XML parsing in ECMAScript¶

When using ECMAScript and Duktape the plugins need to rely on Movian's XML parser. To access do:

var XML = require('showtime/xml');

To parse a string as XML then do as this tiny example:

var doc = XML.parse('<foo><bar baz="123">meh</bar></foo>');

Accessing nodes in the element is easy, just dereference properties

console.log(doc.foo.bar);           // prints 'meh'

Attributes are prepared with an 'at' (@) sign an thus needs to be accessed using the array operator:

console.log(doc.foo.bar["@baz"]);   // prints '123'

Let's step it up by creating a more complicated document

var doc2 = XML.parse('<result>' +
                    '<alpha>1</alpha>' +
                    '<beta>2</beta>' +
                    '<gamma>3</gamma>' +
                    '<delta>4</delta>' +
                    '</result>');

Nodes can be iterating using the standard ECMAScript enumeration technique:

for(v in doc2.result) {
 print(v + " = " + doc2.result[v]);
}

will print

alpha = 1
beta = 2
gamma = 3
delta = 4

Iterating over nodes which have the same name is possible using the filterNodes() method, see the following example:

var doc3 = XML.parse('<dataset>' +
                    '<data>foo</data>' +
                    '<data>bar</data>' +
                    '<data>baz</data>' +
                    '<data>quux</data>' +
                    '<metadata>quuz</metadata>' +
                    '</dataset>');

var all_the_data = doc3.dataset.filterNodes('data');
for(var i = 0; i < all_the_data.length; i++) {
 var data = all_the_data[i];
 print(i + ". " + data);
}

Will print

0. foo
1. bar
2. baz
3. quux

No surprises really.

For some extra info and reference, here is a diff of the change needed to convert Movian's Headweb plugin from JavaScript (SpiderMonkey) to ECMAScript (Duktape)

https://github.com/andoma/showtime-plugin-headweb/commit/dfa43609fde79a65834e3fca7f7ca696fb7f8805