- From: Jeni Tennison <jeni@jenitennison.com>
- Date: Tue, 24 Sep 2002 09:58:16 +0100
- To: xmlschema-dev@w3.org, ht@cogsci.ed.ac.uk (Henry S. Thompson)
- Cc: rainerbecker.mail@t-online.de (Rainer Becker), Dare Obasanjo <dareo@microsoft.com>
Henry, >> Because the whitespace normalization takes place BEFORE the >> attribute value is validated against the schema, the second >> instance should be valid too. Can one confirm please? > > I believe your analysis is correct, and your instance should be OK > -- which validator are you getting the error from? I've done some testing and it seems that MSXML behaves this way for all types from xs:token down. When you try doing: <foo xsi:type="xs:token"> bar </foo> MSXML wrongly complains that " bar " is not a token because "token is the set of strings that do not contain the line-feed (#xA) or tab (#x9) characters, that have no leading or trailing spaces (#x20) and that have no internal sequences of two or more spaces." It appears that MSXML does not carry out whitespace normalization before testing the value, though it does on those data types that are not derived from xs:string. Cheers, Jeni --- Jeni Tennison http://www.jenitennison.com/
Received on Tuesday, 24 September 2002 05:05:38 UTC