I went back to an old ibm-internal document
and found a discussion that this property is used only to control output.
Is this still true?
This is my logic about this:
I.e., useNullValueForDefault applies
only to nullable items, and an element is required and if nothing
is in the logical value when you output, then it "defaults it"
by filling in the logical value with null, and then of course when outputting
that, the null is represented however it is represented.
On input, if an element is required
but is not present in the data at all, and it has a default value normally
that would used as the logical value.
useNullValueForDefault can either be
used in this situation to mean that the default value is to be null (which
can't be expressed in the XSD default="..." syntax), or one can
say this is ignored on input, and the user achieves this affect by putting
empty string on the list of nullValues, and not having any default value.
We need to reexamine this nullValues
as a list thing, That may or may not be sufficient, but no matter what
one certainly needs to be able to indicate that empty-string means null,
so there's no need for useNullValueForDefault on input.
Mike Beckerle
STSM, Architect, Scalable Computing
IBM Software Group
Information Platform and Solutions
Westborough, MA 01581
direct: voice and FAX 508-599-7148
assistant: Pam Riordan
priordan@us.ibm.com
508-599-7046
| Geoff Judd <JUDDG@uk.ibm.com>
12/03/2007 05:16 AM
|
|
To
| Mike Beckerle/Worcester/IBM@IBMUS
|
|
cc
| dfdl-wg@ogf.org, dfdl-wg-bounces@ogf.org
|
|
Subject
| Re: [DFDL-WG] OGF DFDL - useNullValueForDefault |
|
Mike,
The think the primary motivation for
the useNullValueForDefault is to allow out-of-band values to be output
as default values. A default value has to be in the value space of the
missing element but a Null value can be an out-of-band value. One scenario
where useNullForDefaultValue is useful is in the case of COBOL where a
user wants to output HIGH-VALUES or LOW-VALUES for missing elements.
Regards,
Geoff Judd
WebSphere Message Brokers Development
IBM United Kingdom Ltd.
Hursley
Telephone: +44-1962-818461
E-mail: JUDDG@uk.ibm.com
Mike Beckerle <beckerle@us.ibm.com>
Sent by: dfdl-wg-bounces@ogf.org
30/11/2007 20:34
|
|
To
| dfdl-wg@ogf.org
|
|
cc
|
|
|
Subject
| [DFDL-WG] OGF DFDL - useNullValueForDefault |
|
I'm trying to simplify something in the nulls/defaults space.
There's this property proposed: useNullValueForDefault.
This is a boolean.
I'm trying to understand what this is for.
On input, when the element is required but no content is found, and you
therefore want to create a default value, you would look, see that the
element is nillable, see that useNullValueForDefault is true, and set the
logical value of that element to null.
On output when no logical value is provided but the element is required,
you therefore want to create a default value, so you look and see that
the element is nillable and see that useNullValueForDefault is true and
so you provide the logical value null as the value. Then you output as
if the logical value was null in the first place, so you would output the
first of the dfdl:nullValues list as the representation.
These two are not symmetric.
I understand wanting empty content to cause null as the logical value of
a nillable element. Symmetry argues that one wants in this case for the
output side to output empty content as the representation of null logical
value.
The way I would express this is to provide a means for the list of dfdl:nullValues
to be able to contain the empty string as its first member of the list.
E.g., some syntax like:
dfdl:nullValues=" '' 'null' 'NULL' '\ ' "
Once you have this, there is no need for useNullValueForDefault. It's functionality
is subsumed by allowing nullValues to contain empty-string as a null representation.
This leaves the case where an element is both nullable and has a non-null
default value.
Is this case important or can we just say that an element is either nullable
or has default values, but not both. We already have some constraints
on this stuff in that we don't allow nullable complex types, only simple.
I believe allowing all combinations of nullability with default values
is probably overkill.
Comments?
I understand wanting a missing
Mike Beckerle
STSM, Architect, Scalable Computing
IBM Software Group
Information Platform and Solutions
Westborough, MA 01581
direct: voice and FAX 508-599-7148
assistant: Pam Riordan
priordan@us.ibm.com
508-599-7046
--
dfdl-wg mailing list
dfdl-wg@ogf.org
http://www.ogf.org/mailman/listinfo/dfdl-wg
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number
741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU