Being <x> any valid element, using htmlPrefilter() to parse a <br /> element inside an attribute of an <x> element adds a spurious </x> after it:

$.htmlPrefilter('<x foo="<br />"></x>');

This returns <x foo="<br ></x>"></x> (note the extra </x> inside the foo attribute).

I browsed other issues and I saw that a related one was marked as invalid because void elements must not have closing tags in HTML. This might be the same case, but the docs say that "[htmlPrefilter] will greedily ensure that all tags are XHTML-compliant", and I understand that <br /> is definitely valid XHTML (and this behavior ironically makes it invalid).

Is this a documentation thing, a bug (probably in the rxhtmlTag regex?) or am I missing something else?