if(fethingPages && textRead) {
				loadingText.innerText = STR("searching_pages");
				search(searchText);
			}
			else throw new Error("fethingPages=" + fethingPages + " textRead=" + textRead + " pages=" + JSON.stringify(pages));
			
		}
		
		function search(searchText) {
			// Text content is hopefully no that big so we can make a full text search instead of a stream search
			var found = false;
			var results = 0;
			var pagesSearched = 0;
			var regString = escapeRegExp(searchText);
			var reg = new RegExp(regString, "ig");
			var maxSearchResults = 10;
			
			while(resultDiv.firstChild) resultDiv.removeChild(resultDiv.firstChild);
			
			for(var url in pages) {
				if(pages[url].text && pages[url].text.search(reg) != -1) {
					found = true;
					results++;
					//console.log("Found searchText=" + searchText + " on page url=" + url);
					showSearchResult(url);
					if(results >= maxSearchResults) break;
				}
				pagesSearched++;
			}
			
			if(found) {
				
				scrollIntoViewIfOutOfView(resultDiv);
				
				// Try to hide the keyboard on mobile devices
				resultDiv.focus(); // This didin't work
				websearch_input.blur(); // This worked!
				
				var ruler = document.createElement("hr");
				resultDiv.appendChild(ruler);
			}
			else {
				var searchFailed = document.createElement("p");
				var failString = STR("search_fail", undefined, {searchText: searchText, pagesSearched: pagesSearched});
				searchFailed.appendChild(document.createTextNode(failString));
				//alert(failString);
				resultDiv.appendChild(searchFailed);
			}
			
			function showSearchResult(url) {
				
				var resultBoxHeading = document.createElement("h4");
				resultBoxHeading.setAttribute("class", "websearch_result_heading");
				
				var title = pages[url].title || getFilenameFromPath(url);
				
				resultBoxHeading.appendChild(document.createTextNode(title));
				
				var resultBoxLink = document.createElement("a");
				resultBoxLink.setAttribute("href", url);
				resultBoxLink.appendChild(resultBoxHeading);
				
				var resultBoxBody = document.createElement("p");
				resultBoxBody.setAttribute("class", "websearch_result_body");
				
				
				var contextStringsFound = 0;
				var maxContextStrings = 3;
				
				var text = pages[url].text;
				var arrContextStrings = [];
				
				var arrExec;
				while((arrExec = reg.exec(text)) !== null  && contextStringsFound < maxContextStrings) addContextString(arrExec);
				
				arrContextStrings = removeRepetition(arrContextStrings);
				
				for (var i=0; i<arrContextStrings.length; i++) {
					arrContextStrings[i] = dotsToString(arrContextStrings[i]);
					
					// Highlight the search string
					arrContextStrings[i] = arrContextStrings[i].replace(new RegExp(regString, "igm"), "<b>" + searchText + "</b>") + "<br>";
				} 
				
				resultBoxBody.innerHTML = arrContextStrings.join("");
				
				
				// Reset reg
				reg.lastIndex = 0;
				
				var resultBox = document.createElement("div");
				resultBox.setAttribute("class", "websearch_resultbox");
				
				resultBox.appendChild(resultBoxLink);
				resultBox.appendChild(resultBoxBody);
				
				resultDiv.appendChild(resultBox);
				
				function addContextString(arrExec) {
					contextStringsFound++;
					
					var contextLength = 100;
					var contextStart = Math.max(0, arrExec.index-Math.round(contextLength/2) - searchText.length);
					var contextEnd = Math.min(text.length, contextStart+contextLength);
					
					console.log("contextStart=" + contextStart + " contextEnd=" + contextEnd + " url=" + url);
					
					// End before a line break if possible
					var lineBreakAfter = text.indexOf("\n", arrExec.index);
					if(lineBreakAfter < arrExec.index - contextLength/2) lineBreakAfter = text.indexOf("\n", arrExec.index + contextLength/4);
					
