MediaWiki:ExtractFirst.xsl
Jump to navigation
Jump to search
<?xml version="1.0" encoding="UTF-8"?>
<!-- This is for extracting the first definition of a word from wiktionary, that can be used in a cross site manner. Consider:
http://en.wiktionary.org/w/api.php?action=parse&prop=text&page=word&format=xml&xslt=MediaWiki:extractFirst.xsl -->
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output method='html'/>
<!-- for translation. Also see JS -->
<xsl:variable name="dir">ltr</xsl:variable>
<xsl:variable name="more">» More</xsl:variable>
<xsl:variable name="error">Error: </xsl:variable>
<xsl:variable name='audio'>Pronunciation info</xsl:variable>
<xsl:variable name='audio-separator'> • </xsl:variable>
<xsl:variable name="copyright"> © <a href="http://en.wiktionary.org/wiki/">Wiktionary</a>. Released under <a href="http://creativecommons.org/licenses/by-sa/3.0/" rel="license copyright">CC-BY-SA 3.0</a></xsl:variable>
<xsl:variable name="contentLang" select="'en'"/> <!-- make sure quoted-->
<!-- END XSLT VARIABLES TO TRANSLATE. SEE JS as well -->
<!-- other variables. DO NOT TRANSLATE THESE -->
<xsl:variable name="pageName" select="concat('http://', $contentLang, '.wiktionary.org')"/>
<!-- end xslt constants -->
<xsl:template match="/">
<html dir="{$dir}" lang="{$contentLang}" xml:lang="{$contentLang}">
<head>
<meta name="generator" content="Wiktionary Extract XSLT 1.10-EN"/>
<base target='_blank' href="{$pageName}" />
<title> Wiktionary extract</title>
<style>
#wordThisIsFor { font-weight:bold;}
a.wtif1 { color: black; text-decoration: none;}
a.wtif1:hover {text-decoration: underline;}
.disambig-see-also, .disambig-see-also-2 {display:inline;}
#container {background-color:white; padding: 0.5em; border: solid black thin;}
a.new {color: red;}
#error {color: red;font-size:larger;}
div.exit {float:right;font-weight:bold;font-family:sans-serif;border:outset;padding:0.1em}
div.exit a {color: black; text-decoration:none;}
</style>
<script type='text/javascript'>
/*<![CDATA[*/
function setup () {
//Stuff to translate:
var preferLang = {'en': 'English', 'fr': 'French', 'de': 'German', 'es': 'Spanish', 'it': 'Italian', 'pt': 'Portuguese', 'ja': 'Japanese', 'pl': 'Polish', 'ru': 'Russian', 'nl': 'Dutch', 'qqqAny': null}; //for now.
var extractSeeAlso = /<div class=\"disambig-see-also(?:-2)?\">[\s\S]*?<\/div>/; //no subexpressions!
var see_also_process = function (sa) {return sa;}
var createLink = '» Create'; // text only.
var not_found = "Could not retrieve definition of $1.";
//END stuff to translate (there is one more translation block below)
//Stuff not to translate in general (setup).
var rd = location.search.match(/\&rd\=([^&]*)/); //is this from redirect. + converts to numeric.
rd = rd ? (+rd[1] + 1) : 1; //redirection level.
var showWord = 0; //default to not showing. 0 = none, 1 = bold, 2 = bold link.
var showWordRaw = location.search.match(/\&showWord\=([^&]*)/);
showWordRaw = showWordRaw ? showWordRaw[1] : 'none';
if (showWordRaw !== "none") {
showWord++;
}
if (showWordRaw === "link") {
showWord++;
}
var useAudio = 0;
if (location.search.match(/\&audio\=(?!none)/)) useAudio++;
if (location.search.match(/\&audio\=autoplay/)) useAudio++;
var closeLink = false;
if (location.search.match(/&exit\=(?!false)/)) closeLink = true;
var numbDfn = location.search.match(/\&count\=([^&]*)/); //count. + converts to numeric.
