NAME

Gideon::Manual::Intro - Gideon Manual Intro

VERSION

version 0.0.3

DESCRIPTION

Writing code to retrieve objects from data stores, such as databases, is not only time consuming, but it is also tedious. Sometimes what it's even worst is the fact code is written in a way that only works with a particular data store or data store's dialect. This elevates the cost of maintaining that piece of code.

Gideon aims to reduce your need to write, data store specific code in your application. It doesn't eliminate the needs to write such code and there are still plenty of applications where Gideon (or any other data mapper) will not be the best fit.

Setup

Setting up Gideon is really simple, all you need to do is register a data store and associate any Moose class to a particular data store.

  # Registering a database
  use Gideon::Registry;
  my $dbh = DBI->connect(...);
  Gideon::Registry->register_store( mydb => $dbh );

The previous snippet of code is usually found in your application's start-up script. In this case we have registered a database connection under the mydb name. The following thing you need to do is to associate any Moose class with mydb data store. For example consider the following class

  package MusicRecord;
  use Moose;

  has id => ( is => 'rw', isa => 'Num' );
  has artist => (is => 'rw', isa => 'Str' );

  1;

In order to make associate that to mydb the code needs to be changed to:

  package MusicRecord;
  use Gideon driver => 'DBI';

  has id => ( 
      is => 'rw', 
      isa => 'Num', 
      serial => 1, 
      primary_key => 1,
      traits => [ 'Gideon::DBI::Column' ]
  );

  has artist => ( 
      is => 'rw',
      column => 'artist_name',
      isa => 'Str', 
      traits => [ 'Gideon::DBI::Column' ]
  );

  __PACKAGE__->meta->store('mydb:musicrecord');
  1;

The disruption in your model is minimal. The particular details for a each data store vary, but the philosophy is the same. Let's break down into smaller pieces.

  use Gideon driver => 'DBI';

First of all, in order to reduce boilerplate code, Gideon exports Moose into your model class. The driver = 'DBI'> tells Gideon that this object will use the DBI driver to access the data store.

  has id => ( 
      is => 'rw', 
      isa => 'Num', 
      serial => 1, 
      primary_key => 1,
      traits => [ 'Gideon::DBI::Column' ]
  );

We added serial, primary_key and traits meta-attributes to the id attribute of the class MusicRecord.

In this case serial is a hint for Gideon and tells that for new objects the data store will provide the id. This is very common in the RDB world, in order to ensure consistency, the database provides the id for each inserted record.

primary_key is another hint for Gideon telling that this field is required to uniquely identify a particular record from another. This is helpful when trying to update or remove a particular object.

Finally <traits = [ 'Gideon::DBI::Column']>> is necessary for each attribute that is persisted in the database and what's adds the other meta-attributes.

  has artist => ( 
      is => 'rw',
      column => 'artist_name',
      isa => 'Str', 
      traits => [ 'Gideon::DBI::Column' ]
  );

The only different to the id attribute is the column option, The Gideon DBI driver allows you to use a mapping between your column names and attributes, in this case the attribute artist will be mapped to the column artist_name. If not provided Gideon will assume that the attribute name is the column name.

Finally to associate the class to a registered data store we need to incorporate the following code

  __PACKAGE__->meta->store('mydb:musicrecord');

Here mydb is the data store and musicrecord is the resource within the data store, in the case of DBI the resource is table name. And that's it, your application is ready to use.

Using Gideon

Gideon exports find, find_one, update and remove methods to your model classes and it's the common interface that all your classes will have regardless of the data store. Now all you need to do is operate in with your model classes using the API provided.

  # Create a record
  my $new_record = MusicRecord->new( artist => 'Coldplay' );
  $new_record->save;

  # Finding some records
  my @beatles_albums = MusicRecord->find( artist => 'The Beatles');

  # Updating records
  $_->update( artist => 'The Beatles!') for @beatles_albums;

  # Removing records
  $new_record->remove;

These are really basic use cases and they exhibit the full potential of Gideon, please refer to Gideon::Manual::Creating, Gideon::Manual::Finding, Gideon::Manual::Updating and Gideon::Manual::Removing for a complete explanation of those methods.

NAME

Gideon Intro

VERSION

version 0.0.3

AUTHOR

Mariano Wahlmann, Gines Razanov

COPYRIGHT AND LICENSE

This software is copyright (c) 2013 by Mariano Wahlmann, Gines Razanov.

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.

cpanm Gideon

perl -MCPAN -e shell
install Gideon