momo33 has asked for the wisdom of the Perl Monks concerning the following question:

Hello Monks,

I would like to manipulate a cpio archive. CPAN has http://search.cpan.org/~pixel/Archive-Cpio-0.07/ This looks just what I need. Unfortunately, it does not what I expect.

Im my cpio archive test.cpio are three files: test1, test2, and test3

use Archive::Cpio;
my $cpio = Archive::Cpio->new;
$cpio->read('test.cpio');
$cpio->remove('test1');
$cio->write('newtest.cpio');

[download]

Unfortunately, test.cpio is identical to newtest.cpio; I want test1 to be gone. Is this a bug or my mistake?

One step further, I would like to insert a new file: new1. The docs give me $cpio->add_data($filename, $data, $opthashref).
$cpio->add_data('new1',$data, ???) did not do it. In addition, I have no idea about $opthashref. Any suggestions?

Thank you.

Replies are listed 'Best First'.
Re: manipulating cpio archive
by Corion (Patriarch) on Jan 16, 2011 at 20:29 UTC

    Seeing that Archive::Cpio::read is implemented as

    sub read {
    my ($cpio, $file) = @_;
    
    my $IN;
    if (ref $file) {
    $IN = $file;
    } else {
    open($IN, '<', $file) or die "can't open $file: $!\n";
    }

    read_with_handler($cpio, $IN, sub { 
        my ($e) = @_;
    push @{$cpio->{list}}, $e;
    });

    [download]

    and Archive::Cpio::remove is implemented as

    sub remove {
    my ($cpio, @filenames) = @_;
    $cpio->{list} or die "can't remove from nothing\n";

    my %filenames = map { $_ => 1 } @filenames;

    @{$cpio->{list}} = grep { !$filenames{$_} } @{$cpio->{list}};
}

    [download]

    I would assume that there is no file test1 in your cpio archive. Maybe there is a directory name prepended? Have a look at what information Archive::Cpio actually reads from your archive, and/or what your archive actually contains.

[reply]
[d/l]
[select]
      Thank you for your reply.
      I tried it again, and made better notes. My archive is td.cpio: 
      $ cpio --list -i <td.cpio
.
t2
t1
t3
1 block
$

      [download]
      My script is 
      #!/opt/ActivePerl-5.12/bin/perl
use Archive::Cpio;
my $cpio = Archive::Cpio->new;
$cpio->read('td.cpio');
$cpio->remove('t1');
$cpio->write('tdnew.cpio');

      [download]
      The output tdnew.cpio is identical to td.cpio

      I think the answer is in the Archive::Cpio code. The entry should be removed in

      sub remove {
    my ($cpio, @filenames) = @_;
    $cpio->{list} or die "can't remove from nothing\n";
    my %filenames = map { $_ => 1 } @filenames;
    @{$cpio->{list}} = grep { !$filenames{$_} } @{$cpio->{list}};
}

      [download]
      I cannot decipher that last line but I do not like it (I may be completely wrong).

      How can I get this right? Thank you.
[reply]
[d/l]
[select]

        I would now look inside the Archive::Cpio object (using Data::Dumper) to see what it contains, before and after your modification.

[reply]

Back to Seekers of Perl Wisdom