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


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Application Design

-----Original Message-----
From: Mike.Champion@SoftwareAG-USA.com
Sent: Saturday, August 11, 2001 11:58 AM
To: xml-dev@lists.xml.org
Subject: RE: Application Design

>>I'm playing the unusual role (for me) of being the XML cheerleader in this
thread partly to figure out what "best practice" really is in this area.  I
wouldn't be crushed to find out that XSLT doesn't really belong in the
"solid core of XML," but would like to see us share experiences about what
people have done with it, or failed to do with it, in practice.

Well, I've hit one thing I couldn't see how to do in XSLT:

I had a situation where the RDBMS guys had taken a perfectly good tree and
turned it into a
table with an extra column identifying the parent record. In order to
convert it back, I needed
a stack to push the parents on to. A list would have worked, but I couldn't
figure out any way
to get dynamic temporary storage in XSLT.

In the same app they had taken a table where every row was a meaningful
record and exported it
in column-major form, so that each row I recieved was actually a column. A
for-each worked
to get the data back into meaningful form, but I couldn't figure out how to
do the other
direction, again without temporary storage to deal with the unknown number
of rows. I'm not
sure that this wasn't possible in XSLT, but since I was already writing
procedural code, I
punted when my head started to hurt.

So there's the limit I've observed: no dynamic storage for any looping that
can't be done with a for-each.