--- /dev/null
+#!/usr/bin/perl
+
+use strict;
+use warnings;
+
+use lib './lib';
+
+use Moose;
+
+=pod
+
+=head1 ROADMAP
+
+This is the roadmap for the moosedoc utility. It is just a rough
+sketch of what I am thinking for this.
+
+First question, should it be source-file oriented? or class oriented?
+
+In other words, should I have to do this:
+
+ > moosedoc --target ./my_project/lib/
+
+And have moosedoc traverse the ./my_project/lib/ directory looking for
+.pm files, loading each one and then creating a .pod for it based on the
+moose introspection?
+
+Or should it do this:
+
+ > moosedoc --target ./my_project/script.pl
+
+And have moosedoc then ask Moose what classes/types/subtypes/etc. I
+loaded, and create some kind of .pod for them?
+
+Second question, should I create a large source repository like javadoc?
+or should it just be a file-per-file thing?
+
+If I do it like javadoc, then I would need an index file, a frameset, a
+file for all types/subtypes made, one for all classes, one for all roles,
+etc. At that point, POD may not make sense, and we are into pure HTML
+(for the hyperlinking of course). This then restricts the type of output.
+
+Hmmm,.. gotta do some thinking.
+
+=cut
around 'validate' => sub {
my $c = shift;
my ($self, $field) = @_;
- if ($c->($self, $self->validation_value($field))) {
- return undef;
- }
- else {
- return $self->error_message;
- }
+ return undef if $c->($self, $self->validation_value($field));
+ return $self->error_message;
};
sub validation_value {
around 'validate' => sub {
my $c = shift;
my ($self, $field) = @_;
- if ($c->($self, $self->validation_value($field))) {
- return undef;
- }
- else {
- return $self->error_message;
- }
+ return undef if $c->($self, $self->validation_value($field));
+ return $self->error_message;
};
sub validation_value {