numbDfn = numbDfn ? (parseInt(numbDfn[1])) : 1; //default to 1
var pageURL = '/w/index.php?title=' + encodeURIComponent(decodeURIComponent(location.search.match(/\&page\=([^&]*)/)[1]));
var src = document.getElementById('src');
var display = document.getElementById('word-list');
var loc = location.search.match(/\&page\=([^&]*)/)[1]; //this is not escaped
var escWord = decodeURIComponent(loc).replace(/&/, '&').replace(/>/, '<').replace(/</, '>');
//note: wordEsc does not escape quotes. DO NOT PUT AS ATTRIBUTE VALUE
var preferLangCode = location.search.match(/\&lang\=([^&]*)/);
if (preferLangCode) { preferLangCode = preferLangCode[1]; }
else {preferLangCode = 'qqqAny';}
src.normalize();
var html = src.firstChild.data;
var def = html //may be redefined later.
//stuff you might need to translate, but hopefully won't
var subSectRegex = new RegExp('<h2>[^<]*?(<span[^<]*?<a[^<]*?<\/a[^<]*?</span[^<]*?)?<span class="mw-headline" id="' + preferLang[preferLangCode] + '"[^>]*>[\\s\\S]*$');
var extractCurLangName = /<span class="mw-headline" id[^>]*>([\s\S]*?)<\/span>/; //first subexpression
//End stuff you hopefully won't need to translate.
try {
//this assumes attribute order doesn't change!!!
html = html.replace(/<div id="toctitle">[\s\S]*?<\/div>/, '');
if (preferLangCode && preferLang[preferLangCode]) {
try {
//strip off all definitions before tagret lang.
var subSect = html.match(subSectRegex)[0];
if (subSect.match(/<ol>[\s\S]*?<li>/)) {
//if it has content
def = subSect;
}
} catch (e) { /*alert(e)*/}
}
var lang = def.match(extractCurLangName)[1];
var intro = "(" + lang + ") ";
//Start testing to see if we can play audio.
var audioCreditLink;
var audioLink = html.match(/\<a\s*\S*?\s*href\=\"?(http:\/\/upload.wikimedia.org\/\S*.(?:oga|ogg))\"?[^>]*>/);
if (!audioLink) audioLink = html.match(/\<button[\s\S]*?\sonclick=\"[\s\S]*?"videoUrl":\s*"(http:\/\/upload.wikimedia.org\/\S*?)"[^"]*?"isVideo":\sfalse[^"]*?".*?>/);
if (audioLink) {
audioCreditLink = audioLink[1].match(/\/([^\/]*)$/);
if (audioCreditLink) audioCreditLink = 'File:' + audioCreditLink[1]
if (useAudio > 0) {
var supportsHTML5Audio = (typeof Audio !== 'undefined');
var audioAutoplay = (useAudio === 2) ? "autoplay" : "";
audioLink = encodeURI(decodeURI(audioLink[1])); //should be already encoded, but better safe.
intro += ' <audio style="float:right;" controls src="' + audioLink + '" ' + audioAutoplay; + ' >';
//the cortando stuff is inspired (stolen) from oggHandler.
//the object tag interferes with the audio tag on firefox for autoplay
//specificly it interperts autoplay differently then mplayer does.
if (!supportsHTML5Audio) {
intro += '<object align="right" type="application/ogg" height="18" width="180" data="' + audioLink + '" autoplay="' + (audioAutoplay === 'autoplay') + '" >';
}
intro += '<applet code="com.fluendo.player.Cortado.class" ' +
' width=180' +
' height=20' +
' archive="http://upload.wikimedia.org/jars/cortado.jar">' +
' <param name="url" value="' + audioLink + '" />' +
' <param name="seekable" value="true"/>' +
' <param name="autoPlay" value="' + (audioAutoplay === 'autoplay') + '"/>' +
' <param name="showStatus" value="show"/>' +
' <param name="showSpeaker" value="false"/>' +
' <param name="statusHeight" value="18"/>' +
'<p style="display:none"></p></applet>';
if (!supportsHTML5Audio) {
intro += '</object>';
}
intro += '</audio>';
}
}
if (showWord) intro = '<a href="' + pageURL + '" id="wordThisIsFor" class="wtif' + showWord + '" >' + escWord + "</a> " + intro ;
var definitions_matched;
//FIXME: in both cases the extraction method does not properly strip nested divs. This results in image thumbnails being left behind
if (numbDfn === 1) {
definitions_matched = def.match(/<ol>[\s\S]*?<\/ol>/)[0].replace(/<dl>[\s\S]*?<\/dl>/g, '').replace(/<div[^>]*>[\s\S]*?<\/div>/g, '').replace(/<\/div>/g, '').replace(/<ul>[\s\S]*?<\/ul>/g, '').replace(/<a href="(#[^"]*)">/g, '<a href="' + pageURL + '$1">').match(/<li>([\s\S]*?)<\/li>/);
display.innerHTML = intro + definitions_matched[1];
} else {
//this use not well supported...
