This is a very good point. With no delimiters specified the lenthKind delimited effectively becomes terminated by end of data. On Jun 27, 2012 12:59 PM, "Steve Hanson" <smh@uk.ibm.com> wrote:
After discussion by the DFDL WG, it was noted that applying a default lengthKind of 'delimited' to all the elements in a DFDL model for CSV would cause the root element to be using 'delimited' without any delimiters in scope. Further spec section 12.3.2 specifically allows for end-of-data being a valid reason for successfully ending a delimited scan. Therefore it is not correct for 'delimited' and no delimiters to give rise to an SDE.
Regards
Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, *OGF DFDL Working Group* <http://www.ogf.org/dfdl/> IBM SWG, Hursley, UK* **smh@uk.ibm.com* <smh@uk.ibm.com> tel:+44-1962-815848
From: Steve Hanson/UK/IBM To: Mike Beckerle <mbeckerle.dfdl@gmail.com> Cc: dfdl-wg@ogf.org, dfdl-wg-bounces@ogf.org Date: 21/06/2012 15:49 Subject: Re: [DFDL-WG] test case length_delimited_12_03 ------------------------------
Hi Mike
I think the test case can be explained by the fact that IBM DFDL does not yet support dfdl:lenghKind = 'endOfParent', hence any test case would try to use 'delimited'.
I don't have an issue with saying that encountering dfdl:lengthKind 'delimited' while parsing but having no delimiters in scope, is an SDE. It's almost certainly an indication the model is wrong.
Regards
Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, *OGF DFDL Working Group* <http://www.ogf.org/dfdl/> IBM SWG, Hursley, UK* **smh@uk.ibm.com* <smh@uk.ibm.com> tel:+44-1962-815848
From: Mike Beckerle <mbeckerle.dfdl@gmail.com> To: dfdl-wg@ogf.org Date: 21/06/2012 13:20 Subject: [DFDL-WG] test case length_delimited_12_03 Sent by: dfdl-wg-bounces@ogf.org ------------------------------
IBM published a set of test cases for DFDL. One of them is length_delimited_12_03.
I want to discuss whether this is in fact the behavior we want for DFDL.
In this example, there is data "abcde", and the default format is lengthKind="delimited", but no delimiters are defined anywhere.
<!-- parent with specified length -->
<xs:element name=*"myStringSeq2"* dfdl:lengthKind=*"explicit"* dfdl:length=*"5"*>
<xs:complexType>
<xs:sequence dfdl:initiatedContent=*"no"* dfdl:separatorPosition=*"infix"*
dfdl:sequenceKind=*"ordered"* dfdl:separator=* ""* >
<xs:element name=*"Test1"* type=*"xs:string"*>
</xs:element>
</xs:sequence>
</xs:complexType>
</xs:element>
The only way for this to parse is for the inner element Test1 to be delimited by the end-of-data coming from the explicit surrounding box created by the myStringSeq2.
I thought this should be a SDE because no delimiter is defined anywhere. I see that it can work, if we adopt a policy that delimited, but with no delimiters, means delimited by end-of-available-data.
But, I thought that is what lengthKind="endOfData" is for.
Can anyone comment as to why the lengthKind="delimited" but with no delimiter strings defined should NOT be an SDE? -- dfdl-wg mailing list dfdl-wg@ogf.org https://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
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
-- dfdl-wg mailing list dfdl-wg@ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg