This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.

Bug 15869 - NMTOKENS_minLength001_1464 should be invalid
Summary: NMTOKENS_minLength001_1464 should be invalid
Status: RESOLVED FIXED
Alias: None
Product: XML Schema Test Suite
Classification: Unclassified
Component: Microsoft tests (show other bugs)
Version: 2006-11-06
Hardware: All All
: P2 normal
Target Milestone: ---
Assignee: C. M. Sperberg-McQueen
QA Contact: XML Schema Test Suite mailing list
URL:
Whiteboard:
Keywords: decided
Depends on:
Blocks:
 
Reported: 2012-02-03 13:51 UTC by Andreas Meissl
Modified: 2012-03-09 23:55 UTC (History)
2 users (show)

See Also:


Attachments

Description Andreas Meissl 2012-02-03 13:51:43 UTC
In test NMTOKENS_minLength001_1464 of set MS-DataTypes2006-07-15 a simple type restricts xs:NMTOKENS with minLength="0". As defined at http://www.w3.org/TR/xmlschema11-2/#NMTOKENS, NMTOKENS has minLength="1" making a restriction with minLength="0" less restrictive and therefore invalid.

In my opinion this testcase should be invalid for schema 1.1 and (but not as clear as for schema 1.1) also for 1.0.

In schema 1.0 the minLength="1" of NMTOKENS is only defined by "A Schema for Datatype Definitions" - which is normative. The textual description only specifies the value space of NMTOKENS as the "set of finite, non-zero-length sequences of NMTOKENs", but doesn't explicitly define a value for the minLength facet. So there is some place left for discussions in case of schema 1.0.
Comment 1 Andreas Meissl 2012-02-06 11:02:50 UTC
The same for test NMTOKENS_minLength004_1467 and a similar case in NMTOKENS_maxLength001_1468 (maxLength="0" of restricted type isn't greater than or equal to minLength="1" of xs:NMTOKENS).
Comment 2 Andreas Meissl 2012-02-06 14:34:52 UTC
Other similar cases are positiveInteger_maxExclusive001_1834 and positiveInteger_maxExclusive002_1835 in the same test set, where maxExclusive="1" in the derived type conflicts with minInclusive="1" of the built-in type definition xs:positiveInteger.
Comment 3 David Ezell 2012-03-09 17:16:02 UTC
Resolution:  WG agrees that the test should be invalid for both 1.0 and 1.1.
Comment 4 Michael Kay 2012-03-09 23:55:01 UTC
I have fixed the expected results of these tests in CVS.