Sponsored by

Chewy

Chewy is an ODM and wrapper for the official Elasticsearch client.

Why Chewy?

• Multi-model indices.

  Index classes are independent from ORM/ODM models. Now, implementing e.g. cross-model autocomplete is much easier. You can just define the index and work with it in an object-oriented style. You can define several types for index - one per indexed model.

• Every index is observable by all the related models.

  Most of the indexed models are related to other and sometimes it is necessary to denormalize this related data and put at the same object. For example, you need to index an array of tags together with an article. Chewy allows you to specify an updateable index for every model separately - so corresponding articles will be reindexed on any tag update.

• Bulk import everywhere.

  Chewy utilizes the bulk ES API for full reindexing or index updates. It also uses atomic updates. All the changed objects are collected inside the atomic block and the index is updated once at the end with all the collected objects. See Chewy.strategy(:atomic) for more details.

• Powerful querying DSL.

  Chewy has an ActiveRecord-style query DSL. It is chainable, mergeable and lazy, so you can produce queries in the most efficient way. It also has object-oriented query and filter builders.

• Support for ActiveRecord, Mongoid and Sequel.

Installation

Add this line to your application's Gemfile:

gem 'chewy'

And then execute:

$ bundle

Or install it yourself as:

$ gem install chewy

Usage

Client settings

There are two ways to configure the Chewy client: the Chewy.settings hash and chewy.yml

You can create this file manually or run rails g chewy:install.

# config/initializers/chewy.rb
Chewy.settings = {host: 'localhost:9250'} # do not use environments

# config/chewy.yml
# separate environment configs
test:
  host: 'localhost:9250'
  prefix: 'test'
development:
  host: 'localhost:9200'

The resulting config merges both hashes. Client options are passed as is to Elasticsearch::Transport::Client except for the :prefix, which is used internally by Chewy to create prefixed index names:

  Chewy.settings = {prefix: 'test'}
  UsersIndex.index_name # => 'test_users'

The logger may be set explicitly:

Chewy.logger = Logger.new(STDOUT)

See config.rb for more details.

Index definition

1. Create /app/chewy/users_index.rb

class UsersIndex < Chewy::Index

end

2. Add one or more types mapping

class UsersIndex < Chewy::Index
  define_type User.active # or just model instead_of scope: define_type User
end

Newly-defined index type class is accessible via UsersIndex.user or UsersIndex::User

3. Add some type mappings

class UsersIndex < Chewy::Index
  define_type User.active.includes(:country, :badges, :projects) do
    field :first_name, :last_name # multiple fields without additional options
    field :email, analyzer: 'email' # Elasticsearch-related options
    field :country, value: ->(user) { user.country.name } # custom value proc
    field :badges, value: ->(user) { user.badges.map(&:name) } # passing array values to index
    field :projects do # the same block syntax for multi_field, if `:type` is specified
      field :title
      field :description # default data type is `string`
      # additional top-level objects passed to value proc:
      field :categories, value: ->(project, user) { project.categories.map(&:name) if user.active? }
    end
    field :rating, type: 'integer' # custom data type
    field :created, type: 'date', include_in_all: false,
      value: ->{ created_at } # value proc for source object context
  end
end

See here for mapping definitions.

4. Add some index- and type-related settings. Analyzer repositories might be used as well. See Chewy::Index.settings docs for details:

class UsersIndex < Chewy::Index
  settings analysis: {
    analyzer: {
      email: {
        tokenizer: 'keyword',
        filter: ['lowercase']
      }
    }
  }

  define_type User.active.includes(:country, :badges, :projects) do
    root date_detection: false do
      template 'about_translations.*', type: 'string', analyzer: 'standard'

      field :first_name, :last_name
      field :email, analyzer: 'email'
      field :country, value: ->(user) { user.country.name }
      field :badges, value: ->(user) { user.badges.map(&:name) }
      field :projects do
        field :title
        field :description
      end
      field :about_translations, type: 'object' # pass object type explicitly if necessary
      field :rating, type: 'integer'
      field :created, type: 'date', include_in_all: false,
        value: ->{ created_at }
    end
  end
end

See index settings here. See root object settings here.

