W3C home > Mailing lists > Public > www-forms@w3.org > September 2009

regarding Test 5.1.b - validity of normalizedString and token

From: Joern Turner <joern.turner@googlemail.com>
Date: Tue, 22 Sep 2009 13:35:29 +0200
Message-ID: <67f00ca70909220435s6fcb7034n222f820cfb619b4f@mail.gmail.com>
To: www-forms@w3.org
Test Case 5.1.b tests the validity of several XForms builit-in
datatypes and checks this by setting valid or invalid values for the
respective datatypes. We wondered why we never got invalid events
dispatched for xsd:normalizedString and xsd:token from our processor
and did a bit of research to find the reason. While that's hard to
extract from the original Schema Spec we looked a various other
sources and found this cite in XML Schema, O'Reilly from Eric van der
Vlist (Page 26):

"There are no additional constraints on normalized strings. Any value
that is valid as a xs:string is also valid as a xs:normalizedString
but its tabs, linefeed and CR will be replaced by spaces. The
difference is the whitespace processing that is applied when the
lexical value is calculated'.

We also looked into the sources of Xerces J and found that whitespace
normalization is applied BEFORE the actual check of validity which
seems to be in sync with the above cite. In our interpretation this
means that the mentioned test case cannot work for xs:normalizedString
and xs:token as these types simply cannot become invalid.

It would be fine if the WG could confirm or reject our assumptions so
we can fix this testcase.

Joern Turner
Received on Tuesday, 22 September 2009 11:40:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 10 March 2012 06:22:17 GMT