Skip to content

Ardis is a simple library that allows you to leverage the NoSQL Redis store to index, group, rank or associate traditional ActiveRecord objects.

License

Notifications You must be signed in to change notification settings

cardinalblue/ardis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

28 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Ardis

Ardis is a simple library that allows you to leverage the NoSQL Redis store to index, group, rank or associate traditional ActiveRecord objects. Internally, Ardis only stores the ids of the ActiveRecord objects in Redis data structures, but provides an ActivRecord association-like interface.

Installation

Add this line to your application's Gemfile:

gem 'ardis'

And then execute:

$ bundle

Or install it yourself as:

$ gem install ardis

Usage

Basic usage is using the DSL to create a class-level collection (i.e. a "Series").

require 'ardis'

class Collage < ActiveRecord::Base
  include Ardis
  series_list name: 'featured', global: true
end

c = Collage.create
Collage.featured << c
Collage.featured.first
 => #<Collage id: 1, ... >

Or you can create an instance-level Series:

class Collage < ActiveRecord::Base
  include Ardis
  series_list name: 'likers', relation: User
end

c = Collage.create
c.likers << User.create(name: 'John Doe')
c.likers << User.create(name: 'Jane Tow')
c.likers.limit(2).to_a
 => [ #<User id: 1, ... >, #<User id: 2, ... > ]

Which would result in:

Diagram with PostgreSQL and Redis

The instance that holds the Series (in the example above the Collage c) is referred to as the "container".

Of course you can always create a Series manually:

s = Ardis::RedisAdapter::ListSeries.new(key: 'collages:featured', relation: Collage)
s << Collage.create

When retrieving objects from a collection, most of the methods familiar from ActiveRecord relations work:

m.users.limit(10).offset(2).reverse_order
m.users.page(2)                      # `Kaminari` integration
m.users.includes?(@joe)              # Check if in collection
m.users.includes(:some_association)  # Eager-loading
m.users.count                        # Can be faster, unlike PostgreSQL
m.users.empty?
m.users.delete(@joe)
m.users.where(gender: 'male')        # Will `nil`-out objects that don't match
                                     # (but only remove if `autocompact`, see below).

Redis keys

If using the DSL, Ardis will choose the appropriate Redis keys based on the Class name, container and given name, and depending if the Series is global or not, but the key can always be overridden manually:

series_sorted_set name: 'mycollection', key: 'custom:redis:key', global: true

or calculated on the fly:

series_sorted_set name: 'mycollection', key: ->(container){ "mycollection:#{container.id}:somename" }

Series types

Implementations are provided for the basic Redis data structures:

Inverse Series

Two Series' can be declared inverses of each other in a reciprocal association, so that inserting into one, automatically inserts into the other one.

For example, suppose we have Users that like Photos.

class User < ActiveRecord::Base
  series_list name: 'liked_photos', relation: Photo, inverse_of: 'likers'
end

class Photo < ActiveRecord::Base
  series_list name: 'likers', relation: User, inverse_of: 'liked_photos'
end

u = User.create name: 'Jaime'
p = Photo.create caption: 'The beach!'
u.liked_photos << p
p.likers.to_a
 => [ #<User @name='Jaime' ... > ]

Internally they are two different Redis lists, but you only have to insert and remove from one of them.

Autocompact

If the underlying ActiveRecord row is deleted, Ardis will return nil for that object. Running a query with autocompact automatically purges that id from the Redis datastructure.

s = Ardis::RedisAdapter::ListSeries.new name: 'list', relation: User
s << User.create name: 'Bill'
s << User.create name: 'Clay'
s.to_a
 => [ #<User name: 'Bill'>, #<User name: 'Clay'> ]

User.where(name: 'Bill').destroy_all

s.limit(2).to_a
 => [ nil, #<User name: 'Clay'> ]
s.limit(2).autocompact.to_a
 => [ #<User name: 'Clay'> ]
s.limit(2).to_a
 => [ #<User name: 'Clay'> ]

Sorted Sets

One of the most useful Redis data structures is the Sorted Set. With Ardis, each of the ActiveRecord objects (or rather its id) associated with a score, by which they are sorted. The score can be configured to be read automatically from an attribute:

class User < ActiveRecord::Base
  series_sorted_set global: true, name: :queue_by_age, attr_score: :age
end

@joe = User.create age: 21
User.queue_by_age << @joe
User.queue_by_age.score_for(@joe)
 => 21.0
User.queue_by_age.incr(@joe, 1)

The score can also be calculated on the fly and set automatically upon reading (to avoid extra trips to the database):

class User < ActiveRecord::Base
  series_sorted_set global: true, name: :queue_by_bmi, attr_score: :bmi

  attr_writer :bmi
  def bmi
    @bmi ||= weight / (height * height)
  end
end

@bob = User.create height:186, weight:80
User.queue_by_bmi << @bob
users = User.queue_by_age.with_scores.to_a
 => [ #<User @bmi=0.0023 ... > ]

Advanced usage

(Documentation coming soon)

  • initializer
  • Custom Redis client, for pipelining/multi

Development

After checking out the repo, run bin/setup to install dependencies. Then, run rake test to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/cardinalblue/ardis.

License

The gem is available as open source under the terms of the MIT License.

About

Ardis is a simple library that allows you to leverage the NoSQL Redis store to index, group, rank or associate traditional ActiveRecord objects.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published