Project

unique_by

0.02
No commit activity in last 3 years
No release in over 3 years
unique_byodedniv/unique_byHomepageDocumentationSource CodeBug TrackerWiki
Allows uniqueness of a record when sharding (specifying the shard ID as the group) or span accross tables (receipts).
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
 Popularity
Downloads
957,929
Stars
0
Forks
0
Watchers
2
 Releases
Current version
3.0.0
Total releases
5
First release
2014-09-16
Latest release
2014-12-11
 Development
Primary Language
Ruby
Licenses
UNLICENSE
Average date of last 50 commits
2014-10-22
Reverse Dependencies
0

 Dependencies

Development

bundler
~> 1.6
rspec
~> 3.1
rspec-its
~> 1.0

Runtime

generate_method
~> 1.0
 Project Readme

unique_by Gem Version

This simple gem allows specifying uniqueness groups for an attribute, giving you something to expose to the outside world.

When do you need this?

  • If you're sharding your database.
  • If you have multiple tables that you want to expose a unique ID for.

Installation

Add this line to your application's Gemfile:

gem 'unique_by'

And then execute:

$ bundle

Or install it yourself as:

$ gem install unique_by

Usage

Sharding example

You first need to specify a unique group in your model:

# == Schema Info
#
# Table name: medical_bills
#
#  id                  :integer(11)    not null, primary key
#  client_id           :integer(11)
#

class MedicalBill < ActiveRecord::Base
  unique_by client_id: 50 # total of 50 clients
end

then, you can use these basic methods:

bill1 = MedicalBill.find(123) # from a DB shard for client_id = 1
bill2 = MedicalBill.find(123) # from a DB shard for client_id = 2

bill1.unique_id
=> 7873
bill2.unique_id
=> 7874

You can use the singleton methods:

MedicalBill.unique_id_from(123, client_id: 1) # gives the unique_id
=> 7873
MedicalBill.id_from(7873) # gives the id
=> 123
MedicalBill.id_group_from(7874) # gives the client_id
=> { client_id: 2 }

You can specify multiple unique group attributes:

class MedicalBill < ActiveRecord::Base
  unique_by client_id: 50, client_part: 5 # total of 50 clients and 5 parts
end

It is recommended to create finder methods that will find records according to their id group, like so:

class MedicallBill < ActiveRecord::Base
  unique_by client_id: 50, client_part: 5 # total of 50 clients and 5 parts

  def self.find_by_unique_id(unique_id)
    withing_client_connection(id_group_from(unique_id)[:client_id]) do
      find(id_from(unique_id))
    end
  end
end

Multiple tables example

You can supply a block to give a custom mechanism for determining the group:

class MedicalBill < ActiveRecord::Base
  unique_by(type_index: 2) { { type_index: 0 } }
end
class UtilityBill < ActiveRecord::Base
  unique_by(type_index: 2) { { type_index: 1 } }
end

Groups can also be served as a singleton method:

class MedicalBill < ActiveRecord::Base
  unique_by type_index: 2
  def self.type_index
    0
  end
end
class UtilityBill < ActiveRecord::Base
  unique_by type_index: 2
  def self.type_index
    1
  end
end

In either case, if all the groups are specified as singleton methods or served within the block WITHOUT using the instance (self), you can use the singleton methods without explicitly specifying the groups for this class:

MedicalBill.unique_id_from(123)
=> 247
UtilityBill.unique_id_from(123)
=> 246

It is recommended to create finder methods that will find records according to their id group, like so:

module Bill
  module_function
  def find_by_unique_id(unique_id)
    case MedicalBill.id_group_from(unique_id)[:type_index]
      when MedicalBill.type_index then MedicalBill.find(MedicalBill.id_from(unique_id))
      when UtilityBill.type_index then UtilityBill.find(UtilityBill.id_from(unique_id))
    end
  end
end

You can supply both group attributes and a block, and the block can also return more than one field:

class MedicalBill < ActiveRecord::Base
  unique_by(client_id: 50, client_part: 5, xy: 10, halfz: 20) do
    { xy: self.x * self.y, halfz: self.z / 2 }
  end
end

Not ActiveRecord

The generator module is already included in ActiveRecord::Base, but if you want the above methods in another class you can extend it:

class MyClass
  extend UniqueBy::Generator

  unique_by ..., primary_key: :id
  #
  # OR
  #
  def self.primary_key
    :id # or 'id'
  end
end

See also

After adding the unique groups to the id, the unique_id might turn out pretty large. You could use rebase_attr to fix that:

class MedicalBill < ActiveRecord::Base
  unique_by client_id: 32**2 # two base32 letters
  rebase_attr :unique_id, to: 32, readable: true # digits and leters, without '0', 'o', '1' and 'l'
end

bill = MedicalBill.find(3528918) # from a DB shard for client_id = 78
bill.unique_id
=> "ywr3b2e"
bill.unique_id_without_rebase
=> 1806806094

Contributing

  1. Fork it ( https://github.com/odedniv/unique_by/fork )
  2. Create your feature branch (git checkout -b my-new-feature)
  3. Commit your changes (git commit -am 'Add some feature')
  4. Push to the branch (git push origin my-new-feature)
  5. Create a new Pull Request