Re: [ubl-dev] UBL Code List Value Validation Methodology version 0.7 Draft 1

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

At 2006-09-25 16:33 +0100, Stephen Green wrote:
>How about adding something for Identifiers too.

To the documentation?  As it stands trading 
partners can add their own lists of coded values 
associated with identifier-based information items.

>One example could be a pattern for UUIDs. Is
>there scope in genericode for this too.

Ahh ... patterns?  No.  But an enumeration of a 
pattern, yes.  The humungous 
ContainerSizeTypeCode-2.0.gc has an enumeration 
of numerous combinations of patterned values.

>I guess
>one might assume that if it has a pattern and
>not a codelist then it is an Identifier and not
>a Code, but the margins do tend to blur.

Indeed ... but if you *really* wanted to 
implement "pattern checking" then this could 
probably be done as a business rule included with the synthesized coded rules.

Version 0.7 is modular in the use of Schematron 
patterns (not the same meaning as "pattern" in 
your phrase "pattern for UUIDs") ... the 
methodology creates Schematron patterns that are 
pulled together by users into Schematron 
schemas.  To add to the constraints, users create 
their own Schematron patterns and include them as 
well into the Schematron schemas.

>I notice
>TaxCategory has an identifier in UBL 2, not a code.
>This would presumably cover tax codes such as
>L, A, X, Z ,etc which tend to have to be agreed
>between trading partners - in the UK, for VAT, at
>least. These are typical examples for this method
>but, as I say, are Identifiers rather than 'Codes'
>in UBL 2 so how would the methodology apply to
>them as it stands?

By created a "code list" in genericode of the 
values allowed, with all the meta data and 
documentation you wish, and then addressing the 
related information items as if they were of type CodeType.

>If the methodology were
>extended to cover Identifiers then maybe it
>should cater for identifiers for which there are
>'codelists' (identifier lists :-) even if such lists have
>to be agreed between parties as with tax category
>'identifiers')

Right ... just by mimicking code lists with 
identifier lists, but using genericode 
expressions of both.  Of course you could modify 
the generation of Schema patterns to use other 
than genericode and pull in the Schematron 
patterns into the final Schematron schema by some 
as yet not considered expression language.

>but also for patterns such as for UUID.

But genericode doesn't express algorithmic combinations.

>The UUID would benefit from the methodology in
>that there would then be a more standard way to
>document in trading partner agreements the
>associated attribute values for schemeID, etc.
>For other identifiers these supplementary
>component metadata values would be more
>important to agree in advance and use for applying
>constraints in the secondary validations.

Well, enumerations are handled but algorithmic 
patterns are not ... by genericode ... I'm not 
sure how one could create an expression of the 
pattern short of using regular expressions, but 
regular expressions are not supported in XSLT 
1.0.  That doesn't mean one couldn't come up with 
an XSLT 2.0 implementation of Schematron which 
would enable an outboard expression of the reg-ex 
of the pattern you need to match ... but not in 
the short time frame of what we can realistically produce.

Does it warrant adding to the 0.7 documentation 
that the use of genericode files for an 
enumeration of identifier values is not 
inappropriate?  Just that that isn't done by 
default (because there are no built-in 
enumerations of identifiers in UBL; if there were 
then this would have been done), and nothing prevents users from doing so.

. . . . . . . . . . . . Ken

--
UBL/XML/XSLT/XSL-FO training: Vårø, Denmark 2006-10-02/06,11-20/24
UBL International 2006  2006-11-13/17 http://www.ublconference.com
World-wide corporate, govt. & user group UBL, XSL, & XML training.
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/u/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Cancer Awareness Aug'05  http://www.CraneSoftwrights.com/u/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal