Contents of xref should use page title if contents is not specified
If the contents (i.e., text) in an xref macro is not specified, Antora should use the page title as the contents.
While this issue seems simple at first, there's a chicken-egg problem. It's possible that the target page has not yet been loaded, so Antora doesn't know what the page title is. Either Antora as to load all the page titles before starting document conversion, or Antora needs to perform a post-conversion step to fill in page titles where needed.
- If the text (i.e., contents) of an xref from one page to another is empty, it is filled in automatically from the xreftext of the target page
- If the text (i.e., contents) of an xref to a page in a nav file is empty, it is filled in automatically from the navtitle attribute of the target page, if present, otherwise the xreftext of that page