					if(lineBreakAfter != -1 && lineBreakAfter-arrExec.index < contextLength/1.5) {
						contextEnd = lineBreakAfter;
						console.log("Corrected contextEnd=" + contextEnd + " because of ending line break");
					}
					else {
						// End after a dot (end of sentence) if possible
						var lastDot = text.indexOf(".", arrExec.index);
						if(lastDot != -1 && lastDot-arrExec.index < contextLength/1.5) {
							contextStart = firstDot;
							console.log("Corrected contextStart=" + contextStart + " because of lastDot");
						}
						else {
							// End before a space
							var lastSpace = text.lastIndexOf(" ", contextEnd);
							if(lastSpace != -1 && (contextEnd - lastSpace) < contextLength/4) {
								contextEnd = lastSpace;
								console.log("Corrected contextEnd=" + contextEnd + " because of last space");
							}
						}
					}
					
					if((contextEnd - contextStart) < contextLength && contextStart > 0) {
						contextStart = Math.max(0, contextEnd - contextLength);
						console.log("Corrected contextStart=" + contextStart + " because not enough length");
					}
					
					// Start after a line break if possible
					var lineBreak = text.lastIndexOf("\n", arrExec.index);
					if(lineBreak != -1 && (lineBreak - arrExec.index) < contextLength/1.5) {
						contextStart = lineBreak;
						console.log("Corrected contextStart=" + contextStart + " because of starting line break");
					}
					else {
						// Start after a dot (end of sentence) if possible
						var firstDot = text.lastIndexOf(".", arrExec.index);
						if(firstDot != -1 && arrExec.index-firstDot < contextLength/1.5) {
							contextStart = firstDot;
							console.log("Corrected contextStart=" + contextStart + " because of firstDot");
						}
						else {
							// Start after first space
							var firstSpace = text.indexOf(" ", contextStart);
							if(firstSpace != -1 && (contextStart - firstSpace) < contextLength/4) {
								contextStart = firstSpace + 1;
								contextEnd = Math.min(text.length, contextStart+contextLength);
								console.log("Corrected contextStart=" + contextStart + " contextEnd=" + contextEnd + " because of first space");
							}
						}
					}
					
					// Sanity check
					if(!(arrExec.index >= contextStart && arrExec.index < contextEnd)) {
						throw new Error("Something went wrong, search string is not in range!\n" +
						"contextStart=" + contextStart + " contextEnd=" + contextEnd + " arrExec.index=" + arrExec.index);
					}
					
					var contextString = text.substring(contextStart, contextEnd);
					
					// Update reg so that we don't show the same string again
					reg.lastIndex = contextEnd;
					
					arrContextStrings.push(contextString);
					
					//console.log("contextString=" + contextString);
					
				}
				
				function dotsToString(contextString) {
					var firstChar = contextString.charAt(0);
					contextString = contextString.trim(); // Trim after getting firstChar so that firstChar can be a line break of tab
					var lastChar = contextString.charAt(contextString.length-1);
					
					//console.log("firstChar=" + firstChar + " lastChar=" + lastChar);
					
					if(!(lastChar == "." || lastChar == ";" || lastChar == "?" || lastChar == "!")) contextString += " ...";
					
					if(!(firstChar == "\n" || firstChar == "\t" || firstChar == ".")) contextString = "... " + contextString;
					
					return contextString;
				}
				
				function removeRepetition(arr) {
					//console.log("Remove repetition: arr=" + JSON.stringify(arr, null, 2));
					// Recursive, removes strings the repeats
					for(var i=0; i<arr.length; i++) {
						for(var j=0; j<arr.length; j++) {
							if(arr[i].indexOf(arr[j].trim()) != -1 && j != i) {
								//console.log("arr[" + i + "]=" + arr[i] + "\ncontains arr[" + j + "]=" + arr[j]);
								arr.splice(j, 1);
								return removeRepetition(arr);
							}
						}
					}
					return arr;
				}
			}
			
		}
		
		
		// Some (old) browsers doesn't like this function!! Why?
		function fetchPages() {
			//alert("Fetching pages ...");
			
			if(!gotPages) throw new Error("Have not yet got the pages from RSS file(s)");
			
			console.time("Fetch pages for searching");
			fethingPages = true;
			var pagesCount = Object.keys(pages).length;
			var pagesFetched = 0;
			var pagesFailed = 0;
			var retries = {};
			
			if(pagesCount == 0) console.warn("No pages to fetch!");
			else {
				for(var url in pages) {
					retries[url] = 0;
					fetch(url, pageFetched);
				}
			}
			
			function pageFetched(err, url) {
				
				if(err) {
					console.warn("Failed to fetch url=" + url + " code=" + err.code + ": " + err.message);
				}
				else {
					console.log("Successfully fetched url=" + url);
				}
				
				pagesFetched++;
				