See mapping.rb for more details.

5. Add model-observing code

class User < ActiveRecord::Base
  update_index('users#user') { self } # specifying index, type and back-reference
                                      # for updating after user save or destroy
end

class Country < ActiveRecord::Base
  has_many :users

  update_index('users#user') { users } # return single object or collection
end

class Project < ActiveRecord::Base
  update_index('users#user') { user if user.active? } # you can return even `nil` from the back-reference
end

class Badge < ActiveRecord::Base
  has_and_belongs_to_many :users

  update_index('users') { users } # if index has only one type
                                  # there is no need to specify updated type
end

class Book < ActiveRecord::Base
  update_index(->(book) {"books#book_#{book.language}"}) { self } # dynamic index and type with proc.
                                                                  # For book with language == "en"
                                                                  # this code will generate `books#book_en`
end

Also, you can use the second argument for method name passing:

update_index('users#user', :self)
update_index('users#user', :users)

In the case of a belongs_to association you may need to update both associated objects, previous and current:

class City < ActiveRecord::Base
  belongs_to :country

  update_index('cities#city') { self }
  update_index 'countries#country' do
    # For the latest active_record changed values are
    # already in `previous_changes` hash,
    # but for mongoid you have to use `changes` hash
    previous_changes['country_id'] || country
  end
end

You can observe Sequel models in the same way as ActiveRecord:

class User < Sequel::Model
  update_index('users#user') { self }
end

However, to make it work, you must load the chewy plugin into Sequel model:

Sequel::Model.plugin :chewy_observe  # for all models, or...
User.plugin :chewy_observe           # just for User

Multi (nested) and object field types

To define an objects field you can simply nest fields in the DSL:

field :projects do
  field :title
  field :description
end

This will automatically set the type or root field to object. You may also specify type: 'objects' explicitly.

To define a multi field you have to specify any type except for object or nested in the root field:

field :full_name, type: 'string', value: ->{ full_name.strip } do
  field :ordered, analyzer: `ordered`
  field :untouched, index: 'not_analyzed'
end

The value: option for internal fields would no longer be effective.

Geo Point fields

You can use Elasticsearch's geo mapping with the geo_point field type, allowing you to query, filter and order by latitude and longitude. You can use the following hash format:

field :coordinates, type: 'geo_point', value: ->{ {lat: latitude, lon: longitude} }

or by using nested fields:

field :coordinates, type: 'geo_point' do
  field :lat, value: ->{ latitude }
  field :long, value: ->{ longitude }
end

See the section on Script fields for details on calculating distance in a search.

Crutches™ technology

Assume you are defining your index like this (product has_many categories through product_categories):

class ProductsIndex < Chewy::Index
  define_type Product.includes(:categories) do
    field :name
    field :category_names, value: ->(product) { product.categories.map(&:name) } # or shorter just -> { categories.map(&:name) }
  end
end

Then the Chewy reindexing flow would look like the following pseudo-code (even in Mongoid):

Product.includes(:categories).find_in_batches(1000) do |batch|
  bulk_body = batch.map do |object|
    {name: object.name, category_names: object.categories.map(&:name)}.to_json
  end
  # here we are sending every batch of data to ES
  Chewy.client.bulk bulk_body
end

But in Rails 4.1 and 4.2 you may face a problem with slow associations (take a look at rails/rails#19423). Also, there might be really complicated cases when associations are not applicable.