definitions_matched = def.match(/<ol>[\s\S]*?<\/ol>/)[0].replace(/<dl>[\s\S]*?<\/dl>/g, '').replace(/<div[^>]*>[\s\S]*?<\/div>/g, '').replace(/<\/div>/g, '').replace(/<ul>[\s\S]*?<\/ul>/g, '').replace(/<a href="(#[^"]*)">/g, '<a href="' + pageURL + '$1">').match(/<li>([\s\S]*?)<\/li>/g);
var tmp = intro + ' <ul>';
for (var i = 0; i < numbDfn && i < definitions_matched.length; i++) {
tmp += definitions_matched[i];
}
display.innerHTML = tmp + '</ul>';
}
}
catch (e) {
//alert(e)
//page does not exist, not well formed, these regexs suck, etc
//note, this is appending a text node, thus it is ok, that loc is not escaped.
display.appendChild(document.createTextNode(not_found.replace("$1", decodeURIComponent(loc))));
document.getElementById('more-link').firstChild.data = createLink;
if (rd < 9) { //arbitrary to prevent infinite loops
//make sure don't have loops.
var newLoc; //this should not be urlEncoded.
var remAlt = false;
var dLoc = decodeURIComponent(loc);
newLoc = dLoc.charAt(0).toLowerCase() + dLoc.substring(1, loc.length);
//try some other redirections.
if (newLoc === dLoc) newLoc = dLoc.toLowerCase();
if (newLoc === dLoc && location.search.match(/\&alt\=([^&]*)/)) {
newLoc = decodeURIComponent(location.search.match(/\&alt\=([^&]*)/)[1]);
remAlt = true;
}
if (newLoc !== dLoc) { //redir
var newURL = location.href.replace(/(^[\s\S]*?\&page\=)[^&]*([\s\S]*$)/, '$1'+ encodeURIComponent(newLoc) + '$2');
newURL = newURL.replace(/&rd\=[^&]*/, ''); //strip old redirect header.
if (remAlt) {
location.href.replace(/&alt\=[^&]*/, '');
}
location = newURL + '&rd=' + rd;
}
}
}
var sa = html.match(extractSeeAlso);
if(sa) {
document.getElementById('see-also').innerHTML = ' (' + see_also_process(sa) + ')' ;
}
if (closeLink) {
var container = document.getElementById('container');
container.innerHTML = '<div class="exit"><a href="about:blank" title="Hide" target="_self">X</a></div>' + container.innerHTML;
}
document.getElementById('more-link').href= pageURL;
if (audioCreditLink) {
document.getElementById('audio-separator').style.display = 'inline';
var audioA = document.getElementById('audio-link');
audioA.href = '/w/index.php?title=' + encodeURIComponent(decodeURIComponent(audioCreditLink));
audioA.style.display = 'inline';
}
}
/*]]>*/
</script>
</head>
<body onload='setup()'>
<div id='container'>
<div id='word-list'><xsl:apply-templates select='api/error'/></div>
<div><a id='more-link'><xsl:value-of select="$more"/> </a> <span id='audio-separator' style='display:none;'><xsl:value-of select="$audio-separator"/></span> <a id='audio-link' style='display:none;'> <xsl:value-of select="$audio"/></a> <span id='see-also'></span> <small id="copyright-notice"> <xsl:copy-of select="$copyright"/></small>
</div>
</div>
<div id='src' style='display:None'>
<xsl:value-of select='api/parse/text'/>
</div>
</body>
</html>
</xsl:template>
<xsl:template match='api/error'>
<span id='error'><b><xsl:value-of select="$error"/></b> <xsl:value-of select='@info'/></span>
</xsl:template>
</xsl:stylesheet>