				//alert("Fetched page " + pagesFetched + " of " + pagesCount + " url=" + pageUrl);
				if(pagesFetched == pagesCount) {
					textRead = true;
					console.timeEnd("Fetch pages for searching");
					//alert("All pages fetched!");
				}
			}
			
			function fetch(pageUrl, whenDoneFetching) {
				
				if(typeof whenDoneFetching != "function") throw new Error("whenDoneFetching=" + whenDoneFetching);
				
				httpGet(pageUrl, function pageFetched(err, text) {
					
					if(err) return whenDoneFetching(err, pageUrl);
					
					var page = pages[pageUrl];
					
					page.text = text.replace(/\r\n/g, "\n"); // Use LF as line break character
					page.text = stripHtmlComments(page.text);
					page.text = contentOfHtmlTag(page.text, "body");
					page.text = stripHtml(page.text); // We are not interested in the HTML tags
					
					var arrTitle = find(text, "<title>(.*?)</title>\\s{0,}", 1);
					if(arrTitle.length > 0) page.title = arrTitle[0];
					
					return whenDoneFetching(null, pageUrl);
					
				});
			}
		}
		
		function addRssFile(href) {
			if(href.indexOf("://") != -1) var url = href;
			else var url = absolutePath(document.location.href, href)
			
			console.log("RSS url=" + url);
			
			httpGet(url, function gotRssFile(err, text) {
				if(err) throw err;
				
				var items = text.match(/<guid>([^<]+)<\/guid>/ig);
				
				if(!items) console.warn("Unable to find <guid> in RSS url=" + url + "\ntext=" + text);
				else {
					for(var i=0, pageUrl; i<items.length; i++) {
						pageUrl = items[i].replace(/<\/?guid>/ig,'');
						pages[pageUrl] = {};
						//console.log("Found pageUrl=" + pageUrl);
					}
				}
				
				gotPages = true;
			});
		}
		
		function fetchRssFile() {
			
		}
		
	}
	
	function absolutePath(base, relative) {
		// https://stackoverflow.com/questions/14780350/convert-relative-path-to-absolute-using-javascript
		var stack = base.split("/"),
		parts = relative.split("/");
		stack.pop(); // remove current file name (or empty string)
		// (omit if "base" is the current folder without trailing slash)
		for (var i=0; i<parts.length; i++) {
			if (parts[i] == ".")
			continue;
			if (parts[i] == "..")
			stack.pop();
			else
			stack.push(parts[i]);
		}
		return stack.join("/");
	}
	
	function insertAfter(newNode, referenceNode) {
		// https://stackoverflow.com/questions/4793604/how-to-do-insert-after-in-javascript-without-using-a-library
		referenceNode.parentNode.insertBefore(newNode, referenceNode.nextSibling);
	}
	
	function find(text, reString, group) {
		
		var re = new RegExp(reString, "img");
		
		if(group == undefined) group = 1;
		
		var arr = [];
		var result = [];
		
		while ((arr = re.exec(text)) !== null) {
			//log("Found: " + JSON.stringify(arr));
			result.push(arr[group]);
		}
		
		if(result.length == 0) {
			//log("Did not find " + reString + "");
			//result.push("");
		}
		
		return result;
	}
	
	function contentOfHtmlTag(text, tag, nr) {
		
		//log("Finding content of tag: " + tag);
		
		if(nr == undefined) nr = 1;
		
		var originalText = text;
		
		// Ignore the case
		text = text.toLowerCase();
		tag = tag.toLowerCase();
		
		
		
		var xmpStart = -1;
		var xmpEnd = -1;
		var tagStart = -1;
		var condition = true;
		var xmpTags = [];
		var insideXmp = false;
		var count = 0;
		
		// Find all xmp tags
		do {
			xmpStart = text.indexOf("<xmp", xmpEnd);
			xmpEnd = text.indexOf("</xmp", xmpStart);
			
			if(xmpStart != -1) xmpTags.push({start: xmpStart, end: xmpEnd});
			
		} while(xmpStart != -1)
		
		//log("xmpTags=" + JSON.stringify(xmpTags));
		
		// Find tagStart, nr, not inside xml tags
		do {
			
			tagStart = text.indexOf("<" + tag, tagStart+1);
			