Then you can replace Rails associations with Chewy Crutches™ technology:

class ProductsIndex < Chewy::Index
  define_type Product.includes(:categories) do
    crutch :categories do |collection| # collection here is a current batch of products
      # data is fetched with a lightweight query without objects initialization
      data = ProductCategory.joins(:category).where(product_id: collection.map(&:id)).pluck(:product_id, 'categories.name')
      # then we have to convert fetched data to appropriate format
      # this will return our data in structure like:
      # {123 => ['sweets', 'juices'], 456 => ['meat']}
      data.each.with_object({}) { |(id, name), result| (result[id] ||= []).push(name) }
    end

    field :name
    # simply use crutch-fetched data as a value:
    field :category_names, value: ->(product, crutches) { crutches.categories[product.id] }
  end
end

An example flow would look like this:

Product.includes(:categories).find_in_batches(1000) do |batch|
  crutches[:categories] = ProductCategory.joins(:category).where(product_id: batch.map(&:id)).pluck(:product_id, 'categories.name')
    .each.with_object({}) { |(id, name), result| (result[id] ||= []).push(name) }

  bulk_body = batch.map do |object|
    {name: object.name, category_names: crutches[:categories][object.id]}.to_json
  end
  Chewy.client.bulk bulk_body
end

So Chewy Crutches™ technology is able to increase your indexing performance in some cases up to a hundredfold or even more depending on your associations complexity.

Types access

You can access index-defined types with the following API:

UsersIndex::User # => UsersIndex::User
UsersIndex.type_hash['user'] # => UsersIndex::User
UsersIndex.user # => UsersIndex::User
UsersIndex.types # => [UsersIndex::User]
UsersIndex.type_names # => ['user']

Index manipulation

UsersIndex.delete # destroy index if it exists
UsersIndex.delete!

UsersIndex.create
UsersIndex.create! # use bang or non-bang methods

UsersIndex.purge
UsersIndex.purge! # deletes then creates index

UsersIndex::User.import # import with 0 arguments process all the data specified in type definition
                        # literally, User.active.includes(:country, :badges, :projects).find_in_batches
UsersIndex::User.import User.where('rating > 100') # or import specified users scope
UsersIndex::User.import User.where('rating > 100').to_a # or import specified users array
UsersIndex::User.import [1, 2, 42] # pass even ids for import, it will be handled in the most effective way

UsersIndex.import # import every defined type
UsersIndex.import user: User.where('rating > 100') # import only active users to `user` type.
  # Other index types, if exists, will be imported with default scope from the type definition.
UsersIndex.reset! # purges index and imports default data for all types

If the passed user is #destroyed?, or satisfies a delete_if type option, or the specified id does not exist in the database, import will perform delete from index action for this object.

define_type User, delete_if: :deleted_at
define_type User, delete_if: -> { deleted_at }
define_type User, delete_if: ->(user) { user.deleted_at }

See actions.rb for more details.

Index update strategies

Assume you've got the following code:

class City < ActiveRecord::Base
  update_index 'cities#city', :self
end

class CitiesIndex < Chewy::Index
  define_type City do
    field :name
  end
end

If you do something like City.first.save! you'll get an UndefinedUpdateStrategy exception instead of the object saving and index updating. This exception forces you to choose an appropriate update strategy for the current context.

If you want to return to the pre-0.7.0 behavior - just set Chewy.root_strategy = :bypass.

:atomic

The main strategy here is :atomic. Assume you have to update a lot of records in the db.

Chewy.strategy(:atomic) do
  City.popular.map(&:do_some_update_action!)
end

Using this strategy delays the index update request until the end of the block. Updated records are aggregated and the index update happens with the bulk API. So this strategy is highly optimized.

:resque

This does the same thing as :atomic, but asynchronously using resque. The default queue name is chewy. Patch Chewy::Strategy::Resque::Worker for index updates improving.

Chewy.strategy(:resque) do
  City.popular.map(&:do_some_update_action!)
end

:sidekiq

This does the same thing as :atomic, but asynchronously using sidekiq. Patch Chewy::Strategy::Sidekiq::Worker for index updates improving.

Chewy.strategy(:sidekiq) do
  City.popular.map(&:do_some_update_action!)
end

:urgent

The following strategy is convenient if you are going to update documents in your index one by one.

Chewy.strategy(:urgent) do
  City.popular.map(&:do_some_update_action!)
end

This code would perform City.popular.count requests for ES documents update.

It is convenient for use in e.g. the Rails console with non-block notation:

> Chewy.strategy(:urgent)
> City.popular.map(&:do_some_update_action!)

:bypass

The bypass strategy simply silences index updates.

