OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

ubl-lcsc message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: [ubl-lcsc] Excel->XSD, XSD->Excel


>>i am unsure what you mean by 'working on a script to convert...', i
>>thought we had one of those already?  to my knowledge the perl script
>>can generate all XSDs except the re-usable one (it only does 98% of
>>that).  i think the issue was with name spacing and not the content of
>>the components themselves.  However I am sure Gunther (or someone else
>>from tools and techniques) on the list can clarify this.

As I understand it, for now, the internal production process is 
that the Excel sheets get transformed into XSD, then get applied
with minor or no hand-editing to the XSDs.  These XSDs are then
declared as  normative.  To the "outside" world, the XSDs are 
always normative,  but due to the above "production" process, 
the Excel sheets remain a source for now.

At some point, we may want to declare a version/date-time to
reverse the transformation, that Excel sheets are no longer
the definitive source, but the XSDs edited by hand (via tools
or otherwise) are the normative definitions and also the 
production sources.  Once that turning point is reached, 
Excel sheets may continue but only as a transformed by-product
to augment or ease reading.  In other words, any discussion
or changes will be made directly at the XSDs.

Also, I think in any case, it may be better to take out the
entire Excel directory from the official UBL spec releases
and distribute those Excel sheets as a separate archive.
Whilest the committee might want to show all the steps needed
to produce the XSDs, I think in retrospect (regarding the
UBL0p70 release) the existence of Excel directory seems to
cause a bit of anxiety among people who might be concerned
about whether Excel will continue to be production source
(though they may not be normative).

Finally, I'm working on one-shot (ie. no hand-editing of 
content)  transformation tools to see if I could get the XSDs 
when I pipe in Excel.  This resulted in discovery of some 
remaining inconsistencies between Excel sheets and XSDs, 
some of which I've posted to ubl-comment earlier (there 
are more, but I'll try to put them in RTF and send to Bill).  

I've also gotten some mileage towards converting XSDs
back to Excel (again, no hand-editing of content).  On some 
areas, however, I'm not quite certain  yet, as I'm not entirely 
clear on some of the column definitions and why some of 
the formulas differ between rows though  they're along 
the same column.  To which forum or person do you think
I should discuss that aspect?

Thanks.


cheers,
Chin Chee-Kai
SoftML



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]