[nodejs][jsdom][xml][gdata]search and parse data from youtube by node.js

parsegdata.js

// parsing youtube search gdata with jsdom
// [usage]
//   npm install jsdom
//   node youtubesearch.js > result.xml
//   node parsegdata.js

var fs = require("fs");
var jsdom = require("jsdom");
// magic for applying querySelector/querySelectorAll
jsdom.jsdom("", null, {features: {QuerySelector: true}});

var xml = fs.readFileSync("result.xml", "utf-8"); // as string, not buffer

//[reference] xmlns list for youtube feed
var ns = {
    atom: "http://www.w3.org/2005/Atom",
    media: "http://search.yahoo.com/mrss/",
    openSearch: "http://a9.com/-/spec/opensearchrss/1.0/",
    gd: "http://schemas.google.com/g/2005",
    gml: "http://www.opengis.net/gml",
    yt: "http://gdata.youtube.com/schemas/2007",
    georss: "http://www.georss.org/georss",
};

var browser = jsdom.browserAugmentation(jsdom.dom.level3.core, {});
var doc = new browser.Document();
doc.innerHTML = xml;
//console.log(doc.innerHTML);

// parse feed
//   see: http://code.google.com/apis/youtube/developers_guide_protocol.html
// gdata title
console.log(doc.querySelector("feed>title").textContent);
// get as result entry list
var entries = doc.querySelectorAll("feed>entry");

// pickup 1 entry
var entry = entries[0];
//console.log(entry);
// entry id
console.log(entry.querySelector("id").textContent);
// entry title
console.log(entry.querySelector("title").textContent);
// description
console.log(entry.querySelector("content").textContent);
// date
console.log(new Date(entry.querySelector("published").textContent));
// author name
console.log(entry.querySelector("author>name").textContent);

// html page url for the video
console.log(
    entry.querySelector("link[rel='alternate'][type='text/html']").href);
// gdata url (== id): a root of content xml is just this atom entry 
console.log(entry.querySelector("link[rel='self']").href);


// misc info for the entry
//var group = entry.getElementsByTagNameNS(ns.media, "group")[0]; // jsdom bug?
var group = entry.getElementsByTagName("media:group")[0];
//console.log(group);

// swf url
var swf = group.querySelector(
    "[type='application/x-shockwave-flash'][medium='video']");
console.log(swf.getAttribute("url"));

// thumbnails
var thumbs = group.getElementsByTagName("media:thumbnail");
for (var i = 0; i < thumbs.length; i++) {
    var thumb = thumbs[i];
    var tHeight = thumb.getAttribute("height");
    var tWidth = thumb.getAttribute("width");
    var tTime = thumb.getAttribute("time");
    var tUrl = thumb.getAttribute("url");
    console.log(tUrl + " " + tWidth + "x" + tHeight + " at " + tTime);
}

youtubesearch.js

// accessing youtube search gdata
// this code just prints downloaded search results
// for parsing result, see parsegdata.js
var http = require("http");
var url = require("url");

var keywords = "the beatles";

// see: http://code.google.com/apis/youtube/developers_guide_protocol.html
var searchUri = url.format({
    protocol: "http", hostname: "gdata.youtube.com",
    pathname: "/feeds/api/videos", query: {
        vq: keywords,
        orderby: "published",
    }});
var pSearchUri = url.parse(searchUri);

var req = http.request({
    method: "GET",
    host: pSearchUri.host, path: pSearchUri.path,
}, function (res) {
    var body = "";
    res.on("data", function (chunk) {
        body += chunk;
    });
    res.on("end", function () {
        console.log(body);
    });
});
req.end();

It is an example for using jsdom as a XML DOM processor.
The processing style is similer as browser's DOM API.