Twig is nice, as GrandFather says. XML::LibXML is quite nice too.

use strict;
use warnings;
use XML::LibXML;
my $parser = XML::LibXML->new();
# Uncomment next line to try it on a file.
# my $doc = $parser->parse_file( shift || die "Gimme an XML file!\n" )
+;
my $doc = $parser->parse_fh(\*DATA); # This is just for demo.
my @target = $doc->findnodes("//something");
my $whee = "a";
$_->setAttribute("else", $whee++) for @target;
print $doc->serialize(1), $/; # For this demo.
# $doc->toFile("./fixed.xml"); # For file writing.

__DATA__
<?xml version="1.0"?>
<root>
  <something/>
  <something/>
  <something/>
</root>

__END__
<?xml version="1.0"?>
<root>
  <something else="a"/>
  <something else="b"/>
  <something else="c"/>
</root>
[download]

There is only one - XML::Twig¹.

¹Well, there may be others, but that's generally reckoned to be the most comprehensive general purpose XML manipulating module for Perl.

Another option would be to use XML::SAX to construct a pipeline reading at one end, modifying as needed in the middle and writing out the new file at the end. That would be an excellent choice if your files could be too large to load into memory all at once.

-sam

It depends a lot on the complexity of the searches and the size of the XMLs. If the searches are complex and the parsed XML fits into memory then XML::LibXML might be best. If it doesn't fit in memory you'll need to use something else. XML::LibXML::SAX, XML::Twig, XML::Rules ...

It's better to have a look at several and find the one that fits the task and your brain.