Nesting

Strategies are designed to allow nesting, so it is possible to redefine it for nested contexts.

Chewy.strategy(:atomic) do
  city1.do_update!
  Chewy.strategy(:urgent) do
    city2.do_update!
    city3.do_update!
    # there will be 2 update index requests for city2 and city3
  end
  city4..do_update!
  # city1 and city4 will be grouped in one index update request
end

Non-block notation

It is possible to nest strategies without blocks:

Chewy.strategy(:urgent)
city1.do_update! # index updated
Chewy.strategy(:bypass)
city2.do_update! # update bypassed
Chewy.strategy.pop
city3.do_update! # index updated again

Designing your own strategies

See strategy/base.rb for more details. See strategy/atomic.rb for an example.

Rails application strategies integration

There are a couple of predefined strategies for your Rails application. Initially, the Rails console uses the :urgent strategy by default, except in the sandbox case. When you are running sandbox it switches to the :bypass strategy to avoid polluting the index.

Migrations are wrapped with the :bypass strategy. Because the main behavior implies that indices are reset after migration, there is no need for extra index updates. Also indexing might be broken during migrations because of the outdated schema.

Controller actions are wrapped with the :atomic strategy with middleware just to reduce the number of index update requests inside actions.

It is also a good idea to set up the :bypass strategy inside your test suite and import objects manually only when needed, and use Chewy.massacre when needed to flush test ES indices before every example. This will allow you to minimize unnecessary ES requests and reduce overhead.

RSpec.configure do |config|
  config.before(:suite) do
    Chewy.strategy(:bypass)
  end
end

Index querying

scope = UsersIndex.query(term: {name: 'foo'})
  .filter(range: {rating: {gte: 100}})
  .order(created: :desc)
  .limit(20).offset(100)

scope.to_a # => will produce array of UserIndex::User or other types instances
scope.map { |user| user.email }
scope.total_count # => will return total objects count

scope.per(10).page(3) # supports kaminari pagination
scope.explain.map { |user| user._explanation }
scope.only(:id, :email) # returns ids and emails only

scope.merge(other_scope) # queries could be merged

Also, queries can be performed on a type individually:

UsersIndex::User.filter(term: {name: 'foo'}) # will return UserIndex::User collection only

If you are performing more than one filter or query in the chain, all the filters and queries will be concatenated in the way specified by filter_mode and query_mode respectively.

The default filter_mode is :and and the default query_mode is bool.

Available filter modes are: :and, :or, :must, :should and any minimum_should_match-acceptable value

Available query modes are: :must, :should, :dis_max, any minimum_should_match-acceptable value or float value for dis_max query with tie_breaker specified.

UsersIndex::User.filter{ name == 'Fred' }.filter{ age < 42 } # will be wrapped with `and` filter
UsersIndex::User.filter{ name == 'Fred' }.filter{ age < 42 }.filter_mode(:should) # will be wrapped with bool `should` filter
UsersIndex::User.filter{ name == 'Fred' }.filter{ age < 42 }.filter_mode('75%') # will be wrapped with bool `should` filter with `minimum_should_match: '75%'`

See query.rb for more details.

Additional query action.

You may also perform additional actions on the query scope, such as deleting of all the scope documents:

UsersIndex.delete_all
UsersIndex::User.delete_all
UsersIndex.filter{ age < 42 }.delete_all
UsersIndex::User.filter{ age < 42 }.delete_all

Filters query DSL

There is a test version of the filter-creating DSL:

UsersIndex.filter{ name == 'Fred' } # will produce `term` filter.
UsersIndex.filter{ age <= 42 } # will produce `range` filter.

The basis of the DSL is the expression. There are 2 types of expressions:

• Simple function

  UsersIndex.filter{ s('doc["num"] > 1') } # script expression
  UsersIndex.filter{ q(query_string: {query: 'lazy fox'}) } # query expression

• Field-dependent composite expression Consists of the field name (with or without dot notation), a value, and an action operator between them. The field name might take additional options for passing to the resulting expression.

