Some that I've used are XML::Simple, XML::Smart, XML::LibXML, XML::Twig. They all have their pros & cons; your best bet is to read the docs, pick a couple that look best suited to your needs, and give them a try.

Thanks for your response.
But what you have mentioned are Xml parsers (used to read a parser).
But i need to edit a XML file. I dont want to read it using a parser and write it to some other file using a XML writer. as that would certaily take time.
Do we have XML editors like i would read a XML and at a particular point add or remove the required tags.

These can do that. For example:

#!/usr/bin/perl
use strict;
use warnings;
use XML::Smart;
my $xmlstring = q{
<root>
 <member id="1" name="member1">
  <ips>
   <ip1>
    <from></from>
    <to></to>
   </ip1>
   <ip2>
    <from></from>
    <to></to>
   </ip2>
  </ips>
 </member>
</root>};
my $XML = XML::Smart->new($xmlstring);
$XML->{root}{member}{ips}{ip1}{from} = 'I am IPS IP1 From';
print $XML->data(noheader => 1);
__END__

<root>
  <member id="1" name="member1">
    <ips>
      <ip1>
        <from>I am IPS IP1 From</from>
        <to/>
      </ip1>
      <ip2>
        <from/>
        <to/>
      </ip2>
    </ips>
  </member>
</root>
[download]

Hi

Was just about to post a response when I saw your clarification re: keszler. However, I'm still not quite sure what you mean by an editor... Do you mean you want to open the file for a human to change, or you want to change the file without reading the whole thing into memory?

You can "edit" XML as a stream using some of the XML modules already mentioned by keszler. In particular I've used XML::Twig for large files. though I wouldn't say it was exactly fast. Depends how complex the changes are (you have to output to a tempfile and rename, as far as I know).

FalseVinylShrub

Assuming you're talking about using code to edit your XML file, if you try to avoid the XML parser part you'll only have to reinvent it yourself (poorly, as it's harder than you think to make an XML parser). Don't worry about how much time it takes: That's a premature optimization. Until you know it's too slow, don't worry about it. Also, there are some XML parsers available that are pretty fast, so it shouldn't be an issue anyway.

But that's now why I'm writing. What I really wanted to do was suggest that you change your XML a bit. You're using tags <ip1>, <ip2>, etc., in your <ips> section. You should just use <ip> -- if you want to number them, then give them a number or name attribute like this: <ip id="1"> or have another sub-node to give it a name:

<ips>
  <ip>
    <name>foo</name>
    <from></from>
    <to></to>
  </ip>
</ips>
[download]

There are several problems with dynamically changing your node tags. The most problematic is that any programs processing your XML documents are going to have to have some "magic" to interpret the document because you're trying to make a tag convey two pieces of information: (1) that this is an ip item, and (2) which ip item it is.