How to fasten the XML searching compared to database

Question

Hi,
i'm putting the whole scenario here:
It will be so helpful if u can help me.
-------------------the doubt----------------------------
i'm having an xml in the following format
<LOV>
<RECORD NO="1">
<FIELD NAME="EMPNO">E1001</EMPNO>
<FIELD NAME="EMPNAME">MR. X</EMPNO>
<FIELD NAME="TEAM">PRODUCT TEAM</EMPNO>
</RECORD>
<RECORD NO="2">
<FIELD NAME="EMPNO">E1002</EMPNO>
<FIELD NAME="EMPNAME">MR. Y</EMPNO>
<FIELD NAME="TEAM">TECHNOLOGY TEAM</EMPNO>
</RECORD>
</LOV>
like this there will be minimum 5000 records and what i need is to search this xml(using DOM) by specifying a condition like
EMPNO=E1001
or EMPNAME=MR.X or EMPNO=E1* etc etc.
i did this with the help of NodeIterator.
but the problem is :
this is done to avoid database interaction.
and this s taking 2-3 secs to query and once it is stored as an object it s taking .3 secs.
but when i tried to query the same no and sort of records straight from a database it s taking only .1 sec everytime.
Now my question is : Is it possible to use any other method in xml parsing so that i can make the xml quering much faster than database query.
pls help if possible
Ranjith.M

Answer 1

this is not pure JAXP but a behind the scene hack: use the org.apache.xpath.CachedXPathAPI and run XPath queries on your DOM.
simple sample:
// getMyXpathAPI() returns a singleton CachedXPathAPI
Node node = getMyXpathAPI().selectSingleNode(doc, xpath);