RE: Unicode and XSL (was substring())

 
Mike Kay:
> > (The current definition [of translate()], of course, is hopelessly
> > informal).
>
James Clark: 
> It is informal, but I hope it is clear.  Strings are defined as
> sequences of UCS characters (ie character means the same as it does in
> XML).  Apart from the issue of what happens if the second and third
> arguments are of unequal lengths, it seems well defined to 
> me.  If there are other cases where the intended behaviour is not clear, 
> please report them to the XSL editors list.
> 
The conditions where behaviour is not obvious are:
- where arg2 and arg3 are unequal in length
- where the same character appears more than once in arg2

The areas where behaviour is not clearly specified, but fairly intuitive,
include:
- a definition of "corresponding" (a formal definition depends, I think, on
being able to enumerate the characters in a string, something that XSLT has
avoided doing)
- a reference to the rules for equality matching between characters (which
is defined elsewhere in XSLT, if not entirely satisfactorily).

Mike


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list