Last modified: 2013-10-10 23:08:54 UTC
To make it easy to process our output using both XML and HTML tools, we should serialize our output to so-called 'polyglot markup'. This means that our output will be valid XML *and* HTML5 at the same time (effectively XHTML). The spec for polyglot markup is at http://dev.w3.org/html5/html-xhtml-author-guide/. The relevant differences to our current HTML5 serialization should be: * void elements are serialized with trailing / as in <br/> * only a small set of named entities is used, other entities are converted to character entities ( becomes  ) We can add either add this functionality in Domino, or create our own XMLserializer implementation that walks an arbitrary DOM.
+1 ... HTML5's normalized parsing is great, but lots of programming environments only have an XML parser handy.
Change 88904 had a related patch set uploaded by GWicke: Bug 53968: Add XMLSerializer and use it to produce XHTML https://gerrit.wikimedia.org/r/88904
Change 88904 merged by jenkins-bot: Bug 53968: Add XMLSerializer and use it to produce XHTML https://gerrit.wikimedia.org/r/88904