			//log("tagStart=" + tagStart);
			
			insideXmp = false;
			
			for(var i=0; i<xmpTags.length; i++) {
				insideXmp = xmpTags[i].start < tagStart && xmpTags[i].end > tagStart;
				if(insideXmp) break;
			}
			
			if(!insideXmp) count++;
			
		} while (count < nr && tagStart != -1);
		
		if(tagStart == -1) {
			//log("Did not find enough occurencies of " + tag + "");
			return "";
		}
		
		
		// Move to the end of the tagStart
		tagStart = text.indexOf(">", tagStart) + 1;
		
		
		// Find tag end (cant be inside xmp)
		var tagEnd = tagStart;
		//log("Finding end ...");
		do {
			
			tagEnd = text.indexOf("</" + tag, tagEnd+1);
			
			//log("tagEnd=" + tagEnd);
			
			insideXmp = false;
			
			for(var i=0; i<xmpTags.length; i++) {
				insideXmp = xmpTags[i].start < tagEnd && xmpTags[i].end > tagEnd;
				if(insideXmp) {
					tagEnd = xmpTags[i].end; // Start looking from here
					break;
				}
			}
			
		} while (insideXmp && tagEnd != -1);
		
		if(tagEnd == -1) error(new Error("Could not find tag ending </" + tag + ""));
		
		// Return the content of the html tag
		return originalText.substring(tagStart, tagEnd);
		
	}
	
	function getFilenameFromPath(path) {
		if(path.indexOf("/") > -1) {
			return path.substr(path.lastIndexOf('/')+1);
		}
		else {
			// Assume \ is the folder separator
			return path.substr(path.lastIndexOf('\\')+1);
		}
	}
	
	function stripHtml(text) {
		return text.replace(/<(?:.|\n)*?>/gm, '');
	}
	
	function stripHtmlComments(text) {
		text = text.replace(/<!--(?!>)[\S\s]*?-->/g, '');
		
		if(text.indexOf("<!--") != -1) console.warn("Strip comments might have failed! text contains <!--");
		if(text.indexOf("-->") != -1) console.warn("Strip comments might have failed! text contains -->");
		
		return text;
	}
	
	function stripHtmlCommentsOld(text) {
		// https://stackoverflow.com/questions/5653207/remove-html-comments-with-regex-in-javascript
		var COMMENT_PSEUDO_COMMENT_OR_LT_BANG = new RegExp(
		'<!--[\\s\\S]*?(?:-->)?'
		+ '<!---+>?'  // A comment with no body
		+ '|<!(?![dD][oO][cC][tT][yY][pP][eE]|\\[CDATA\\[)[^>]*>?'
		+ '|<[?][^>]*>?',  // A pseudo-comment
		'g');
		
		text = text.replace(COMMENT_PSEUDO_COMMENT_OR_LT_BANG, "");
		
		if(text.indexOf("<!--") != -1) console.warn("Strip comments might have failed! text contains <!--");
		if(text.indexOf("-->") != -1) console.warn("Strip comments might have failed! text contains -->");
		
		return text;
	}
	
	function escapeRegExp(str) {
		return str.replace(/[-[\]{}()*+?.,\\^$|#\s]/g, "\\$&");
	}
	
	function scrollIntoViewIfOutOfView(el) {
		if(!el.scrollIntoView) return console.warn("scrollIntoView not supported for el=" + el);
		
		var topOfPage = window.pageYOffset || document.documentElement.scrollTop || document.body.scrollTop;
		var heightOfPage = window.innerHeight || document.documentElement.clientHeight || document.body.clientHeight;
		var elY = 0;
		var elH = 0;
		if (document.layers) { // NS4
			elY = el.y;
			elH = el.height;
		}
		else {
			for(var p=el; p&&p.tagName!='BODY'; p=p.offsetParent){
				elY += p.offsetTop;
			}
			elH = el.offsetHeight;
		}
		if ((topOfPage + heightOfPage) < (elY + elH)) {
			el.scrollIntoView(false);
		}
		else if (elY < topOfPage) {
			el.scrollIntoView(true);
		}
	}
	
	function httpGet(url, callback, range) {
		var xmlHttp = new XMLHttpRequest();
		var timeoutTimer;
		var timeoutTimeMs = 10000;
		
