- From: Philip Taylor (Webmaster, Ret'd) <P.Taylor@Rhul.Ac.Uk>
- Date: Mon, 19 Jul 2010 13:38:50 +0100
- To: David Dorward <david@dorward.me.uk>
- CC: Richard Ishida <ishida@w3.org>, www-validator@w3.org
David Dorward wrote: > > On 19 Jul 2010, at 13:17, Richard Ishida wrote: > >> What follows is the output of a validation session. The validator seems to cope fine with php PIs other than inside attribute values. Please recognise these as PIs and allow the file to be validated. > > The markup validator validates markup, not PHP source code that generates markup. > > Run the software you have written and validate the output. > > Trying to deal with the PHP code would be, for the simplest cases, impractical and vulnerable to false positives and negatives. For more complex cases it would be a security hazard. Although I have considerable sympathy with this position (it was also my own immediate reaction), I note that the HTML 4.0 specification does in fact recognise the concept of a Processing Instruction (PI) : http://www.w3.org/TR/REC-html40/appendix/notes.html#h-B.3.6 I haven't yet had a chance to work out whether this "recognition" is sufficient for an otherwise-valid document containing valid PIs to therefore qualify as valid itself. Philip Taylor
Received on Monday, 19 July 2010 12:39:45 UTC