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


Test::Command::Simple - Test external commands (nearly) as easily as loaded modules.


Version 0.05


    use Test::Command::Simple;

    run('echo', 'has this output'); # only tests that the command can be started, not checking rc
    is(rc,0,'Returned successfully')
    like(stdout,qr/has this output/,'Testing stdout');
    is(length stderr, 0,'No stderr');


This test module is intended to simplify testing of external commands. It does so by running the command under IPC::Open3, closing the stdin immediately, and reading everything from the command's stdout and stderr. It then makes the output available to be tested.

It is not (yet?) as feature-rich as Test::Cmd, but I think the interface to this is much simpler. Tests also plug directly into the Test::Builder framework, which plays nice with Test::More.

As compared to Test::Command, this module is simpler, relying on the user to feed rc, stdout, and stderr to the appropriate other tests, presumably in Test::More, but not necessarily. This makes it possible, for example, to test line 3 of the output:

    my (undef, undef, $line) = split /\r?\n/, stdout;
    is($line, 'This is the third line', 'Test the third line');

While this is possible to do with Test::Command's stdout_like, some regex's can get very awkward, and it becomes better to do this in multiple steps.

Also, Test::Command saves stdout and stderr to files. That has an advantage when you're saving a lot of text. However, this module prefers to slurp everything in using IPC::Open3, IO::Select, and sysread. Most of the time, commands being tested do not produce significant amounts of output, so there becomes no reason to use temporary files and involve the disk at all.



Runs the given command. It will return when the command is done.

This will also reinitialise all of the states for stdout, stderr, and rc. If you need to keep the values of a previous run() after a later one, you will need to store it. This should be mostly pretty rare.

Counts as one test: whether the IPC::Open3 call to open3 succeeded. That is not returned in a meaningful way to the user, though. To check if that's the case for purposes of SKIPping, rc will be set to -1.


Returns the last run's stdout


Returns the last run's stderr


Returns the last run's full $?, suitable for passing to POSIX's :sys_wait_h macros (WIFEXITED, WEXITSTATUS, etc.)


Returns the exit status of the last run


Shortcut for checking that the return from a command is 0. Will still set stdout and stderr for further testing.

If the first parameter is an integer 0-255, then that is the expected return code instead. Remember: $? has both a return code (0-255) and a reason for exit embedded. This function must make the assumption that you want a "normal" exit only. If any signal is given, this will treat that as a failure.

Note that this becomes three tests: one that IPC::Open3 could create the subprocess with the command, the next is the test that the process exited normally, and the last is the test of the rc.


Darin McBride, <dmcbride at>


Please report any bugs or feature requests to bug-test-command at, or through the web interface at I will be notified, and then you'll automatically be notified of progress on your bug as I make changes.


You can find documentation for this module with the perldoc command.

    perldoc Test::Command::Simple

You can also look for information at:



Copyright 2010 Darin McBride.

This program is free software; you can redistribute it and/or modify it under the terms of either: the GNU General Public License as published by the Free Software Foundation; or the Artistic License.

See for more information.