.AS File Format – XML and E4X

XML and XMLList

• XML – one node: an element, an attribute, a text node, a comment or a processing instruction. nodeKind() says which: "element", "attribute", "text", "comment" or "processing-instruction".
• XMLList – an ordered list of nodes. Most E4X expressions give a list. A list of one item behaves like that item, so book.title can be read as text even though it is a list.

Creating XML

XML can be written straight into the code, without quotes. Braces insert the value of an expression as an element name, an attribute value or content. A string can also be parsed with new XML(text) or XML(text), which throws a TypeError for null, undefined or markup that is not well formed.

var price:Number = 4.5;
var shop:XML =
    <shop name="Corner">
        <item id="1" type="fruit">
            <name>Apple</name>
            <price>0.4</price>
        </item>
        <item id="2" type="bread">
            <name>Loaf</name>
            <price>{price}</price>
        </item>
    </shop>;

var tag:String = "note";
var extra:XML = <{tag} lang="en">Closed on Sunday</{tag}>;

var list:XMLList = <><a/><b/></>;
var parsed:XML = new XML('<point x="3" y="4"/>');

Static settings on XML control parsing and printing. They apply to every XML object created afterwards.

XML.settings(), XML.setSettings(obj) and XML.defaultSettings() save and restore all five at once.

Reading a tree

Inside a filter ( … ), names refer to each item’s children and attributes; other variables in scope can be used too. Element text is a string, so numbers are compared after Number() or int().

An element that is missing gives an empty list, not an error: shop.owner.length() is 0. Testing shop.owner == undefined or shop.owner.length() == 0 checks for it.

Getting text out

• toString() – for a node with simple content (only text), the text without tags; otherwise the markup. String conversion, trace and + "" use it.
• toXMLString() – always the markup, tags included.
• text() – the text nodes, as a list.
• Attributes and element text are always strings: int(shop.item[1].@id) gives a number.

Changing a tree

shop.item[0].price = 0.45;                 // replace an element's text
shop.item[0].@sale = "yes";                // add or replace an attribute
shop.@opened = 1998;                       // values become strings
shop.item[1].name += " (white)";
shop.appendChild(<item id="3"><name>Milk</name></item>);
shop.prependChild(<notice>Open</notice>);
delete shop.item[0].@sale;                 // remove an attribute
delete shop.notice;                        // remove elements
shop.item[1].setName("product");           // rename an element

Assigning to a child that does not exist creates it. Assigning a string to an element replaces its content with a text node; assigning XML puts the XML there. copy() makes an independent copy of a subtree.

XML namespaces

Elements in an XML namespace are reached through a Namespace value and the :: operator, or by making it the default xml namespace.

var soap:Namespace = new Namespace("http://schemas.xmlsoap.org/soap/envelope/");
var body:XMLList = response.soap::Body;

var atom:Namespace = feed.namespace();      // the element's own namespace
for each (var title:XML in feed.atom::entry.atom::title) {
    trace(title);
}

var q:QName = new QName(atom, "entry");
trace(feed.child(q).length());

namespace(prefix), inScopeNamespaces(), namespaceDeclarations(), addNamespace(), removeNamespace() and setNamespace() work with the namespaces of a node; name() gives its QName and localName() the name without the namespace.

XML methods

XMLList has the subset that makes sense for a list: attribute, attributes, child, children, comments, contains, copy, descendants, elements, hasComplexContent, hasSimpleContent, hasOwnProperty, length, normalize, parent, processingInstructions, text, toString, toXMLString and valueOf. Called on a list of one item, the other XML methods work on that item.