Question on DOM Traversal

[Fran Orzel <fm...@us.ibm.com>]

Given an XML file that looks as follows: 

<?xml version="1.0" encoding="UTF-8" standalone="no" ?> 
 <vpdc date="Feb 18 2009" time="17:07:11" version="1.0">
  <some_data info1="E12" info2="MNXK4" info3="1997" > 
       <other_data>this is the data</other_data>
       <other_data1> some other data </other_data1>
    </some_data> 
 <some_records>
 <record1 data1="CAGE" data2="LION" data3="DEFAULT">
  <time_of_day>20090205144231</time_of_day> 
  </record1>
 <record2 data1="BOX" data2="ADOG" data3="45454545">
  <time_of_day>20090205144231</time_of_day> 
  </record2>
<recordn data1="BOWL" data2="ACAT" data3="31313131">
  <time_of_day>20090205144231</time_of_day> 
  </recordn>
</records>
</vpdc>



I want to parse this file and put save the data into linked lists such 
that I have an entry for each major record (<vpdc>, <system>, <record1>, 
<record2> and <recordn>).  For each of these I want to have a linked list 
for the "data" associated with each record.  For example: 

LList1 = 
<vpdc>  with pointer to      linked list of data for date, time and 
version 
<some_data> with pointer to linked list of data for info1, info2, info3, 
other_data and other_data1
<record1> with pointer to linked list of data for data1, data2, data3 and 
time_of_day
<record2> with pointer to linked list of data for data1, data2, data3 and 
time_of_day
<record3> with pointer to linked list of data for data1, data2, data3 and 
time_of_day

The problem I am having is that I do not know how many other_data and 
time_of_day possible entries there may be.  I have looked at the 
DOMCount.cpp sample and wanted to so something similar to this since in 
the future I may have even more "subtrees" of data then I do now....... 

When I implemented a version of the DOMCOunt code tyring to get the 
non-attribute data for the nodes, I get a TEXT_NODE for each of the 
records with no value attached to it, so I can't use the existence of a 
text-node to clearly indicate that I have some data I want to keep (since 
I want to ignore those values).  I also tried a TreeWalker but that would 
mean recognizing that I have to go down additional subtrees and I may not 
know how many of them I have....

Any hints, samples, ideas ?????

Thanks, 

Fran