This is the script I've been using in order to replace creationd id. perl -pi.bak -e 's{creationid=".+?"}{creationid="Simon Simonsen"}g;' file.xml

This works for me, if file.xml contains only the string you provided above. Can you try to whittle your ~1 GB file down to a simpler test case that fails, e.g. by bisecting it? Doing so may prove instructive in identifying the problem; alternatively, it'll make it easier for us monks to help you if we've got both an "it works" case and a "it stops working if I do this" case.

s{creationid="[^"]*"}{creationid="Simon Simonsen"}g
[download]

If you need to change the structure of XML file in a more complex way, you might need to switch to XML::LibXML::Reader, an XML pull-parser.

Thank you choroba and also many thanks to AppleFritter

I do not have to change the structure of the file, it is just about replacing some strings. Chances are this has to be put into a 'proper' script at some stage (I will surely try out the XPath variant)

As for the file, I can post a small snippet, which I had to modify. I changed the content but the structure is the same. Sorry for not having initiallly posted the snippet.

<tu creationdate="12345Z" changedate="12345" creationid="John Doe" cha
+ngeid="John DOE" srclang="en-US">
        <prop type="user-defined">server:xyz</prop>
        <prop type="user-defined">quality:100</prop>
        <prop type="user-defined">storing mode:1</prop>
        <prop type="user-defined">uncontrolled:0</prop>
        <prop type="user-defined">lastuser:unknown</prop>
        <prop type="user-defined">context left:1;2;3;4;5;6;7;8;9;10;11
+;12;13;14;15;16;17;18;19;20</prop>
        <prop type="user-defined">context right:21;22;23;24;25;26;27;2
+8;29;30;31;32;33;34;35;36;37;38;39;40</prop>
        <prop type="user-defined">a lot of junk with letters and &quot
+;90365AF9&quot</prop>
        <prop type="Guid::xxx">123456789</prop>
        <prop type="Att::xxx">name_of_someone</prop>^M
        <tuv xml:lang="en-US"><seg>a lot of junk with letters and &quo
+t;90365AF9&quot<bpt i="1"></bpt>text<ept i="1">&lt;/cf&gt;</ept></seg
+></tuv>
        <tuv xml:lang="en-US"><seg>a lot of junk with letters and &quo
+t;90365AF9&quot<bpt i="1"></ept></seg></tuv>
</tu>
[download]

Thanks a mil for helping me out. Really helps beginners to move on. Looking forward to your replies.

Kind regards C.

Thanks for sharing this piece of data -- however, your oneliner is still working just fine there for me. I understand that you may not be in a position to share confidential data, but on the other hand, I'm sure you'll understand it's a bit difficult to diagnose a problem without a test case showing the problem.

I second what other monks have suggested: if regular expressions aren't working here for whatever reason, a proper XML wrangling module may be the way to go. And XML::Twig is specifically intended for working with very large XML files that don't fit into memory, so that's my recommendation as well.

Works on your example XML (once the structure is corrected):

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

XML::Twig->new(
    keep_spaces => 1,
    twig_print_outside_roots => 1,
    twig_roots => {
        tu => sub {
            my ($twig, $elt) = @_;
            $elt->set_att('creationid','Simon Simonsen');
            $elt->print;
        }
    },
)->parsefile('1097276.xml');
[download]

Prints to STDOUT, so redirect the output wherever you like (perl 1097276.pl >output.xml)

Thank you Anonymous Monk for the code snippet.

Your help is much appreciated. The code works fine on the command line.

However I've been getting this nasty error message.

Wide character in print at /usr/lib/perl5/site_perl/5.14/XML/Twig.pm l
+ine 8403.
Wide character in print at /usr/lib/perl5/site_perl/5.14/XML/Twig.pm l
+ine 8403.
Wide character in print at /usr/lib/perl5/site_perl/5.14/XML/Twig.pm l
+ine 8403.
[download]

I think this is due to the encoding.

Is there any chance I could force utf8 output along the lines of

use warnings;
use strict;
use XML::Twig;
use File::Slurp;
use utf8;
use File::Slurp qw(read_file write_file);
    

my $infile = shift; 
my $filename = $ARGV[2];

XML::Twig->new(
    keep_spaces => 1,
    twig_print_outside_roots => 1,
    twig_roots => {
        tu => sub {
            my ($twig, $elt) = @_;
            $elt->set_att('creationid','Simon Simonsen');
            $elt->print;
        }
    },
)->parsefile( $infile );

write_file $filename, {binmode => ':utf8'}, $infile;
[download]

Thanks in advance for encourageing comments

Kind regards and many thanks

C

Since XML::Twig is writing to STDOUT, I'm not sure what your write_file is supposed to be doing...

Encodings are covered in the XML::Twig docs. Here's some code that works for me:

use warnings;
use strict;
use XML::Twig;

open my $ofh, '>:utf8', '1097276_out.xml' or die $!;
XML::Twig->new(
    keep_spaces => 1,
    twig_print_outside_roots => $ofh,
    twig_roots => {
        tu => sub {
            my ($twig, $elt) = @_;
            $elt->set_att('creationid',"Simon Sim\xF6nsen");
            $elt->print($ofh);
        }
    },
)->parsefile('1097276.xml');
close $ofh;
[download]

Also have a look at the aptly-named output_encoding option in XML::Twig.