[DFDL-WG] MBTK and Daffodil - Intentioning Violating Property Scoping Rules?

Thu Apr 25 10:51:25 EDT 2013

I think what you’re saying is that I’m failing to take into account this part of the schema:

      <xsd:annotation>
            <xsd:appinfo source="http://www.ogf.org/dfdl/">
                  <dfdl:format ref="GeneralPurposeFormat" />
            </xsd:appinfo>
      </xsd:annotation>

This avoids the out of scoping issue by providing a reference to the named <dfdl:format> in the GPF file:

                                    <dfdl:defineFormat name="GeneralPurposeFormat">
                                                <dfdl:format alignment="1" alignmentUnits="bytes"
                                                            binaryBooleanFalseRep="0" binaryBooleanTrueRep="1"
                                                            …

I rewrote some of the schemas to confirm that this works.

Obviously this is my all mistake, so everything I’ve said below is wrong.  I apologize!

From: Tim Kimber [mailto:KIMBERT at uk.ibm.com]
Sent: Thursday, April 25, 2013 10:20 AM
To: Garriss Jr., James P.
Cc: dfdl-wg at ogf.org
Subject: Re: [DFDL-WG] MBTK and Daffodil - Intentioning Violating Property Scoping Rules?

OK. I'll bite...

Some features of DFDL take some time to understand, and some are not perfectly described in the specification. That does make it hard for somebody in your position because you only have the specification to go on, As far as I'm concerned, you should feel free to raise questions like this without apology.

In this case, neither Daffodil nor IBM DFDL is violating the (intended) rules in the specification. The properties from the imported xsd are not being used by a *component* in the main xsd. Instead, the 'ref' attribute of the global format block in the main xsd is referring to a named format block in the imported xsd. In this way, the library xsd is contributing a wide-ranging list of default values into the global format block of the main xsd.

I expect Steve or Mike will add some details to that...

regards,

Tim Kimber, DFDL Team,
Hursley, UK
Internet:  kimbert at uk.ibm.com<mailto:kimbert at uk.ibm.com>
Tel. 01962-816742
Internal tel. 37246742

From:        "Garriss Jr., James P." <jgarriss at mitre.org<mailto:jgarriss at mitre.org>>
To:        "dfdl-wg at ogf.org<mailto:dfdl-wg at ogf.org>" <dfdl-wg at ogf.org<mailto:dfdl-wg at ogf.org>>,
Date:        25/04/2013 15:00
Subject:        [DFDL-WG] MBTK and Daffodil - Intentioning Violating Property        Scoping Rules?
Sent by:        dfdl-wg-bounces at ogf.org<mailto:dfdl-wg-bounces at ogf.org>
________________________________

(Emotions are hard to convey in email; please trust me when I say that I am writing this email with kindness and friendliness!)

In section 8 it says, “The dfdl:format annotation on the top level xs:schema declaration provides defaults for the DFDL representation properties at every DFDL-annotatable component contained in the schema document. They do not apply to any components in any included or imported schema document (these may have their own defaults).”

If I understand this, it means that when properties are defined using <dfdl:format> in one DFDL schema file, they are out of scope for any other DFDL schema file.

So if schema A defines some properties and includes schema B, the properties are out of scope in schema B.
Similarly, if schema A includes schema B and schema B defines some properties, the properties are out of scope in schema A.

Is that right?  I think so, and I have empirically confirmed this in both tools.

Ok, so you know where this going, right?  Why does the following line work?

<xsd:import namespace="http://www.ibm.com/dfdl/GeneralPurposeFormat" schemaLocation="IBMdefined/GeneralPurposeFormat.xsd"/>

According to the spec, it shouldn’t.  Yet both tools support it.

But if you make any changes to the GeneralPurposeFormat, it breaks.  You can’t rename it.  You can’t put it in a different folder.  Etc.

Here’s what I suspect:  Both MBTK and Daffodil hard-coded this as an undocumented exception to the rule.

I think you want to have your cake (properties are out of scope) and eat it, too (except when we want them to be in scope because repeating all the properties in every DFDL file is a pain).

If I’m wrong, just let me know.  It’s entirely possible that I don’t really understand what’s going here.

But if I’m right, then you guys should not do this.

•       If the spec makes sense, then you should follow the spec.
•       If the spec doesn’t make sense, then you should change the spec.
•       If the spec needs an exception to the rule for this one case, then add an exception and follow it.

To intentionally break the spec in an undocumented fashion seems wrong.--
 dfdl-wg mailing list
 dfdl-wg at ogf.org<mailto:dfdl-wg at 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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ogf.org/pipermail/dfdl-wg/attachments/20130425/f9bbf6fa/attachment-0001.html>