XML to JSON: The Mapping Rules and What Gets Lost

XMLBeautify Guides · Updated 2026-10-30 · All guides

There is no standard XML-to-JSON mapping. The converter uses a common one, and this page states it so you can predict the output.

The rules

XML	JSON
Element with only text	A string: `<n>Dune</n>` becomes `"n": "Dune"`
Attribute	A key prefixed with `@`: `<b id="1"/>` becomes `{"@id": "1"}`
Element with attributes and text	Text goes in `#text`
Empty element	`null`
Repeated sibling elements	An array, in document order
Root element	The single top-level key

All values are strings. XML has no number or boolean type, so <qty>2</qty> becomes "2". Converting types is your decision, because only you know whether 007 is a number.

The single-versus-array trap

One <item> becomes an object; two become an array. Code that reads doc.order.item.length works on orders with two items and breaks on orders with one. Handle it at the consumer: wrap a value in an array if it is not one.

What a conversion cannot keep

Order of mixed content. In one two three, the text pieces are joined into one #text, so where  sat is lost.
Comments and processing instructions. They are dropped.
Namespace meaning. Prefixes stay in key names (soap:Body) and xmlns declarations appear as @xmlns:soap attributes, but nothing resolves them.
Element versus attribute. <a b="1"/> and <a>1</a> differ in XML and differ in the output (@b versus b), so a consumer that accepts both forms must handle both.
Whitespace. Text is trimmed, so a value that was only significant whitespace becomes null.

Not well-formed input

The converter refuses to produce JSON for a malformed document and reports the line and column instead of emitting a partial result. See well-formed versus valid.