JDistiller, A DSL for web-crawling built on Node.js and jQuery

A couple weeks ago I posted a web-crawler I'd built for parsing Wikipedia articles (<a href="http://news.ycombinator.com/item?id=4436349" rel="nofollow">http://news.ycombinator.com/item?id=4436349</a>). jDistiller grew out of this project. As a test I rebuilt the crawler in thirty lines of code using the jDistiller API:<p>new jDistiller() .set('title', '#firstHeading') .set('links', '#bodyContent p a', function(element, prev) { var href = element.attr('href'), key = href.replace('/wiki/', '');<p><pre><code> if ( key === href) return; element.replaceWith('[[' + key + ']]'); return [key, { title: element.attr('title'), text: prev[key] ? prev[key].text : element.text(), occurrences: prev[key] ? prev[key].occurrences + 1 : 1 }] }) .set('sections', '#bodyContent p,#firstHeading,h2,h3,img', function(element, prev) { var images = []; if ( element.is('h1') || element.is('h2') ) { this.heading = element.text().trim(); return } return [this.heading, { text: prev[this.heading] ? prev[this.heading].text + element.text() : element.text(), images: element.is('img') &#38;&#38; element.attr('width') &#62; 50 ? [element.attr('src').replace('//', 'http://')] : [] }]; }) .distill('http://en.wikipedia.org/wiki/Dog', function(err, distilledPage) { console.log(JSON.stringify(distilledPage.sections)); });</code></pre>