Hi. My xml files are actually XML definitions of different product installations for IzPack open source Installer. IzPack consists of 2 main part: JAVA written engine. Specific installation definitions in format of XML.
Since our products are numerous and quite complicated - those XML definitions are quite complex and have a lot of interconnections... So - I loaded all them into base-X and use it to perform different queries through whole XML data. But - to remain in world of XQuery/Xpath - I have to rewrite substantial part of JAVA engine, that read/parse/manage those XMLs. It is a lot of work...
2014-01-29 Christian Grün christian.gruen@gmail.com
Dear Moshe,
I agree that such a feature would be desirable. The original byte offset of an XML node will not be preserved as we try to minimize the size of our databases. One more reason is that such information gets invalid anyway after database updates.
One question: What's your reason for looking up the data in the original file? Maybe there's some chance to stay in the XPath/XQuery world for the operations to follow?
Other suggestions from the community are welcome, Christian
On Tue, Jan 28, 2014 at 5:15 PM, Moshe Zuisman zuismanm@gmail.com wrote:
Hi. I am using base-X as kind of repository for my XML files, to perform
queries
over all files I use. Number and size of files are not a problem - it is some 100+ files , each one few hundreds lines. To be more specific - those are xml files of
IzPack
installation. I have loaded all my files into database and perform different kinds of queries there. It works great. Problem is - that after I have found desired data, I have to look at
which
file at which place it really is located. What I am doing now is - just
copy
into clipboard piece of resulting data and performing "search in files"
at
all my XML files directory. Finally - it works, but I am sure - there
have
to be better way to do it... Can I somehow modify my query to cause
base-X
to "say" me - where calculated result is located at original file system?