35800 - class names in non-ASCII charset are not recognized

Status: VERIFIED FIXED

(Core :: CSS Parsing and Computation, defect, P3)

Description

[Marc Attinasi]

The CSSScanner is not designed to handle tokens that are in non-ASCII range.

nsCSSScanner::Next is used to read up tokens, however it assumes that characters 
are less than 256:

PRBool nsCSSScanner::Next(PRInt32& aErrorCode, nsCSSToken& aToken)
{
  PRInt32 ch = Read(aErrorCode);
  if (ch < 0) {
    return PR_FALSE;
  }
  if (ch < 256) {
    // OUCH! what about I18L ???

Comment 1

[Marc Attinasi]

This bug is a blocker for supporint @charset in an external style sheet. If a 
style sheet has character values > 256 then the CSS parser will not deal with 
it, so supporting @charset is of very dubious value.

Comment 2

[Marc Attinasi]

Moving to M20 since this is probably not a big priority and the changes required 
are quite extensive and risky. Workaround is to put classname in quotes.

Comment 3

[Marc Attinasi]

See bug 38038: possible patch proposed, and additional test case.

Comment 4

[Marc Attinasi]

Patch checked in, bug fixed. Thanks again [e-mail]!

Comment 5

[Hixie (not reading bugmail)]

Netscape's standard compliance QA team reorganised itself once again, so taking 
remaining non-tables style bugs. Sorry about the spam. I tried to get this done 
directly at the database level, but apparently that is "not easy because of the 
shadow db", "plus it screws up the audit trail", so no can do...

Comment 6

[Hixie (not reading bugmail)]

test case passes