SmartSearch

A simple search plug-in which provides a full-text search for ActiveRecord Models. It builds the search tags based upon the attributes you define in your model.

It requires a MySQL Database, and the mysql2 gem!

First run

rake smart_search:install:migrations

To get the data which will be used for your index there are two ways:

# Attributes can be symbols:
# Use symbols to access your models instance methods and attributes
smart_search :on => [:first_name, :last_name]

# Attributes can be strings:
# Use this to access nested attributes
smart_search :on => ["customer.first_name", "cutomer.last_name"]

You can define the attributes for the index in your model in two different ways. With and without evalutation:

# Without evaluation
# All attributes have the save importance 
smart_search :on => ["customer.first_name", "cutomer.last_name"]

# With evaluation
# Higher Boost == higher importance
smart_search :on => [
  {:field_name => "customer.first_name", :boost => 2.5}, 
  {:field_name => "cutomer.last_name", :boost => 1}
]

# So a customer with matching first_name will be higher ranked 
# then one with matching last name

smart_searched models will automatically update their search tags after save. To set search tags for all rows in database, use

Model.set_search_index   

Find entries by using:

Model.find_by_tags("your search tags")

To add default conditions:

smart_search :on => [- define attributes here-], :conditions => "- SQL Conditions here -"

Will only find matching tags and the given condition

Lets say you have a Customer model which has the columns 'first_name', 'last_name' and 'email' and also belongs to a User, who also has the column 'name'.

To build your search for Customer, add:

smart_search :on => [:first_name, :last_name, :email, 'user.name']

to your model. For accessing columns of the model, use symbols, for accessing methods use strings.

class User < ActiveRecord::Base

  belongs_to :office, :class_name => "Office", :foreign_key => "office_id"

  smart_search :on => [:full_name, 'office.name']

  def full_name
    "#{self.first_name} #{self.last_name}"
  end  
end

Saving an User with first_name = "Test" and last_name = "User" the created search tags would be "test user". User.find_by_tags("Test User") would find the User. If the User belongs to an office with name "Headquarter", the search tags would be "test user headquarter". User.find_by_tags("Test Headquarter") would find the User.

User.find_by_tags("Test", :conditions => {:office_id => 1}) 

Also works as expected:-)

class Customer < ActiveRecord::Base
  smart_search :on => [:first_name, :last_name, 'user.full_name', 'user.office.name', :birthday], :conditions => "is_private = 0"

  def user
    User.find(self.user_id)
  end  

end        

class Office < ActiveRecord::Base
  smart_search :on => [:name, :user_names]

  def user_names
    self.users.map {|u| u.full_name }.join(" ")
  end

  def users
    User.find_all_by_office_id(self.id)
  end    
end  

This is still very experimental. THis gem now allows to build up a table with word similarities and combine your search query with other words you might be looking for.

See class 'SmartSimilarty' for more infos...

Documentation coming soon

Added rake tasks for loading similarity data

rake smart_search:similarity_from_file           # Load similarity data from file - Use FILE=path/to/file to specify file
rake smart_search:similarity_from_query_history  # Load similarity data from query history
rake smart_search:similarity_from_url            # Load similarity data from url - Use URL=http://.../ to specify url - Requires 'curl'

These will fill the similarity table with word similarity data.

When using rvm, calling the rake task cron tab must be called like that

source "$HOME/.rvm/scripts/rvm" && rvm use 2.1.0 &&  cd /path/to/app  && RAILS_ENV=production rake smart_search:similarity_from_query_history

Files:       7

Classes:     6 (2 undocumented)
Modules:     3 (0 undocumented)
Constants:   5 (0 undocumented)
Attributes:  0 (0 undocumented)
Methods:    19 (3 undocumented)

Total:      33 (5 undocumented)
 84.85% documented

• Maybe add a search controller, and some views for quick starting...
• Documentation for similarities
• make configurable with initializer
• Dont split searchable attributes if no boost is used

Florian Eck it-support@friends-systems.de

The MIT License (MIT)

Copyright (c) [2014] [Florian Eck]

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.