Detect new unresolved xref structure
In Antora 2.1, the unresolved xref will change to preserve the content (i.e., the linked text).
Consider the following unresolved xref:
xref:no-such-page.adoc[No Such Page]
Here's the current output:
<a href="#">no-such-page.adoc</a>
Here's the new output:
<a href="#no-such-page.adoc" class="page unresolved">No Such Page</a>
The page ID spec (and optional fragment) gets encoded as an href fragment. The class "unresolved" is appended to the element (along with the "page" class). The content is preserved.
The xref validator should be updated to look for and parse this new structure when upgrading the pipeline to Antora 2.1 or better.
See the related issue for more details: antora#421 (closed)