W3C home > Mailing lists > Public > public-sml@w3.org > June 2007

[Bug 4686] Use schema terminorlogies to describe "xml schema valid"

From: <bugzilla@wiggum.w3.org>
Date: Thu, 21 Jun 2007 17:44:13 +0000
CC:
To: public-sml@w3.org
Message-Id: <E1I1QhZ-0001Ma-Lg@wiggum.w3.org>

http://www.w3.org/Bugs/Public/show_bug.cgi?id=4686

           Summary: Use schema terminorlogies to describe "xml schema valid"
           Product: SML
           Version: unspecified
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Core+Interchange Format
        AssignedTo: cmsmcq@w3.org
        ReportedBy: sandygao@ca.ibm.com
         QAContact: public-sml@w3.org


The "Model Validation" section has words like "document MUST be XML Schema
valid" or "must be valid under the xml schema". But XML Schema doesn't define
what qualifies as XML schema valid.

Schema assessment produces PSVIs, which is the only output. SML needs to
describe the expected behavior in terms of PSVI properties/values. There are 2
obvious options.

1. The [validity] PSVI property for the document element must be "valid".

2. The [validity] PSVI property for the document element must be "valid" and
there is no descendant information items (element or attribute) whose
[validity] is "invalid".

The difference between these 2 options is that if a subtree is laxly assessed
and something is marked [validity]=invalid in that subtree, the [validity] of
the room is not affected. In this case, option 1 would say it's valid; while
option 2 says it's not valid.

Depending on the context, we may want to pick one of these 2 alternatives (or
others).

A couple of concrete examples. In the core SML spec, section "Model
Validation". It's not obviously which alternative should be used here.

And in the SML-IF spec, section "The Basics", SML-IF documents are required to
be valid. Here it seems option #1 is more desirable.
A concrete example SML-IF doc
Received on Thursday, 21 June 2007 17:44:30 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:16:53 GMT