[
Lists Home |
Date Index |
Thread Index
]
- From: richard@cogsci.ed.ac.uk (Richard Tobin)
- To: xml-dev@xml.org
- Date: 8 Mar 2000 23:26:42 GMT
>BTW: is there anywhere an updated version of the spec including the errata
>corrections?
No, but at some point there will be.
>So you would say, that `a b' should be invalid too?
No. The normalised value is "a b", and that matches production 8.
Erratum 8 clarifies that the validity constraints are applied after
normalisation.
>Section 3.3.3 says, that attribute-value normalization takes place before
>checking of the value. And since "[8] Nmtokens ::= Nmtoken (S Nmtoken)" I
>would say, both 'a b' 'a b' (checked as 'a b' and 'a\nb') are valid.
Right. I think the spec should be changed to require that the
normalised value match a production with #x20 instead of S. Otherwise
the normalisation doesn't have the obviously (!) intended effect of
passing the application a string of tokens separated by spaces. And
in any case I believe this change is required for SGML compatibility.
-- Richard
--
Spam filter: to mail me from a .com/.net site, put my surname in the headers.
"The Internet is really just a series of bottlenecks joined by high
speed networks." - Sam Wilson
***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************
|
