xmlserializer serializes a DOM subtree or DOM document into XML/XHTML.
It understands documents generated by parse5 and regular browser DOMs (and thus can act as a drop-in replacement for XMLSerializer which for some browsers only serializes true XML documents).
Limitations
See the wiki for limitations in HTML to XML conversion.
Currently some cases are treated differently to the XMLSerializer implementation of the browsers:
- Invalid characters (ASCII control characters) that are invalid in XML 1.0 are removed on serialization. The browsers silently include those characters and on reparsing those documents throw a parser exception.
- Dashes in comments are escaped to provide valid comments in XHTML. Firefox does not do this.
- The
xmlnsattribute has higher precedence than the type of the DOM object passed to the serializer. - Small differences in style: no space in self-closing tag, empty value for boolean attributes, quoting of single apostrophes in attribute values.
This behaviour might become optional in the future.
Demo
See https://runkit.com/embed/siinwcyjdcjw
Run tests and build for the browser
$ npm install && npm test
Example
For a browser based example run npm test and see example.html.
The same code for Node.js:
var xmlserializer = ;var { var parser = dom = parser; return xmlserializer;};console;// => <html xmlns="http://www.w3.org/1999/xhtml"><head/><body><br/></body></html>