& XSLT >         I tried using XML::XPath::XMLParser and XML::XPath modules to 
>  parse an xml files. Its really works good and still working well only 
>  the problem is its taking so much time and resources. The xml 
>  files are 
>  approximately 20MB each. My CPU load jump to 16 and the server almost 
>  don't want to accept keyboard input anymore.

DOM-style parsing (which XML::XPath does) reads the entire file into memory.
You would be better off using XML::Twig, which has optimizations for larger
files. It will stream to what you are interested in, then build little
mini-trees for convenient processing.

A simple example:

  my $twig = XML::Twig-> new( twig_handlers => 
         {
          'Object[@Class="Pens"]' =>  sub { $_->print }
         }
  );
  $twig-> parsefile('doc.xml');

I think you'll find this to be much faster for large files.

-- 
Mark Thomas 
Internet Systems Architect
_______________________________________
BAE SYSTEMS Information Technology 
2525 Network Place
Herndon, VA  20171  USA 




>  -----Original Message-----
>  From: perl-xml-bounces@[...].com 
>  [mailto:perl-xml-bounces@[...].com] On Behalf 
>  Of Joseph C. Bautista
>  Sent: Tuesday, September 20, 2005 7:39 AM
>  To: Joseph C. Bautista
>  Cc: perl-xml
>  Subject: Re: Getting XML Nodes based on attributes value
>  
>  Hi All,
>  
>         I tried using XML::XPath::XMLParser and XML::XPath modules to 
>  parse an xml files. Its really works good and still working well only 
>  the problem is its taking so much time and resources. The xml 
>  files are 
>  approximately 20MB each. My CPU load jump to 16 and the server almost 
>  don't want to accept keyboard input anymore.
>  
>         My server is P4 1.7GHz with 512MB RAM running in linux 
>  redhat 9.
>  
>         Is there any way to do this in more optimize and faster way?
>  
>  
>         Thank you and all sugestions/ideas are highly appreciated.
>  
>  Joseph
>  
>  >>> Hi All,
>  >>>
>  >>>    I was wondering if there's a module that i can used to 
>  get only 
>  >>> the nodes in an XML file based on the attributes value?
>  >>>
>  >>>   Example XML is:
>  >>>
>  >>>    <Things>
>  >>>       <Object Class="Pens" Location="Room">
>  >>>             <p name="length">4<\p>
>  >>>             <p name="thickness">.5</p>
>  >>>             <p name="color">yellow</p>
>  >>>       </Object>
>  >>>       <Object Class="Papers" Location="Room">
>  >>>             <p name="length">9<\p>
>  >>>             <p name="thickness">2</p>
>  >>>             <p name="color">yellow</p>
>  >>>       </Object>
>  >>>       <Object Class="Pens" Location="Sala">
>  >>>             <p name="length">4<\p>
>  >>>             <p name="thickness">.5</p>
>  >>>             <p name="color">Black</p>
>  >>>       </Object>
>  >>>       <Object Class="Papers" Location="Sala">
>  >>>             <p name="length">9<\p>
>  >>>             <p name="thickness">3</p>
>  >>>             <p name="color">white</p>
>  >>>       </Object>
>  >>>   </Things>
>  >>>
>  >>>    If I want to retrieve only the objects with class 
>  "Pens" then i 
>  >>> would just execute something like
>  >>>
>  >>>          $xml->node("Object", "Class" => "Pens")
>  >>>      and the module(s) will give me a hashref with two 
>  values. All 
>  >>> objects, including its parameters, with class "pens" (1 in "Sala" 
>  >>> and 1 in "Room")?
>  >>>
>  >>>
>  >>>    Thanks...
>  >>>
>  >>> Br,
>  >>> Joseph
>  >>>
>  >>>
>  >>>
>  >>
>  >> _______________________________________________
>  >> Perl-XML mailing list
>  >> Perl-XML@[...].com
>  >> To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
>  >>
>  >
>  >
>  
>  
>  -- 
>  This message has been scanned for viruses and dangerous 
>  content by host-center.net and is believed to be clean.
>  
>  _______________________________________________
>  Perl-XML mailing list
>  Perl-XML@[...].com
>  To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
>  
>  

_______________________________________________
Perl-XML mailing list
Perl-XML@[...].com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs