Consider using XML::Twig instead of regexes to process the file. Easier to apply changes to a data element before creating the flat file rather a complete line afterwards.

#!/usr/bin/perl
use strict;
use XML::Twig;

my $xml = join '',<DATA>;
my $twig = XML::Twig->new( 
  twig_handlers => {'datafield' => \&datafield}
);          
$twig->parse( $xml );

sub datafield {
  my( $t, $e ) = @_;

  my %subfield = ();
  for my $elem ($e->children('subfield')){
    $subfield{$elem->att('code')} = $elem->text;
  }

  my @f = ();
  $f[0] = $e->att('tag');
  $f[1] = $e->att('ind1').$e->att('ind2');
  my @tmp;
  for (sort keys %subfield){
    push @tmp,$_.$subfield{$_};
  }
  $f[2] = join '|',@tmp;
  
  # change
  if ($subfield{'x'} =~ /^(isbn13|isbn)$/){
    $f[0] =~ s/024/020/;
  }
  
  # flat format for output
  printf ".%s. %s|%s\n",@f if ($f[2]); # skip blanks
}   

#.245. 14|aThe solar system|cChris Oxlade [Author] 
#.024. 3#|a9780750247092|xisbn13  
#.024. 3#|a0750247096|xisbn  
__DATA__
<collection>
<record>
  <datafield ind1="1" ind2="4" tag="245">
    <subfield code="a">The solar system</subfield>
    <subfield code="c">Chris Oxlade [Author]</subfield>
  </datafield>
</record>
<record>  
  <datafield ind1="3" ind2="#" tag="024">
    <subfield code="a">a9780750247092</subfield>
    <subfield code="x">isbn13</subfield>
  </datafield>
<record>
</record>
  <datafield ind1="3" ind2="#" tag="024">
    <subfield code="a">a0750247096</subfield>
    <subfield code="x">isbn</subfield>
  </datafield>
</record>
<record>
  <datafield ind1="3" ind2="#" tag="024">
  </datafield>
</record>
</collection>
[download]