Re: [CSS21] WD 4.2: end-of-string vs. end-of-stylesheet

I attach a .tar.gz archive containing an HTML document that tests
whether UAs give precedence to BAD_STRING or the "unexpected end of
style sheet" rules.  It similarly tests the corresponding BAD_URI and
BAD_COMMENT questions, and tests whether UAs interpret "all open
constructs" to include the case of FUNCTION with no matching ')'.

The document makes some provisional proposed textual changes according
to the results I saw; though of course it would be good to check some
other UAs before deciding what behaviour(s) to allow.

The document can also be viewed at

  http://bowman.infotech.monash.edu.au/~pmoulder/html-tests/premature-eof/premature-eof.html

pjrm.