polyglototron

In a bug thread on the polyglot spec

http://www.w3.org/Bugs/Public/show_bug.cgi?id=11064#c10

The question came up as to whether the constraints in that spec could or 
should be checked mechanically.

The attached is a first stab at such a mechanised check.

I believe (modulo a few bugs here and there) that a conforming html5 
document that is well formed xml meets the conditions of being a 
polyglot document just if the attached schematron file reports no errors.

there are two constraints not checked,

6.3.3 Attribute Values
7. Attributes


6.33 is a cross constraint between the document and any css or 
javascript that may interact with it, so may not be mechanicaly 
checkable, and is certainly not checkable with the tools being used here,

explict newlines in attributes are not visible to xml processing so can 
not be checked using tools based on xml parsing.

David