XML Formatter

Worked Examples

Before formatting

<root><user id="1"><name>Alice</name><email>alice@example.com</email></user></root>

After formatting

<root>
  <user id="1">
    <name>Alice</name>
    <email>alice@example.com</email>
  </user>
</root>

Declaration, comment and CDATA

The output below is what the tool returned in the browser for a one-line order document:

<?xml version="1.0" encoding="UTF-8"?><!-- order 1042 --><order id="1042" status="paid"><item sku="A-1" qty="2"/><note><![CDATA[Leave at <door> & ring]]></note><total currency="EUR">59.90</total></order>

Format with 2 spaces:

<?xml version="1.0" encoding="UTF-8"?>
<!-- order 1042 -->
<order id="1042" status="paid">
  <item sku="A-1" qty="2"/>
  <note>
    <![CDATA[Leave at <door> & ring]]>
  </note>
  <total currency="EUR">59.90</total>
</order>

The XML declaration is copied as written, because the browser’s DOM has no node for it. The comment and the CDATA section keep their content, so <door> and & inside CDATA are not escaped. Minify on the formatted text gives back the original single line. Attribute order and quote style follow the parsed document: every attribute is written with double quotes, and " < > & ' inside a value are escaped.

Mixed content changes

In documents where text and elements mix, such as XHTML or DocBook, the whitespace around inline elements carries meaning. The formatter does not keep it. <p>Hello <b>world</b>!</p> formats to:

<p>
  Hello
  <b>world</b>
  !
</p>

and minifies to <p>Hello<b>world</b>!</p>, which drops the space before “world”. Use the tool on data-style XML (configuration, API payloads, feeds); for markup documents, check the result before you replace the source.

Error messages

Errors come from the browser’s XML parser. The status line shows “Error: ” and the first error with its line and column. In Chrome 152 (measured on 2026-10-08), <company>AT&T</company> gives error on line 1 at column 14: EntityRef: expecting ';', and <a><b>x</a> gives error on line 1 at column 12: Opening and ending tag mismatch: b line 1 and a. Firefox words the same errors differently, and the message stays in English on every page language.

Whitespace-only elements

Both outputs below start from the same input. With two-space indentation, the first value loses its surrounding spaces. The blank element becomes a self-closing tag in formatted output. Minify also removes the blank text, but keeps that element’s opening and closing tags because the parsed element still has a text node. The already empty element stays self-closing.

Input

<root><value>  x  </value><blank> </blank><empty/></root>

Formatted with 2 spaces

<root>
  <value>x</value>
  <blank/>
  <empty/>
</root>

Minified

<root><value>x</value><blank></blank><empty/></root>

Attribute quotes and escaped text

The input uses single quotes for its attribute. Both operations write double quotes instead. The ampersand in the attribute and the less-than sign in the text are escaped again when the parsed values are serialized. Formatting this single text child adds no extra line breaks.

Input

<item name='A &amp; B'>1 &lt; 2</item>

Formatted or minified output

<item name="A &amp; B">1 &lt; 2</item>

Common Errors

The formatter parses input with the browser’s XML parser, which stops at the first well-formedness error and shows the browser’s message. These are the usual causes:

Limits

• Declared entities are expanded. The XML declaration and the DOCTYPE, including its internal subset, are copied to the output unchanged, but a reference to an entity declared there, such as &company;, is written as its replacement text.
• Text is trimmed. Leading and trailing whitespace inside elements is removed, and xml:space="preserve" is not honored: <v xml:space="preserve"> x </v> becomes <v xml:space="preserve">x</v>, although XML 1.0 §2.10 asks applications to keep that whitespace.
• During formatting, empty elements are self-closed. <note></note> and <note> </note> both become <note/>.
• Well-formedness only. The tool does not validate against a DTD or XML Schema.