Fix "XML is not defined" — migrating E4X to XmlService

ReferenceError: XML is not defined
SyntaxError: Unexpected token '<'
SyntaxError: Invalid or unexpected token     (from item.@id)

Why it happens

Rhino implemented E4X (ECMAScript for XML): XML literals in code, new XML(text), xml.@attr, xml..descendant. V8 has none of it. Literals and operators break parsing; new XML() fails at runtime.

The fix: XmlService

// Rhino (E4X)
var feed = new XML(text);
for each (var item in feed..item) {
  Logger.log(item.title + ' ' + item.@id);
}

// V8 (XmlService)
var root = XmlService.parse(text).getRootElement();
var ns = root.getNamespace();
root.getDescendants().forEach(function (node) {
  var el = node.asElement();
  if (!el || el.getName() !== 'item') return;
  Logger.log(el.getChildText('title', ns) + ' ' + el.getAttribute('id').getValue());
});

E4X	XmlService
`new XML(text)`	`XmlService.parse(text).getRootElement()`
`x.child`	`x.getChild('child', ns)`
`x.child.toString()`	`x.getChildText('child', ns)`
`x.@id`	`x.getAttribute('id').getValue()`
`x..item`	filter `x.getDescendants()` by name
`<a>{v}</a>` literal	`XmlService.createElement('a').setText(v)`
`x.toXMLString()`	`XmlService.getPrettyFormat().format(doc)`

Gotchas

Namespaced feeds (Atom, SOAP) need the Namespace object for every getChild call — getChild('entry') returns null otherwise.
Building XML by string concatenation is fine, but escape & < > in values.

How V8Shift handles it

E4X can't be rewritten safely by rules, so every literal, new XML, .@ and .. is listed with file and line in the manual checklist. (If an AI key is configured for paid jobs, those files get an extra refinement pass.)