[jdom-interest] Regarding XPath

Soumen Das Das Soumen at silchar.com
Tue Apr 17 21:46:55 PDT 2001

See the main problem of using xml file as configuration file is that, for every data value pair that I need, I've to load the whole xml file and then by using XPath I've to iterate. And XPath only takes Document as argument. 
Is there any other way to iterate through the DOM tree. Right now I'm using DOMParser for that. 
To Create xml file for that DataDictionary is simple, but to retrieve value from the xml-file is costly.

--- James Sinnamon <sinnamon at usq.edu.au>
> wrote:
>Thanks for your response.   Altough the query did not originate from me,
>I found your response helpful and interesting. If you have a moment, Can you tell me:
>Markus Bernhardt wrote:
>> James Sinnamon wrote:
>> > (I think that this should have been set to the list, rather than
>> > to me, personally, or else I have an undeserved reputation for being
>> > a guru of JDOM - james )
>> >
>> > -------- Original Message --------
>> > From: "Soumen Das" <soumend at mindtree.com>
>>  And performance wise which one will be better ?
>> The runtime performance of a xml based solution is really hard to tell.
>> At the one hand you might have an increased size of your data because of the tags.
>> At the other hand it is possible that the size will decrease, because you have a
>> tree.
>> We had for example a data source with 50 million data sets and 40 GB size.
>> After using xml we had 22 million sets and 24 GB size.
>Is this with or without any compression?
>I ask because it seems plausible to me  that
>the DBMS's of the future may simply be XML document servers.  Where the amount of
>data is realtively small a database may simply be the XML files served from a
>or some other similar type of server.  Where the volume gets larger and larger, say
>approaching the
>sort of databases need to keep accounts, or other types of records, for a larger
>or government departments, then some kind of special servers using compression and
>decompression protocols could be used.
>Possibly the same sort of systems which were used to manage the older IMS style
>hierarchical databases could be used in these special 'XML database' servers.
>(I am aware of the fact there are criticisms of Relational style databases, and
>the older style hierarchical databases had a lot going for them.)
>Can you say if you think that these ideas have any merit?  Is it possible that
>somewhere in the world, someone may be acting on such ideas?
>(maybe www.xmldb.org)
>> The main advantage of xml I see is the performance of your development team.
>The more I look at it the more it seems to be an inherently natural way to store
>Thank you.
>James Sinnamon  sinnamon at usq.edu.au
>ph 07 46315566,  0412 319669
>PO Box 517 Darling Heights QLD 4350

Win $10 million in freelotto lottery!! Click and Try!!

More information about the jdom-interest mailing list