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 3054 - Availability of type definition
Summary: Availability of type definition
Status: RESOLVED FIXED
Alias: None
Product: XML Schema
Classification: Unclassified
Component: Structures: XSD Part 1 (show other bugs)
Version: 1.0/1.1 both
Hardware: PC Linux
: P2 normal
Target Milestone: ---
Assignee: C. M. Sperberg-McQueen
QA Contact: XML Schema comments list
URL:
Whiteboard: medium, easy
Keywords: resolved
Depends on:
Blocks:
 
Reported: 2006-03-28 02:06 UTC by C. M. Sperberg-McQueen
Modified: 2006-10-14 14:14 UTC (History)
0 users

See Also:


Attachments

Description C. M. Sperberg-McQueen 2006-03-28 02:06:29 UTC
Sections 3.2.5 and 3.3.5 specify that a type definition is known
for an element or attribute (only) if it is valid.

The consequence is that when an element or attribute is locally
invalid, because it does not conform to the relevant type
definition, the PSVI (as defined in 1.0 and the 1.1 status quo)
tells the downstream application that the element or attribute
is invalid, but does not vouchsafe any information about
which type definition it was found to be invalid against.

In simple cases, this can be found by examining the element or
attribute declaration and looking for the declared type 
definition.  In other cases, it requires redoing the work
already done by the schema validator, figuring out which type
definition the element or attribute should be validated against.
If there is any confusion about which type that is, or about
whether the validator actually used the correct type definition,
the current definition of the PSVI does not provide any useful
information.

The type definition should be described as being part of the
PSVI for every element or attribute which is locally assessed,
whether valid or invalid.
Comment 1 C. M. Sperberg-McQueen 2006-09-23 03:32:46 UTC
The Working Group discussed this issue at its telcon of 22 September 2006
and concluded that declarations and type definitions which govern 
elements and attributes should be identified in the PSVI whenever
a governing declaration or type definition was found and used for
schema-validity assessment.

There was brief discussion of the possibility that making this
change would involve loss of information in the PSVI, since the
presence of a type definition currently distinguishes certain
cases where an element is invalid (it helps distinguish whether
the error is in local validity or in a dependent).  We eventually
concluded, however, that the relevant information is also 
accessible through the [schema error code] property in the PSVI,
so there was in the end no objection to this change.

The editors are to prepare a wording proposal.  
Comment 2 C. M. Sperberg-McQueen 2006-10-14 14:14:51 UTC
A wording proosal was prepared along the lines described in
comment #1, and adopted by the working group at its meeting of
13 October 2006.  It has now been integrated into the status
quo documents.