  UsersIndex.filter{ name == 'Name' } # simple field term filter
  UsersIndex.filter{ name(:bool) == ['Name1', 'Name2'] } # terms query with `execution: :bool` option passed
  UsersIndex.filter{ answers.title =~ /regexp/ } # regexp filter for `answers.title` field

You can combine expressions as you wish with the help of combination operators.

UsersIndex.filter{ (name == 'Name') & (email == 'Email') } # combination produces `and` filter
UsersIndex.filter{
  must(
    should(name =~ 'Fr').should_not(name == 'Fred') & (age == 42), email =~ /gmail\.com/
  ) | ((roles.admin == true) & name?)
} # many of the combination possibilities

There is also a special syntax for cache enabling:

UsersIndex.filter{ ~name == 'Name' } # you can apply tilde to the field name
UsersIndex.filter{ ~(name == 'Name') } # or to the whole expression

# if you are applying cache to the one part of range filter
# the whole filter will be cached:
UsersIndex.filter{ ~(age > 42) & (age <= 50) }

# You can pass cache options as a field option also.
UsersIndex.filter{ name(cache: true) == 'Name' }
UsersIndex.filter{ name(cache: false) == 'Name' }

# With regexp filter you can pass _cache_key
UsersIndex.filter{ name(cache: 'name_regexp') =~ /Name/ }
# Or not
UsersIndex.filter{ name(cache: true) =~ /Name/ }

Compliance cheatsheet for filters and DSL expressions:

• Term filter

  {"term": {"name": "Fred"}}
  {"not": {"term": {"name": "Johny"}}}

  UsersIndex.filter{ name == 'Fred' }
  UsersIndex.filter{ name != 'Johny' }

• Terms filter

  {"terms": {"name": ["Fred", "Johny"]}}
  {"not": {"terms": {"name": ["Fred", "Johny"]}}}
  
  {"terms": {"name": ["Fred", "Johny"], "execution": "or"}}
  
  {"terms": {"name": ["Fred", "Johny"], "execution": "and"}}
  
  {"terms": {"name": ["Fred", "Johny"], "execution": "bool"}}
  
  {"terms": {"name": ["Fred", "Johny"], "execution": "fielddata"}}

  UsersIndex.filter{ name == ['Fred', 'Johny'] }
  UsersIndex.filter{ name != ['Fred', 'Johny'] }
  
  UsersIndex.filter{ name(:|) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(:or) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(execution: :or) == ['Fred', 'Johny'] }
  
  UsersIndex.filter{ name(:&) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(:and) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(execution: :and) == ['Fred', 'Johny'] }
  
  UsersIndex.filter{ name(:b) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(:bool) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(execution: :bool) == ['Fred', 'Johny'] }
  
  UsersIndex.filter{ name(:f) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(:fielddata) == ['Fred', 'Johny'] }
  UsersIndex.filter{ name(execution: :fielddata) == ['Fred', 'Johny'] }

• Regexp filter (== and =~ are equivalent)

  {"regexp": {"name.first": "s.*y"}}
  
  {"not": {"regexp": {"name.first": "s.*y"}}}
  
  {"regexp": {"name.first": {"value": "s.*y", "flags": "ANYSTRING|INTERSECTION"}}}

  UsersIndex.filter{ name.first == /s.*y/ }
  UsersIndex.filter{ name.first =~ /s.*y/ }
  
  UsersIndex.filter{ name.first != /s.*y/ }
  UsersIndex.filter{ name.first !~ /s.*y/ }
  
  UsersIndex.filter{ name.first(:anystring, :intersection) == /s.*y/ }
  UsersIndex.filter{ name.first(flags: [:anystring, :intersection]) == /s.*y/ }

• Prefix filter

  {"prefix": {"name": "Fre"}}
  {"not": {"prefix": {"name": "Joh"}}}

  UsersIndex.filter{ name =~ re' }
  UsersIndex.filter{ name !~ 'Joh' }

• Exists filter

  {"exists": {"field": "name"}}

  UsersIndex.filter{ name? }
  UsersIndex.filter{ !!name }
  UsersIndex.filter{ !!name? }
  UsersIndex.filter{ name != nil }
  UsersIndex.filter{ !(name == nil) }

