XML structure. Need advice...

Hello

I m trying to build a word processor like Windows application.

One of the main features of this application is AutoSave. In hope of 
reducing the time taken for the AutoSave, i was planning to save only 
the modified sections.

So I structured my XML files like below.

//----------------------------------------------

Page01.xml
<page id=”3464” modified_date=”2004/04/01”>
   <block id=”213”>
     <p>P4567.xml</p>
     <p>P2345.xml</p>
     <p>P6734.xml</p>
   </block>
   <block id=”156”>
     <p>P4567.xml</p>
     <p>P2345.xml</p>
     <p>P6734.xml</p>
   </block>
</page>
//----------------------------------------------

P4567.xml
<p modified_date=”2004/04/01”>
   <text>blah blah blah</text>
   <other_data/>
</p>
//----------------------------------------------

I have been reading some articles and posts on the net, trying to see if 
i m on the right track. But couldn't find anything that could tell me so...

Here are some of the questions that has been raised:
1) Will i lose performance on loading the page, because of this structure?
2) Will i gain performance on saving parts of the page by doing this?
3) If there are serveral more index files (like Page01.xml) that needs 
to be updated when paragraph file (i.e. P4567.xml) gets updated. Will it 
generate too much complexty?
4) What type of parser should i use for this kind of structure?
5) If i want my application to rebuild the page file from paragraph 
files in case of corruption. What other attributes or data should i 
store in paragraph files? Or do i need to think of different structure?

I will really appreciate any guide from you.

Regards,

Jeapil

-- 
Programer - SplashMedia
http://splashmedia.co.nz