Running::Commentary(3) call system cleanly, with tracking messages

VERSION

This document describes Running::Commentary version 0.000005

SYNOPSIS


use Running::Commentary;
# Set a lexically scoped flag for all subsequent calls...
# (No announcements, if this flag set)
run_with -nomessage if !$verbose;
# Act like system(), only louder and cleaner...
run 'Resetting' => "rm -rf '$ROOT_DIR'"
or die "Couldn't reset";
# Act like system(), but croak() if the command fails...
run -critical, 'Building Makefile' => 'perl Makefile.PL';
# Calls to run() may be nested, to allow subtasks to be tracked...
run 'Running tests'
=> sub {
for my $file (@profiled_files) {
push @profiles, "$NAMING_ROOT/$file.out";
local $ENV{NYTPROF} = "file=$profiles[-1]";
run -nooutput, "Testing $file"
=> "perl -d:NYTProf $profiled_path/$file >& /dev/null";
}
};

DESCRIPTION

This module provides a single subroutine: "run()" which is designed to be a more informative and less error-prone replacement for the built-in "system()".

It also provides a compile-time keyword: "run_with" with which you can set lexically scoped default options for "run()".

INTERFACE

"run $MESSAGE => $SYSTEM_CMD;"
This acts like "system $SYSTEM_CMD", except that it returns true on success and false on failure, and it announces what it's doing. For example:

    run 'Resetting directories' => "rm -rf @STD_DIRS"

...would first output:

    Resetting directories...

...then execute the system command, and finish the message:

    Resetting directories...done

If the command failed for some reason, the completion would reflect the problem:

    Resetting directories...
    rm: tets: No such file or directory
    Resetting directories...exited with value 1

Or:

    Resetting directories...failed to execute: No such file or directory
"run $MESSAGE => sub {...};"
This form of the command expects a subroutine reference, rather than a string, as its second argument. Once again it prints the tracking message, then executes the subroutine, then prints the outcome.

The subroutine is run inside an "eval" block, so any exceptions it throws are intercepted, and reported as the outcome at the end of the tracking message. To have exceptions inside the subroutine propagate back out of the call to "run()", use the "-critical" option (see below).

For example:

    run 'Printing your data' => sub {
        for my $datum (@data) {
            say "    $datum->{key}: $datum->{value}";
        }
    }

Would output:

    Printing your data...
        Name: Fred
        Age: 28
        Score: 87
    Printing your data...done

You can also nest calls to "run()" using this form. For example:

    run 'Running your request' => sub {
        for my $cmd (split /\n/, $request) {
            run "Running '$cmd'" => $cmd;
        }
    }

Would produce:

    Running your request...
        Running 'rm source'...done
        Running 'rebuild_files'...done
        Running 'make test'.......done
    Running your request...done
"run $SYSTEM_CMD;"
"run sub {...};"
When called without a message, "run()" simply executes the system command or subroutine without printing any kind of progress message. In other words, it merely acts as a (quietly) better "system()".
"run_with @OPTIONS;"
The "run_with" keyword can be called with any of the options available to "run()" (see ``OPTIONS''). It takes the options given to it and makes them the default arguments to "run()" for the remainder of the current lexical scope.

For example, to cause any subsequent failed command to throw an exception...

    {
        run_with -critical;
        run "loading"     => $LOAD_CMD;
        run "checking"    => $CHECK_CMD;
        run "installing"  => $INSTALL_CMD;
        run "cleaning up" => $CLEANUP_CMD;
    }

...or to silence message printing on request:

    {
        run_with -nomessage if $opt{-quiet};
        run "loading"     => $LOAD_CMD;
        run "checking"    => $CHECK_CMD;
        run "installing"  => $INSTALL_CMD;
        run "cleaning up" => $CLEANUP_CMD;
    }

Note that "run_with" is a compile-time keyword, not a subroutine, so it should only be called as a statement (i.e. in void context).

