The London Perl and Raku Workshop takes place on 26th Oct 2024. If your company depends on Perl, please consider sponsoring and/or attending.


IO::Capture::Stdout - Capture any output sent to STDOUT


    # Generic example (Just to give the overall view)
    use IO::Capture::Stdout;

    $capture = IO::Capture::Stdout->new();

    $capture->start();          # STDOUT Output captured
    print STDOUT "Test Line One\n";
    print STDOUT "Test Line Two\n";
    print STDOUT "Test Line Three\n";
    $capture->stop();           # STDOUT output sent to wherever it was before 'start'

    # In 'scalar context' returns next line
    $line = $capture->read;
    print "$line";         # prints "Test Line One"

    $line = $capture->read;
    print "$line";         # prints "Test Line Two"

    # move line pointer to line 1

    $line = $capture->read;
    print "$line";         # prints "Test Line One"

    # Find out current line number
    $current_line_position = $capture->line_pointer;

    # In 'List Context' return an array(list)
    @all_lines = $capture->read;

    # More useful example 1 - "Using in module tests"
    #  Note: If you don't want to make users install 
    #        the IO::Capture module just for your tests,
    #        you can just install in the t/lib directory
    #        of your module and use the lib pragma in  
    #        your tests. 

    use lib "t/lib";
    use IO::Capture::Stdout;

    use Test::More;

    my $capture =  IO::Capture::Stdout->new;

    # execute with a bad parameter to make sure get
    # an error.

    ok( ! $test("Bad Parameter") );



The module IO::Capture::Stdout, is derived from the abstract class IO::Capture. See IO::Capture. The purpose of the module (as the name suggests) is to capture any output sent to STDOUT. After the capture is stopped, the STDOUT filehandle will be reset to the previous location. E.g., If previously redirected to a file, when IO::Capture->stop is called, output will start going into that file again.

Note: This module won't work with the perl function, system(), or any other operation involving a fork(). If you want to capture the output from a system command, it is faster to use open() or back-ticks.

       my $output = `/usr/sbin/ls -l 2>&1`;



  • Creates a new capture object.

  • An object can be reused as needed, so will only need to do one of these.

    • Be aware, any data previously captured will be discarded if a new capture session is started.


  • Start capturing data into the IO::Capture Object.

  • Can not be called on an object that is already capturing.

  • Can not be called while STDOUT tied to an object.

  • undef will be returned on an error.


  • Stop capturing data and point STDOUT back to it's previous output location I.e., untie STDOUT


  • In Scalar Context

    • Lines are read from the buffer at the position of the line_pointer, and the pointer is incremented by one.

          $next_line = $capture->read;
  • In List Context

    • The array is returned. The line_pointer is not affected.

          @buffer = $capture->read;
  • Data lines are returned exactly as they were captured. You may want to use chomp on them if you don't want the end of line character(s)

        while (my $line = $capture->read) {
            chomp $line;
            $cat_line = join '', $cat_line, $line;


  • Reads or sets the line_pointer.

        my $current_line = $capture->line_pointer;


Adding Features

If you would like to sub-class this module to add a feature (method) or two, here is a couple of easy steps. Also see IO::Capture::Overview.

  1. Give your package a name

        package MyPackage;
  2. Use this IO::Capture::Stdout as your base class like this:

        package MyPackage;
        use base qw/IO::Capture::Stdout/;
  3. Add your new method like this

        package MyPackage;
        use base qw/IO::Capture::Stdout/;
        sub grep {
            my $self = shift;
            for $line (

See Also





Mark Reynolds

Jon Morgan


Copyright (c) 2003, Mark Reynolds. All Rights Reserved. This module is free software. It may be used, redistributed and/or modified under the same terms as Perl itself.