xmldom does not report parsererror (on XML without proper start tag)

[edi9999]

Compare this : (xmldom) :

> (new (require("@xmldom/xmldom").DOMParser)).parseFromString("a></a>", "text/xml").getElementsByTagName("parsererror")[0].textContent
Uncaught TypeError: Cannot read properties of undefined (reading 'textContent')

To this : (Firefox DOMParser) :

(new (DOMParser)).parseFromString("a></a>", "text/xml").getElementsByTagName("parsererror")[0].textContent
XML Parsing Error: syntax error
Location: about:newtab
Line Number 1, Column 1: newtab:1:1
"XML Parsing Error: syntax error
Location: about:newtab
Line Number 1, Column 1:a></a>

To this (Chrome DOMParser) :

(new (DOMParser)).parseFromString("a></a>", "text/xml").getElementsByTagName("parsererror")[0].textContent
'This page contains the following errors:error on line 1 at column 1: Document is empty\nBelow is a rendering of the page up to the first error.'

It seems that for the input "a>", xmldom says everythings ok will its not.

[karfau]

Thank you for reporting this.
Let me elaborate on the involved issues:

1. xmldom currently never adds the parsererror element as specified in the HTML spec (but only applies to parsing XML, not HTML mime types.
2. The specific snippet you provide has multiple issues when using it with xmldom:
   • it is not well-formed XML
   • xmldom doesn't detect it properly
   • the Document.documentElement is null when it's returned from parseFromString
   • the types provided by xmldom currently do not indicate that the documentElement could be null
3. .getElementsByTagName("parsererror")[0] will only ever work with a document that has errors, the NodeList returned by getElementsByTagName has a length of 0, so accessing the first element returns undefined:

https://stackblitz.com/edit/js-xmldom349?devtoolsheight=33&file=index.js

Regarding 1. At some point I want to introduce the option to add this node, but it would be a massive change in behavior and implementation, so it' currently not really "planned".

Regarding 2. We currently prioritize issues with well formed or even valid XML hihger as those regarding non well formed XML. But the goal should be to at least report a fatalError if there is no documentElement after parsing. And when doing that, we could also add the parsererror tag just in that case.

Regarding 3. not going to be solved by this lib :)

[brodybits]

I would like to chime in with a couple meta points:

• @karfau added the documentation tag to imply that updating the documentation would likely be the primary or most important part
• contributions of code & documentation updates from the community are always very welcome and encouraged for serious consideration (time permitting, of course but please feel free to follow up in case we drop the ball)

And thanks again to @karfau for essentially taking this project over!!

[karfau]

We label issues the explicitly want contributions with help-wanted and/or good first issue

Yes, documenting that would be helpful and contributions are always welcome, but I have not thought about where to best document this. We also have to document the current error handling in general (see #224), most likely this will happen prior to doing anything on this issue.
If somebody wants to contribute to this issue, please discuss the options here first.

[karfau]

starting from https://github.com/xmldom/xmldom/releases/tag/0.9.0-beta.9 (available under @xmldom/xmldom@next), a ParseError is thrown when the documentElement is not present after parsing is done.
This behavior was introduced as part of #454