OPTIONS

The following options can be included anywhere in the argument list of a call to "run()" or "run_with".
"-nomessage"
Run the command without printing the tracking message. Normally used as a conditional lexical option:

    run_with -nomessage if $opt{quiet};

The output of the actual system command is still printed (unless "-nooutput" or "-silent" is also specified)

"-showmessage"
Run the command, printing the tracking message. Useful to turn message printing back on inside a scope where "-nomessage" is already in effect.
"-nooutput"
Run the command without echoing any of its output. The tracking message is still printed (unless "-nomessage" or "-silent" is also specified)
"-showoutput"
Run the command, echoing any output. Useful to turn command echoing back on inside a scope where "-nooutput" is already in effect.
"-silent"
Identical to: "-nomessage, -nooutput"
"-showall"
Identical to: "-showmessage, -showoutput". Useful to override "-silent" in a nested scope.
"-critical"
Normally, if a call to "run()" fails, it simply returns "undef". However, if the "-critical" option is specified, any call to "run" that fails will immediately throw an exception.
"-nocritical"
Revert "run()" to returning "undef" on failure. Useful to override "-critical" in a nested scope.
"-dry"
Instead of executing the specified system command, just print it out. Useful for dry runs during development and testing.
"-colour => \%COLOUR_SPEC"
Specify the colours to be used for messages and output. Colours are specified as the values of the hash, with the keys indicating what purpose each colour is to be used for. For example:

    run_with -colour => {
        MESSAGE => 'white',          # Colour for tracking messages
        DONE    => 'bold cyan',      # Colour for success messages
        FAILED  => 'yellow on_red',  # Colour for failure messages
        OUTPUT  => 'clear'           # Colour for command output
    };

The colour specifications must be single strings, which are split on whitespace and then passed to the "Term::ANSIColor" module. If that module is not available, this option is silently ignored.

This option may also be spelled "-color".

"-nocolour"
Print all messages and output without any special colours.

This option may also be spelled "-nocolor".

ERROR HANDLING

On failure "run()" normally either returns "undef" or throws an exception (if "-critical" is specified).

However, "Running::Commentary" incorporates the "Lexical::Failure" module, so you can also request other failure responses for any particular scope, by passing a named argument when loading the module:

    # Report errors by confess()-ing...
    use Running::Commentary  fail => 'confess';
    # Report errors by returning a failure object...
    use Running::Commentary  fail => 'failobj';
    # Report errors by setting a flag variable...
    use Running::Commentary  fail => \$error;
    # Report errors by calling a subroutine...
    use Running::Commentary  fail => \&error_handler;

For details of the available options, see the documentation of "Lexical::Failure".

DIAGNOSTICS

"Bad argument to 'use Running::Commentary'"
The module accepts only one named argument:

    use Running::Commentary  'fail' => $fail_mode;

(see ``ERROR HANDLING'').

You apparently passed it something else. Or perhaps misspelt 'fail'?

"Useless call to run() with no command"
"run()" expects at least one argument (apart from any configuration options); namely, something to execute. That can be either a string containing a system command, or else a subroutine reference.

You didn't give it either of those, so the call to "run()" was superfluous.

Or, possibly, you wanted "run_with" instead.

CONFIGURATION AND ENVIRONMENT

Running::Commentary requires no configuration files or environment variables.

DEPENDENCIES

This module requires Perl v5.14 or later.

It also requires the modules: "Lexical::Failure", and "Keyword::Simple".

INCOMPATIBILITIES

None reported.

BUGS AND LIMITATIONS

No bugs have been reported.

Please report any bugs or feature requests to "[email protected]", or through the web interface at <http://rt.cpan.org>.

AUTHOR

Damian Conway "<[email protected]>"

LICENCE AND COPYRIGHT

Copyright (c) 2012, Damian Conway "<[email protected]>". All rights reserved.

This module is free software; you can redistribute it and/or modify it under the same terms as Perl itself. See perlartistic.

DISCLAIMER OF WARRANTY

BECAUSE THIS SOFTWARE IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE SOFTWARE, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE SOFTWARE ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE SOFTWARE IS WITH YOU. SHOULD THE SOFTWARE PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR, OR CORRECTION.

IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE SOFTWARE AS PERMITTED BY THE ABOVE LICENCE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE SOFTWARE (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE SOFTWARE TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.