		//console.log("Get: url=" + url + " range=" + JSON.stringify(range));
		
		xmlHttp.onreadystatechange = function httpReadyStateChange() {
			if(xmlHttp.readyState == 4) {
				clearTimeout(timeoutTimer);
				//console.log("Get Status: url=" + url + " xmlHttp.status=" + xmlHttp.status + "");
				//console.log("Headers:\n" + xmlHttp.getAllResponseHeaders());
				if(xmlHttp.status == 200 || xmlHttp.status == 206) {
					
					var err = null;
					
					if(range && xmlHttp.status != 206) err = new Error("Made a range=" + JSON.stringify(range) + " request but got a HTTP status=" + xmlHttp.status + " (expected 206)");
					
					// Nginx seems to give size in characters
					//var size = (xmlHttp.status == 206) ? xmlHttp.getResponseHeader("Size") : xmlHttp.responseText.length;
					var size = xmlHttp.responseText.length;
					
					if(range && range.end) {
						var rangeLength = ( range.end-range.start + 1 );
						if(size != rangeLength) throw new Error("size=" + size + " and rangeLength=" + rangeLength + " does not match! range=" + JSON.stringify(range) + "");
					}
					
					callback(err, xmlHttp.responseText, size);
				}
				else {
					var err = new Error(xmlHttp.responseText + " xmlHttp.status=" + xmlHttp.status + " xmlHttp.readyState=" + xmlHttp.readyState);
					err.code = xmlHttp.status;
					callback(err);
				}
			}
			//else console.log("xmlHttp.readyState=" + xmlHttp.readyState);
		}
		
		// setRequestHeader must be after open, or we'll get a: Uncaught DOMException: Failed to execute 'setRequestHeader' on 'XMLHttpRequest': The object's state must be OPENED.
		xmlHttp.open("GET", url, true); // true for asynchronous
		if(range) xmlHttp.setRequestHeader("Range", "bytes=" + range.start + "-" + (range.end ? range.end : ""));
		xmlHttp.send(null);
		
		timeoutTimer = setTimeout(timeout, timeoutTimeMs);
		
		function timeout() {
			var err = new Error("HTTP GET request timed out. xmlHttp.readyState=" + xmlHttp.readyState);
			err.code = "TIMEOUT";
			xmlHttp.onreadystatechange = null;
			xmlHttp.abort();
			callback(err);
		}
	}
	
	function STR(stringId, lang, vars) {
		
		if(lang == undefined) lang = navigator.language;
		
		var defaultLang = "en";
		
		if(!translations.hasOwnProperty(stringId)) {
			throw new Error("No translation available for stringId=" + stringId);
		}
		else {
			if(!translations[stringId].hasOwnProperty(lang)) {
				console.warn("No translation for stringId=" + stringId + " available for lang=" + lang);
				if(lang != defaultLang) return STR(stringId, defaultLang, vars);
				else throw new Error("No translation for stringId=" + stringId + " with lang=" + lang + " and defaultLang=" + defaultLang);
			}
			else {
				
				var string = translations[stringId][lang];
				
				if(vars) {
					for(var name in vars) {
						if(string.indexOf("%" + name + "%") != -1) {
							string = string.replace("%" + name + "%", vars[name]);
						}
						else throw new Error("variable name=" + name + " does not exist in stringId=" + stringId);
					}
					
					// Check if we forgot a variable. PS. string variables can not have spaces!
					var matches = string.match(/%\w+%/);
					if(matches != null) {
						for(var i=0; i<matches.length; i++) {
							throw new Error("Forgot variable for " + matches[i] + " in " + stringId);
						}
					}
				}
				
				return string;
			}
		}
		
	}
	
	var translations = {
		reading_rss: {
			en: "Reading RRS",
			sv: "Läser in RSS"
		},
		requesting_pages: {
			en: "Requesting pages for searching",
			sv: "Hämtar sidor inför sökning"
		},
		loading_pages: {
			en: "Loading pages for searching",
			sv: "Laddar sidor inför sökning"
		},
		searching_pages: {
			en: "Searching the pages",
			sv: "Söker på sidorna"
		},
		search_fail: {
			en: 'Did not find the string "%searchText%" in any of the %pagesSearched% pages!',
			sv: 'Kunde inte hitta "%searchText%" i någon av de %pagesSearched% sidorna!'
		}
	};
	
	
})();