Re: [ubl] Guidance sought regarding full UBL 2 XPath files

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

At 2006-01-21 09:41 -0500, I wrote:
>I was going to upload the XPath files for WD-UBL-2.0, but I need 
>your guidance please before doing so.
>...
>The ZIP file is more than 338Mb in size.  The unzipped size is 9.6Gb 
>(3.1Gb just for the text files).
>...
>The explosion happens because of the nested combinations of elements 
>... looking just at schema expressions does not always reveal how 
>the combinations build.  A seemingly innocuous addition might create 
>the possibility for a nested structure to be found in a new tree 
>location of the instance, thereby greatly increasing the 
>enumerations of uniquely addressed information items.

I thought that I would run a comparison of the eight document types 
of UBL 1.0 CD2 with the UBL 2 WD 2006-01-08 in order to provide more 
information with which to debate the relevance of these new XPath files.

The differences are remarkable, indicating to me that perhaps some 
inadvertent nesting of constructs was introduced in the UBL 2 
models.  The following are the summary statistics for UBL 1 and UBL 
2, with the UBL 2 models ordered first by the same document types in UBL 1:

UBL 1.0 CD2:

DespatchAdvice:          7255 elements, 20539 attributes, 27794 total
Invoice:                 2320 elements, 6540 attributes, 8860 total
Order:                   9194 elements, 25898 attributes, 35092 total
OrderCancellation:        296 elements, 800 attributes, 1096 total
OrderChange:             9198 elements, 25930 attributes, 35128 total
OrderResponse:           9186 elements, 25886 attributes, 35072 total
OrderResponseSimple:      295 elements, 800 attributes, 1095 total
ReceiptAdvice:           5653 elements, 16019 attributes, 21672 total

UBL 2 WD 2006-01-08:

DespatchAdvice:       185022 elements, 465343 attributes, 650365 total
Invoice:               12810 elements, 32445 attributes, 45255 total
Order:                596323 elements, 1500511 attributes, 2096834 total
OrderCancellation:      1759 elements, 4310 attributes, 6069 total
OrderChange:          596318 elements, 1500509 attributes, 2096827 total
OrderResponse:        596323 elements, 1500508 attributes, 2096831 total
OrderResponseSimple:    1724 elements, 4232 attributes, 5956 total
ReceiptAdvice:        189379 elements, 476304 attributes, 665683 total

ApplicationResponse:    2888 elements, 7212 attributes, 10100 total
AttachedDocument:       1260 elements, 3089 attributes, 4349 total
BillOfLading:          93768 elements, 235730 attributes, 329498 total
Catalogue:              5667 elements, 14028 attributes, 19695 total
CatalogueDeletion:      1326 elements, 3242 attributes, 4568 total
CatalogueItemSpecificationUpdate: 5676 elements, 14051 attributes, 19727 total
CataloguePricingUpdate: 5676 elements, 14051 attributes, 19727 total
CatalogueRequest:       5311 elements, 13163 attributes, 18474 total
CertificateOfOrigin:   97544 elements, 244943 attributes, 342487 total
CreditNote:             3550 elements, 8978 attributes, 12528 total
DebitNote:              3550 elements, 8978 attributes, 12528 total
ForwardingInstruction: 93360 elements, 234739 attributes, 328099 total
FreightInvoice:       102454 elements, 258016 attributes, 360470 total
PackingList:           93730 elements, 235622 attributes, 329352 total
Quotation:            298834 elements, 751896 attributes, 1050730 total
RemittanceAdvice:       6111 elements, 15297 attributes, 21408 total
RequestForQuotation:  150428 elements, 378414 attributes, 528842 total
SelfBilledCreditNote:   3814 elements, 9703 attributes, 13517 total
SelfBilledInvoice:     11988 elements, 30267 attributes, 42255 total
Statement:              7350 elements, 18473 attributes, 25823 total
Waybill:               93766 elements, 235721 attributes, 329487 total


Note, for just one example, how UBL 1 Order has 35,128 information 
items while UBL 2 Order has 2,096,834 information items ... this does 
not feel right to me.  Has Order changed a lot?  If so, my worries 
are unjustified.  If not, then something inadvertent must have 
happened somewhere.

Because of this contrast, I think I will await posting the 338Mb ZIP 
of the 9.6Gb summary of UBL 2 XPath addresses for 206-01-08.  Please 
consider examining the UBL 2 declarations for any inadvertent 
explosion of information item contexts.  I will be pleased to 
recreate these files for any new models, just let me know.

Thanks!

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




--
Upcoming XSLT/XSL-FO hands-on courses:  Denver,CO March 13-17,2006
World-wide on-site corporate, govt. & user group XML/XSL training.
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/o/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Cancer Awareness Aug'05  http://www.CraneSoftwrights.com/o/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal