See method removeAttribute() in XML::LibXML::Element.

OK, got a result, but not easily. I seem to be missing how to descend an xml tree and pick out the element I want to operate on. What I have is:

#!/usr/bin/perl

use strict;
use warnings;
use XML::LibXML;

die "Usage: $0 filename \n";
   unless ( @ARGV > 0 );

my $xml_file = shift;
my $xml      = XML::LibXML->new;
my $dom      = $xml->parse_file( $xml_file );
my $root     = $dom->documentElement;

my @logins = grep { $_->nodeType == XML_ELEMENT_NODE } $root->childNod
+es;
foreach my $login ( @logins ) {
   my @user_infos = grep { $_->nodeType == XML_ELEMENT_NODE } $login->
+childNodes;
   foreach my $user_info ( @user_infos ) {
      my @add_users = grep { $_->nodeType == XML_ELEMENT_NODE } $user_
+info->childNodes;
      foreach my $add_user ( @add_users ) {
         $add-user->removeAttribute( "USER_NAME" );
         $add-user->removeAttribute( "PASSWORD" );
      }
   }
}
my $output = $dom->toString(0);
$output =~ s/(?<=\n)\s*\n//g;
open ( my $FH, '>', 'newilo2' ) or die "Could not open file newilo2 $!
+";
print $FH $output;
close $FH;
[download]

I tried using findnode, but I get an error about perl not being able to find that method. So, thanks for the pointer! It at least is getting me the results I want. :)

No need to do a nested descent through the tree when you know in advance which nodes you want to alter. This revised version of your code does the trick for your sample data:

#!/usr/bin/perl

use strict;
use warnings;
use XML::LibXML;

die "Usage: $0 filename \n"
   unless ( @ARGV > 0 );

my $xml_file = shift;
my $xml      = XML::LibXML->new;
my $dom      = $xml->parse_file( $xml_file );

foreach my $add_user ( $dom->getElementsByTagName('ADD_USER') ) {
    $add_user->removeAttribute( "USER_NAME" );
    $add_user->removeAttribute( "PASSWORD" );
}

my $output = $dom->toString(0);
$output =~ s/(?<=\n)\s*\n//g;
open ( my $FH, '>', 'newilo2' ) or die "Could not open file newilo2 $!
+";
print $FH $output;
close $FH;
[download]

XML::LibXML is pretty powerful, but it's a steep old learning curve. If you do a lot of XML work it would pay to immerse yourself in the documentation for a while to let it all sink in. Good luck.

Edit: better choice of verb

use strict;
use XML::Rules;

my $filter = XML::Rules->new(style => 'filter', rules => {
    ADD_USER => sub {
        my ($tag, $attrs) = @_;
        delete $attrs->{USER_NAME};
        delete $attrs->{PASSWORD};
        return $tag => $attrs;
    }
});

$filter->filter(\*DATA);

__DATA__
<RIBCL VERSION="2.1">
 <LOGIN USER_LOGIN="Administrator">
  <USER_INFO MODE="write">
  <ADD_USER USER_NAME="John Smith" USER_LOGIN="jsmith" PASSWORD="%user
+_password%">
  </ADD_USER>
  </USER_INFO>
 </LOGIN>
</RIBCL>
[download]

You can of course read from and write to other places than DATA and STDOUT and the good thing is that even if the file is huuuuge, the memory footprint is constant.

If you want to change the tag name as well, just change the return $tag => $attrs; to return 'DEL_USER' => $attrs;