• Missing filter

  {"missing": {"field": "name", "existence": true, "null_value": false}}
  {"missing": {"field": "name", "existence": true, "null_value": true}}
  {"missing": {"field": "name", "existence": false, "null_value": true}}

  UsersIndex.filter{ !name }
  UsersIndex.filter{ !name? }
  UsersIndex.filter{ name == nil }

• Range

  {"range": {"age": {"gt": 42}}}
  {"range": {"age": {"gte": 42}}}
  {"range": {"age": {"lt": 42}}}
  {"range": {"age": {"lte": 42}}}
  
  {"range": {"age": {"gt": 40, "lt": 50}}}
  {"range": {"age": {"gte": 40, "lte": 50}}}
  
  {"range": {"age": {"gt": 40, "lte": 50}}}
  {"range": {"age": {"gte": 40, "lt": 50}}}

  UsersIndex.filter{ age > 42 }
  UsersIndex.filter{ age >= 42 }
  UsersIndex.filter{ age < 42 }
  UsersIndex.filter{ age <= 42 }
  
  UsersIndex.filter{ age == (40..50) }
  UsersIndex.filter{ (age > 40) & (age < 50) }
  UsersIndex.filter{ age == [40..50] }
  UsersIndex.filter{ (age >= 40) & (age <= 50) }
  
  UsersIndex.filter{ (age > 40) & (age <= 50) }
  UsersIndex.filter{ (age >= 40) & (age < 50) }

• Bool filter

  {"bool": {
    "must": [{"term": {"name": "Name"}}],
    "should": [{"term": {"age": 42}}, {"term": {"age": 45}}]
  }}

  UsersIndex.filter{ must(name == 'Name').should(age == 42, age == 45) }

• And filter

  {"and": [{"term": {"name": "Name"}}, {"range": {"age": {"lt": 42}}}]}

  UsersIndex.filter{ (name == 'Name') & (age < 42) }

• Or filter

  {"or": [{"term": {"name": "Name"}}, {"range": {"age": {"lt": 42}}}]}

  UsersIndex.filter{ (name == 'Name') | (age < 42) }

  {"not": {"term": {"name": "Name"}}}
  {"not": {"range": {"age": {"lt": 42}}}}

  UsersIndex.filter{ !(name == 'Name') } # or UsersIndex.filter{ name != 'Name' }
  UsersIndex.filter{ !(age < 42) }

• Match all filter

  {"match_all": {}}

  UsersIndex.filter{ match_all }

• Has child filter

  {"has_child": {"type": "blog_tag", "query": {"term": {"tag": "something"}}}
  {"has_child": {"type": "comment", "filter": {"term": {"user": "john"}}}

  UsersIndex.filter{ has_child(:blog_tag).query(term: {tag: 'something'}) }
  UsersIndex.filter{ has_child(:comment).filter{ user == 'john' } }

• Has parent filter

  {"has_parent": {"type": "blog", "query": {"term": {"tag": "something"}}}}
  {"has_parent": {"type": "blog", "filter": {"term": {"text": "bonsai three"}}}}

  UsersIndex.filter{ has_parent(:blog).query(term: {tag: 'something'}) }
  UsersIndex.filter{ has_parent(:blog).filter{ text == 'bonsai three' } }

See filters.rb for more details.

Faceting

Facets are an optional sidechannel you can request from Elasticsearch describing certain fields of the resulting collection. The most common use for facets is to allow the user to continue filtering specifically within the subset, as opposed to the global index.

For instance, let's request the country field as a facet along with our users collection. We can do this with the #facets method like so:

UsersIndex.filter{ [...] }.facets({countries: {terms: {field: 'country'}}})

Let's look at what we asked from Elasticsearch. The facets setter method accepts a hash. You can choose custom/semantic key names for this hash for your own convenience (in this case I used the plural version of the actual field), in our case countries. The following nested hash tells ES to grab and aggregate values (terms) from the country field on our indexed records.

