[DFDL-WG] DFDL: Minutes from OGF WG call, 30 Apr 2008

[Ian W Parkinson]

Hi all,

I've just noticed - I didn't distribute the minutes from last week's WG 
call. Apologies for the delay.

Cheers,
Ian



Open Grid Forum: Data Format Description Language Working Group

Weekly Working Group Conference Call
16:00 GMT, 30 Apr 2008


Attendees
Mike Beckerle (Oco)
Steve Hanson (IBM)
Ian Parkinson (IBM)
Alan Powell (IBM)


1. F2F meeting at IBM Hursley
Mike has confirmed that he will travel to IBM Hursley to further design 
discussions with the IBM team, later in May.


2. Review of specification line items
The meeting began a review of the outstanding work required for 
specification draft 32, but this was not completed during the meeting as a 
number of technical discussions were triggered.

valueCalc Mike has provided material to Alan. No further work is 
anticipated on this item.

nulls/defaults/optionals There has been some discussion on 
nilIndicatorPath and forward references, and this discussion needs to 
continue. The rest of this section is now complete.

UML description of schema components complete

Property Precedence This item is currently marked complete; however it has 
sparked some discussion. On unparse, if the infoset contains less than the 
required number of occurances of some field, then the unparser may be 
expected to output sufficient instances of the default value. We need to 
craft a policy to determine exactly how this achieved; for example, do we 
use the xsd:minOccurs value or do we try to match the result of the 
dfdl:occurs expression? Mike proposed that if an occurs expression is 
present, it must evaluate to a valid number of occurances. Steve suggested 
that we disambiguate dfdl:occurs from xsd:minOccurs and xsd:maxOccurs by 
renaming the property dfdl:occursPath - however, this property can be an 
expression involving calculations, not just a simple path. A number of 
simplifications have been proposed, and these require further discussion.

Alan noted that an XPath expression has no way to access a field's default 
value, and suggested introducing a function to handle this. Also, we may 
need to define count() such that it returns the number of instances of an 
element after any necessary padding and introduce a different way to 
obtain the number of instances found in the infoset.

Stop values and terminators are not considered part of the logical data.

Steve will update the diagram.

Variable Markup Steve had previously proposed a scheme whereby properties, 
such as iniitator or terminator, could be described by another DFDL 
element or type definition. We could use a similar mechanism to handle 
strings with embedded complex structures. This subject will be tackled 
during the F2F.

Regular Expressions for Lengths The meeting agreed to adopt the regular 
expression subset previously distributed by Alan. Regular expressions are 
ignored for unparsing, and are only used to determine the length of an 
element during parsing. Considering repeating elements, Steve contrasted 
the behaviour of regular expressions in DFDL (where the regular expression 
matches just a single instance) with that in IBM's MRM technology (where a 
regular expression matches the entire array), and the meeting discussed 
whether a regular expressions should include any necessary terminators.


Meeting closed, 17:10 GMT



Ian Parkinson
WebSphere ESB Development
Mail Point 211, Hursley Park, Hursley, Winchester, SO21 2JN, UK





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





-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.ogf.org/pipermail/dfdl-wg/attachments/20080506/ea8ba080/attachment.html