My immediate response is that Option 2 (rolling your own in code) is going to 
give you a useable result sooner - especially given the complications you have 
mentioned (multi-values in PERSON, which are translated individually to 
multi-valued fields in other files)

Option 1 may end up giving you a result, but I'd be afraid you would end up 
spending 5 times as long getting the dictionaries right vs.: rolling your own.

Option 3 is going to have you writing code anyway, which may end up being MORE 
difficult as you will be stitching up information from "unstructured" (XML 
documents - unless you have a parser) sources.



Ross Ferris
Stamina Software
Visage - an Evolution in Software Development


[FOOTNOTE: It IS possible to have a "generic" utility to export XML 
representations of arbitrarily complex, related & multi-valued data, BUT you 
need to keep some additional metadata within your dictionaries to get the 
nested relationships "right" - or at least that is the way that we tackled it 
within Visage. Our XML generation routines aren't "language driven" though (ie: 
we don't have an "extension" to ACCESS/PERFORM) - they are "solution driven" - 
we know what fields we need because they have been dropped on a 
form/report/cube/whatever, and we can determine the appropriate XML structure 
using the metadata in our extended dictionaries. Perhaps you CAN achieve what 
you are after using the IBM/U2 XML processing capabilities, and I'll follow 
this thread with interest]


>-----Original Message-----
>From: [EMAIL PROTECTED] [mailto:owner-u2-
>[EMAIL PROTECTED] On Behalf Of Wendy Smoak
>Sent: Friday, 14 January 2005 3:44 AM
>To: [email protected]
>Subject: [U2] Getting XML out of UniData
>
>I need to get an XML representation of a PERSON out of UniData.  This
>isn't a simple LIST statement out of PERSON, the data is in at least
>half a dozen other files.  Sometimes there are keys in PERSON to the
>other files, sometimes the other file is just keyed the same as PERSON.
>
>I can see two extremes - very ugly I-Descriptors added to PERSON (which
>already has over 3000 items, and I don't even know if it will work,
>because some of the data in the other files is itself multivalued-- can
>you have a phrase within a phrase?) or just opening and reading the
>records, and writing out the XML myself from UniData.
>
>A third option which sits in the middle is what I've done before, a
>combination of reading records FORMLISTing some select lists, then using
>DOWNLOAD to create "sub-parts" of the XML document, which I then
>concatenate in UniBasic.
>
>Any ideas appreciated!
>
>--
>Wendy Smoak
>-------
>u2-users mailing list
>[email protected]
>To unsubscribe please visit http://listserver.u2ug.org/
>
>
>--
>No virus found in this incoming message.
>Checked by AVG Anti-Virus.
>Version: 7.0.300 / Virus Database: 265.6.11 - Release Date: 12/01/2005
>

-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 265.6.11 - Release Date: 12/01/2005
-------
u2-users mailing list
[email protected]
To unsubscribe please visit http://listserver.u2ug.org/

Reply via email to