Switching to Moo - adding command line parameters
Returning to the project after 3 weeks break was not easy. I ran the script again and it finished within a few seconds. I looked at the log file that only had two lines. What happened?
Moo-based code
Oh, right, I still have the code that will limit the processing to 20 messages. I remember, this already bothered me earlier that I have to change the code in order to do a limited run. So let's make that conditional on a command line parameter.
There is quite a nice way to do this with Moo and MooX::Options so we are going to do that. Turn the whole script into a Moo-based class and then use MooX::Options for the command line.
use Moo; use MooX::Options;
Replace process($path_to_dir); by
main->new_with_options->process($path_to_dir);
And change the first line for the process function from my ($dir) = @_; to the following code:
my ($self, $dir) = @_;
to accept both the instance object and the directory.
Then I ran the script again. It was still running and still finishing in a few seconds.
Command line parameters
The next step is to start really using MooX::Options for processing the command line options. Currently the script expects exactly one parameter, the path to the mail directory.
We remove the line that gets the parameter from the command line, my $path_to_dir = shift or die "Usage: $0 path/to/mail\n"; we won't need to do it ourselves any more.
Instead we add the declaration of the "path" option:
option path => (is => 'ro', required => 1, format => 's', doc => 'path/to/mail');
Now we don't have the $path_to_dir in the main body of the script so we stop passing it to the process() method:
main->new_with_options->process();
Also inside the process method we are now expecting only the instance object while what we got in the $dir variable earlier will be found in the path attribute of the object:
Replace
sub process { my ($self, $dir) = @_;
by this:
sub process { my ($self) = @_; my $dir = $self->path;
If we run the script as perl bin/mbox-indexer.pl /path/to/mail we will get the usage message as now we have to provide the --path command line option name and run the script like this: perl bin/mbox-indexer.pl --path /path/to/mail.
And the script still runs. Great!
Parameter for processing limit
Now the real thing. We add a command line option called --limit that will accept a number. If it is provided, we will stop processing the mailboxes after the given number of messages.
First thing is that we add a command line option which accepts and integer (format => 'i') and is not required.
option limit => (is => 'ro', required => 0, format => 'i', doc => 'limit number of messages to be processed');
Then in the main loop we replace the hard coded limit: exit if $count > 20;
by this code:
exit if defined $self->limit and $count > $self->limit;
We only compare $count with limit if there was a limit set.
Now if we run the script as above it will take a long time (almost 4 minutes for me) because it processes all the files. If we want to limit it to the first 20 we also need to pass --limit 20 on the command line.
Published on 2015-06-01