The response will include the :facets sidechannel:

< { ... ,"facets":{"countries":{"_type":"terms","missing":?,"total":?,"other":?,"terms":[{"term":"USA","count":?},{"term":"Brazil","count":?}, ...}}

Aggregations

Aggregations are part of the optional sidechannel that can be requested with a query.

You interact with aggregations using the composable #aggregations method (or its alias #aggs)

Let's look at an example.

class UsersIndex < Chewy::Index
  define_type User do
    field :name
    field :rating
  end
end

all_johns = UsersIndex::User.filter { name == 'john' }.aggs({ avg_rating: { avg: { field: 'rating' } } })

avg_johns_rating = all_johns.aggs
# => {"avg_rating"=>{"value"=>3.5}}

It is convenient to name aggregations that you intend to reuse regularly. This is achieve with the .aggregation method, which is also available under the .agg alias method.

Here's the same example from before

class UsersIndex < Chewy::Index
  define_type User do
    field :name
    field :rating, type: "long"
    agg :avg_rating do
      { avg: { field: 'rating' } }
    end
  end
end

all_johns = UsersIndex::User.filter { name == 'john' }.aggs(:avg_rating)

avg_johns_rating = all_johns.aggs
# => {"avg_rating"=>{"value"=>3.5}}

It is possible to run into collisions between named aggregations. This occurs when there is more than one aggregation with the same name. To explicitly reference an aggregation you provide a string to the #aggs method of the form: index_name#document_type.aggregation_name

Consider this example where there are two separate aggregations named avg_rating

class UsersIndex < Chewy::Index
  define_type User do
    field :name
    field :rating, type: "long"
    agg :avg_rating do
      { avg: { field: 'rating' } }
    end
  end
  define_type Post do
    field :title
    field :body
    field :comments do
      field :message
      field :rating, type: "long"
    end
    agg :avg_rating do
      { avg: { field: 'comments.rating' } }
    end
  end
end

all_docs = UsersIndex.filter {match_all}.aggs("users#user.avg_rating")
all_docs.aggs
# => {"users#user.avg_rating"=>{"value"=>3.5}} 

Script fields

Script fields allow you to execute Elasticsearch's scripting languages such as groovy and javascript. More about supported languages and what scripting is here. This feature allows you to calculate the distance between geo points, for example. This is how to use the DSL:

UsersIndex.script_fields(
  distance: {
    params: {
      lat: 37.569976,
      lon: -122.351591
    },
    script: "doc['coordinates'].distanceInMiles(lat, lon)"
  }
)

Here, coordinates is a field with type geo_point. There will be a distance field for the index's model in the search result.

Script scoring

Script scoring is used to score the search results. All scores are added to the search request and combined according to boost mode and score mode. This can be useful if, for example, a score function is computationally expensive and it is sufficient to compute the score on a filtered set of documents. For example, you might want to multiply the score by another numeric field in the doc:

UsersIndex.script_score("_score * doc['my_numeric_field'].value")

Boost Factor

Boost factors are a way to add a boost to a query where documents match the filter. If you have some users who are experts and some who are regular users, you might want to give the experts a higher score and boost to the top of the search results. You can accomplish this by using the #boost_factor method and adding a boost score of 5 for an expert user:

UsersIndex.boost_factor(5, filter: {term: {type: 'Expert'}})

Objects loading

It is possible to load source objects from the database for every search result:

scope = UsersIndex.filter(range: {rating: {gte: 100}})

scope.load # => scope is marked to return User instances array
scope.load.query(...) # => since objects are loaded lazily you can complete scope
scope.load(user: { scope: ->{ includes(:country) }}) # you can also pass loading scopes for each
                                                     # possibly returned type
scope.load(user: { scope: User.includes(:country) }) # the second scope passing way.
scope.load(scope: ->{ includes(:country) }) # and more common scope applied to every loaded object type.

scope.only(:id).load # it is optimal to request ids only if you are not planning to use type objects

The preload method takes the same options as load and ORM/ODM objects will be loaded, but the scope will still return an array of Chewy wrappers. To access real objects use the _object wrapper method:

UsersIndex.filter(range: {rating: {gte: 100}}).preload(...).query(...).map(&:_object)

See loading.rb for more details.

