1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
|
---
layout: post
category: perl
title: Modules I like Getopt::Long and MooseX::Getopt
---
h3. Getopt::Long
"Getopt::long":http://search.cpan.org/perldoc?Getopt::Long is a useful module to parse command line arguements.
A basic usage is something like this:
{% highlight perl %}
#!/usr/bin/perl -w
use strict;
use YAML::Syck;
use Getopt::Long;
GetOptions('config=s' => \my $cfg_file,);
my $config = LoadFile $cfg_file
{% endhighlight %}
In *GetOptions*, we require a value for config with *config=s*. If we wante an integer, we replace 's' with 'i', and for a floating point, with 'f'.
Call your script :
{% highlight bash %}
script.pl --config=file.yml #this one works
script.pl --config file.yml #this one too!
script.pl -c file.yml #and this one too
{% endhighlight %}
The three syntaxes are understood.
A good practices is to combine this module with "Pod::Usage":http://search.cpan.org/perldoc?Pod::Usage. Let's do some modifications on the example:
{% highlight perl %}
#!/usr/bin/perl -w
use strict;
use YAML::Syck;
use Getopt::Long;
use Pod::Usage;
GetOptions('config=s' => \my $cfg_file,) or pod2usage(2);
pod2usage(2) unless @ARGV > 0;
my $config = LoadFile $cfg_file
__END__
=head1 NAME
uberscript
=head1 SYNOPSIS
uberscript [options]
Options:
--config config file
=head1 Options
=over 4
=item B<config>
Path to the config file
{% endhighlight %}
then
{% highlight bash %}
$ perl uberscript
Usage:
uberscript [options]
Options:
--config config file
{% endhighlight %}
From now if we call our script without argument, the POD will be printed on STDIN.
h3. MooseX::Getopt
"MooseX::Getopt":http://search.cpan.org/perldoc?MooseX::Getopt) is a Role that add a <code>new_with_options</code> to your object. We create a basic Object :
{% highlight perl %}
package OurShinyObject;
use Moose;
with qw/MooseX::Getopt/;
has 'config' => (isa => 'Str', is => 'ro', required => 1);
has 'context' => (
isa => 'HashRef',
is => 'rw',
lazy => 1,
traits => ['NoGetopt'],
default => sub { LoadFile shift->config }
);
...
{% endhighlight %}
create a script to call this object
{% highlight perl %}
#!/usr/bin/perl -w
use strict;
use OurShinyObject;
my $obj = OurShinyObject->new_from_options();
{% endhighlight %}
bc. script.pl --config file.yml
The role will set our attribute **context** using the value from the argument set on the command line.
The
{% highlight perl %}
traits => ['NoGetopt']
{% endhighlight %}
indicate that this attributes will be not be read from the command line. An alternate way to do this is to prefix the attributes with *_*.
h3. conclusion (?)
When you write a script, even if you're sure you will never need to have more than one argument, or that you never will have to update the code, *please* consider to use of *Getopt::Long* instead of a *shift @ARGV*, because we all know that you will at a certain point update this script and you will more than one argument :).
|
