Update: Fixed a typo in the parenthesed sentence about having to store the data in an array.

#!/usr/bin/perl --
use strict;
use warnings;
use XML::LibXML 1.70; ## for load_html/load_xml/location

my $xml = q{<contrib-group>
  <contrib contrib-type="author">
.
some code
.
<x>, </x></contrib>
  <contrib contrib-type="author">
.
some code
.
<x>, </x></contrib>
  <contrib contrib-type="author">
.
some code
.
<x> and </x></contrib>
  <contrib contrib-type="author">
.
some code
.
</contrib>
</contrib-group>
};

my $dom = XML::LibXML->new(qw/ recover 2 /)->load_xml(
#~     location => $filepath_or_http,
    string => $xml,
);

print $dom->find(q{   //contrib/x[ contains( . , ' and ' ) ]      })->
+shift->nodePath, "\n";

if( my $cx = $dom->findnodes(q{    //contrib/x    })->pop ){
    print $cx->nodePath, "\n";
    print "$cx\n";
    my $tx = $cx->textContent;
    $tx =~ s/ and / , /;
    $cx->removeChildNodes;
    $cx->appendText( $tx ); 
    print "$cx\n\n\n";
}

print "$dom\n";
__END__
/contrib-group/contrib[3]/x
/contrib-group/contrib[3]/x
<x> and </x>
<x> , </x>
[download]

IMHO, “an XML-parsing module” is the only way to go, and XML::LibXML is an excellent choice, since it employs a binary module that is quite likely to be the same one that was used to produce the file.   A very real consideration when dealing with XML files “manually” is that, sooner or sooner, the line-by-line format of the file will vary from whatever your hand-rolled program was built to expect.   (And it is most likely to do so at 4:45 PM on the evening of your child’s piano recital ... don’t miss it, Dad ...)

Therefore:   use a module/library.   Period.   There are actually several good ones to choose from ... XML::Twig is another good war-horse aimed especially at b-i-g files.   And leave us not forget XML::Simple.   Treat the XML file as an abstract thing, which the module magically knows how to manipulate for you.   You will be rewarded with a durable, long-lasting solution to the problem that won’t have to be revisited, no matter what the next XML file looks like.

“an XML-parsing module” is the only way to go ...

Yeah, right, I agree in principle, but what are you gonna do in the event that when this otherwise very fine module chokes on an error in the XML syntax? This is not a rhetorical question (sh*t happens, you know), I have seen that happening several times, although I don't remember precisely which module it was (and also with HTML modules, BTW).