ActiveSupport::Notifications support

Chewy has notifying the following events:

search_query.chewy payload

• payload[:index]: requested index class
• payload[:request]: request hash

import_objects.chewy payload

• payload[:type]: currently imported type

• payload[:import]: imports stats, total imported and deleted objects count:

  {index: 30, delete: 5}

• payload[:errors]: might not exists. Contains grouped errors with objects ids list:

  {index: {
    'error 1 text' => ['1', '2', '3'],
    'error 2 text' => ['4']
  }, delete: {
    'delete error text' => ['10', '12']
  }}

NewRelic integration

To integrate with NewRelic you may use the following example source (config/initializers/chewy.rb):

ActiveSupport::Notifications.subscribe('import_objects.chewy') do |name, start, finish, id, payload|
  metric_name = "Database/ElasticSearch/import"
  duration = (finish - start).to_f
  logged = "#{payload[:type]} #{payload[:import].to_a.map{ |i| i.join(':') }.join(', ')}"

  self.class.trace_execution_scoped([metric_name]) do
    NewRelic::Agent.instance.transaction_sampler.notice_sql(logged, nil, duration)
    NewRelic::Agent.instance.sql_sampler.notice_sql(logged, metric_name, nil, duration)
    NewRelic::Agent.record_metric(metric_name, duration)
  end
end

ActiveSupport::Notifications.subscribe('search_query.chewy') do |name, start, finish, id, payload|
  metric_name = "Database/ElasticSearch/search"
  duration = (finish - start).to_f
  logged = "#{payload[:type].presence || payload[:index]} #{payload[:request]}"

  self.class.trace_execution_scoped([metric_name]) do
    NewRelic::Agent.instance.transaction_sampler.notice_sql(logged, nil, duration)
    NewRelic::Agent.instance.sql_sampler.notice_sql(logged, metric_name, nil, duration)
    NewRelic::Agent.record_metric(metric_name, duration)
  end
end

Rake tasks

Inside the Rails application, some index-maintaining rake tasks are defined.

rake chewy:reset # resets all the existing indices, declared in app/chewy
rake chewy:reset[users] # resets UsersIndex only

rake chewy:update # updates all the existing indices, declared in app/chewy
rake chewy:update[users] # updates UsersIndex only

rake chewy:reset performs zero-downtime reindexing as described here. So basically rake task creates a new index with uniq suffix and then simply aliases it to the common index name. The previous index is deleted afterwards (see Chewy::Index.reset! for more details).

Rspec integration

Just add require 'chewy/rspec' to your spec_helper.rb and you will get additional features: See update_index.rb for more details.

If you use DatabaseCleaner in your tests with the transaction strategy, you may run into the problem that ActiveRecord's models are not indexed automatically on save despite the fact that you set the callbacks to do this with the update_index method. The issue arises because chewy indexes data on after_commit run as default, but all after_commit callbacks are not run with the DatabaseCleaner's' transaction strategy. You can solve this issue by changing the Chewy.use_after_commit_callbacks option. Just add the following initializer in your Rails application:

#config/initializers/chewy.rb
Chewy.use_after_commit_callbacks = !Rails.env.test?

TODO a.k.a coming soon:

• Typecasting support
• Advanced (simplified) query DSL: UsersIndex.query { email == 'my@gmail.com' } will produce term query
• update_all support
• Maybe, closer ORM/ODM integration, creating index classes implicitly

Contributing

1. Fork it (http://github.com/toptal/chewy/fork)
2. Create your feature branch (git checkout -b my-new-feature)
3. Implement your changes, cover it with specs and make sure old specs are passing
4. Commit your changes (git commit -am 'Add some feature')
5. Push to the branch (git push origin my-new-feature)
6. Create new Pull Request

Use the following Rake tasks to control the Elasticsearch cluster while developing.

rake elasticsearch:start # start Elasticsearch cluster on 9250 port for tests
rake elasticsearch:stop # stop Elasticsearch