Writer::new_with_indent() inserts white space into <![CDATA[ content tags

[klausi]

Problem: the pretty print functionality should not insert spaces new lines before and after <![CDATA[ tags.

Example input:

<?xml version="1.0" encoding="UTF-8"?><jobs><job><jobid><![CDATA[00d46e4494e1]]></jobid></job></jobs>

Output with Writer::new_with_indent(out_file, b" "[0], 2);

<?xml version="1.0" encoding="UTF-8"?>
<jobs>
  <job>
    <jobid>
      <![CDATA[00d46e4494e1]]>
    </jobid>
  </job>
</jobs>

This has now changed the content of the jobid tag, which is bad.

Correct with xmllint --format:

<?xml version="1.0" encoding="UTF-8"?>
<jobs>
  <job>
    <jobid><![CDATA[00d46e4494e1]]></jobid>
  </job>
</jobs>

This has correctly preserved the content of the jobid tag.

[tafia]

Thanks for opening the issue.

Do you have a specification to refer to? This is not obvious to me that the current behavior is wrong.

[klausi]

Could not find anything concrete in the XML specification, only http://xml.silmaril.ie/whitespace.html which talks about significant and insignificant whitespace.

A similar bug was reported to xml-js nashwaan/xml-js#14 where they also decided to not indent CDATA sections.

[klausi]

Another data point is the PHP DOMDocument output which formats the same way as xmllint and does not indent CDATA sections. https://3v4l.org/DviGR

[klausi]

Same with Python

import xml.dom.minidom

dom = xml.dom.minidom.parseString('<?xml version="1.0" encoding="UTF-8"?><jobs><job><jobid><![CDATA[00d46e4494e1]]></jobid></job></jobs>')
print(dom.toprettyxml())

<?xml version="1.0" ?>
<jobs>
    <job>
        <jobid><![CDATA[00d46e4494e1]]></jobid>
    </job>
</jobs>

Not sure if Python, PHP and xmllint all rely on the same XML parsing library, but I think these are pretty good indicators that quick-xml should do the same.

[cdwilson]

FWIW, Altova XMLSpy also reformats this input as @klausi suggested: