sub get_descriptions { my ($self, $packages) = @_; my @descriptions = (); for my $class (@{$packages->{classes}}) { my %local_hash = (); $local_hash{id} = $class->{classid}; $local_hash{description} = $class->{classdescription}; push @descriptions, {%local_hash}; for my $method (@{$class->{methods}}) { my %local_hash = (); $local_hash{id} = $method->{methodid}; $local_hash{description} = $method->{methoddescription} +; push @descriptions, {%local_hash}; } for my $attribute (@{$class->{attributes}}) { my %local_hash = (); $local_hash{id} = $attribute->{attributeid}; $local_hash{description} = $attribute->{attributedescri +ption}; push @descriptions, {%local_hash}; } } @descriptions; }
For your note: I agree for the example the arrays are not neccessary, but there can be multiple methods or attributes (it's an UML XMI generator). As far as I understood the code, there are no accessor methods (the already existing part of the generator retrieves its data in the same way).
Thank you again, this helps me a lot!
NikoPS: Is there a good reference where I can learn the subtleties of scalars, arrays and hashes? I understand the basic difference, but I'm totally at a loss when to use which prefix (e. g. my %local_hash = (); vs. $local_hash{id} vs. $class->{methods}).
In reply to Re^2: Retrieving data from deep structure
by enikao
in thread Retrieving data from deep structure
by enikao
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |