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 536 - setParameter in domconfigschemalocation1
Summary: setParameter in domconfigschemalocation1
Status: ASSIGNED
Alias: None
Product: DOM TS
Classification: Unclassified
Component: DOM Level 3 (show other bugs)
Version: unspecified
Hardware: Other other
: P2 normal
Target Milestone: ---
Assignee: Philippe Le Hegaret
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2004-02-18 12:32 UTC by nnissar
Modified: 2007-03-29 18:48 UTC (History)
0 users

See Also:


Attachments

Description nnissar 2004-02-18 12:32:25 UTC
canSetParameter can return true for value
'sampleSchemaLocation' since the 'schema-location' parameter could be  
set with this value.  However, setParameter (with value  
'sampleSchemaLocation') may not be successful since this parameter is ignored 
unless the "schema-type" parameter value is set. In this case, a  
NOT_SUPPORTED_ERR may be raised.

clarification is being requested from the WG
Comment 1 Andrew Clover 2004-02-19 09:49:04 UTC
IMO canSetParameter with any URI should also be able to return false for
non-validating implementations.
Comment 2 Curt Arnold 2004-02-19 16:14:06 UTC
There was a brief discussion related to this test on the WG conference call on 
18 Feb 2004.  The general feeling was that you should not throw an exception 
on setParameter if canSetParameter returned true.

The 'schema-location' parameter is optional, so non-validating implementations 
can just not support the parameter.  The note on ignoring 'schema-location' 
when 'schema-type' was not set eliminates the need to spell out a specific 
sequence (set schema-type first, then schema-location).  It seems reasonable 
that 'schema-location' would also be ignored if 'validate' and 'validate-if-
schema' are false.

At this point, I believe the test is correct.  I'm reassigning the bug to 
Phillippe so he emailed on any additional discussion.

p.s. I raised issues on the description of this parameter in 
http://lists.w3.org/Archives/Public/2004JanMar/0048.html
Comment 3 Andrew Clover 2004-05-22 20:36:57 UTC
In any case, having canSetParameter return false for schema-location seems
reasonable for an imp that does not support Schema. The test should either have
implementationAttribute schemaValidating, or allow canSetParameter for a URI to
return false.

Incidentally, the assertion that setting schema-location to a non-string,
non-null type (boolean true in this case) must fail is not supportable under a
dynamically-typed language such as Python. (Metaclasses allow an object to be a
both boolean and string!)