BigIndex¶ ↑

A Rails plugin that drops into models and provides indexing functionality. Uses an adapter/repository pattern inspired by Datamapper to abstract the actual indexer used in the background, and exposes the model to a simple indexing API.

This should be used in conjunction with BigRecord in order to provide a more complete ORM.

Supported search servers¶ ↑

Solr
Sphinx (planned)

Installation¶ ↑

(1) Install the Solr search server into your Rails application with the bigindex-solr package [github.com/openplaces/bigindex-solr]. Alternatively, if you want to setup your own Solr server, you’ll need to read the [Solr schema] section below and take note of the Solr url to connect to.

(2) In your Rails application, add Bigindex as a gem to your config/environment.rb file:

config.gem "bigindex", :source => "http://gemcutter.org"

and run the following rake task to install all the gems listed for your Rails app:

[sudo] rake gems:install

(3) Add the following line to the bottom of your RAILS_ROOT/Rakefile

require 'big_index/tasks'

NOTE: Skip steps (4) and (5) if you’re using the bigindex-solr package.¶ ↑

(4) Create a Bigindex config file for your Rails application

rake bigindex:generate_config

(5) Modify the config file config/bigindex.yml to correspond to your Solr server.

Getting Started¶ ↑

Modify your Ruby class/model similarly to the following:

class Model < BigRecord::Base
  include BigIndex::Resource                    # 1. Include the BigIndex::Resource module into your model.

  column :name, :string
  column :description, :text

  index :name,  :string                         # 2. Define each attribute you want to index along with its type.
  index :description                            # Defaults to type :text
end

BigIndex will then override the default Model.find() method to pass through the indexer first. Model.find() will also accept the option “:bypass_index => true”, which bypasses the indexed #find method and dispatches it to the original Model.find() method, e.g. Model.find(:all, :bypass_index => true). Alternatively, you can use Model.find_without_index(:all) for the same functionality.

The {BigIndex::Model#index} macro here is used to define which fields will be indexed, and how they will be returned. Please refer to that method for more information.

Rebuilding your index¶ ↑

Once you have your models setup with Bigindex, you will need to rebuild the index. To do so, run the rake task:

rake bigindex:rebuild_index

This rake task accepts the following options:

* BATCH_SIZE: Defaults to 100. For each model, it will batch process this number of records at a time.
* CLEAR: Defaults to true. Before rebuilding the model, drop the current index before rebuilding it.
* OPTIMIZE: Defaults to true. A Solr option to optimize the index after it's created.

An example usage is:

rake bigindex:rebuild_index BATCH_SIZE=150 CLEAR=true OPTIMIZE=true

Drop current index¶ ↑

To drop the index of your current environment, use:

rake bigindex:drop_index

This will drop the index of any model in your RAILS_ROOT/app/models folder that includes BigIndex::Resource. Be very careful with this task!

Solr schema¶ ↑

Bigindex has a defined schema to use with Solr. If you need to set up Solr yourself, you’ll need to either use the schema.xml defined in this project, or create a merged version of it with your own. You can find the schema file here: github.com/openplaces/bigindex/blob/master/schema.xml

bigindex

Development

Runtime