Henri Sivonen wrote:
There's no DTD for XHTML5 and writing one would be a bad idea. Also, the
conformance checking service runs the XML parser in the non-validating
mode without resolving external entities, because XHTML browsers
generally do not perform DTD validation and do not resolve external
entities and it would be a bad idea for a conformance checker to examine
a document tree that is different from what browsers see.
It would seem to me to be less work to write an appropriate RELAX NG
schema (probably a better idea than a DTD) to catch all these issues
than to write individual rules. External entities could be checked in a
separate step. But you're doing the work, so whatever you want to do. "-)
If so, wouldn't this catch any such issues?
It wouldn't. It would be part of the problem! (The document could
declare a random attribute to be of type ID in the DTD.)
One assumes you'd validate against the DTD or schema you that the spec
specified rather than the one the document specified.
--
Elliotte Rusty Harold [EMAIL PROTECTED]
Java I/O 2nd Edition Just Published!
http://www.cafeaulait.org/books/javaio2/
http://www.amazon.com/exec/obidos/ISBN=0596527500/ref=nosim/cafeaulaitA/
