Daedal

This repository contains a set of Ruby classes designed to make ElasticSearch query creation simpler and easier to debug. Type checking and attribute coercion are handled using Virtus to make it harder to construct invalid ElasticSearch queries before sending them to the server.

The ElasticSearch Query DSL has a tremendous amount of flexibility, allowing users to finely tune their search results. However, elaborate queries often take the form of a complex, deeply nested hash, which can become difficult to create or traverse. By wrapping the core components of the query DSL into Ruby objects, Daedal addresses the following issues:

Constructing a large nested hash can be a headache, and using some of the popular hash extensions may result in performance issues
Remembering all the optional parameters each query can take, where they reside within the query structure, and what values they can take can be challenging
Improperly structured queries, or queries with bad parameters, are hard to catch before sending to the server (and receiving an error)
Debugging invalid queries can be a grueling task

Daedal also makes it easy to define custom queries tailored to your specific use case - you can see a simple example at the end of the documentation.

Installation

From the terminal:

$ gem install daedal

or in your Gemfile:

gem 'daedal'

Then, it's as simple as including the line:

require 'daedal'

Usage

See the Daedal Wiki for some examples.

ElasticSearch Query DSL

Other Ruby packages for ElasticSearch allow you to create queries either as hashes or by constructing raw JSON:

author_query = {'match' => {'author' => {'query' => 'Beckett'}}}

For simple queries like the example above, this works just fine. However, as queries become more complicated, the hash can quickly take on a life of its own. Inspired by ElasticSearch's Java API, Daedal contains Ruby classes designed to make query construction more manageable.

Queries

Queries are contained within the Daedal::Queries module. You can construct query components like:

author_query = Daedal::Queries::MatchQuery.new(field: 'author', query: 'Beckett')

Each query has #to_json defined for easy conversion for use with any of the Ruby ElasticSearch clients out there:

author_query.to_json # => "{\"match\":{\"author\":{\"query\":\"Beckett\"}}}"

The benefits of using Daedal become more obvious for aggregate queries such as the bool query:

bool_query = Daedal::Queries::BoolQuery.new(must: [author_query])
bool_query.to_json # => "{\"bool\":{\"should\":[],\"must\":[{\"match\":{\"author\":{\"query\":\"Beckett\"}}}],\"must_not\":[]}}"

lines_query = Daedal::Queries::MatchQuery.new(field: 'lines', query: "We're waiting for Godot")
bool_query.should << lines_query

bool_query.to_json # => "{\"bool\":{\"should\":[{\"match\":{\"lines\":{\"query\":\"We're waiting for Godot\"}}}],\"must\":[{\"match\":{\"author\":{\"query\":\"Beckett\"}}}],\"must_not\":[]}}"

Currently, the following queries have been implemented:

On deck:

Queries I'm not planning on implementing at all, since they're deprecated:

Filters

Filters are contained within the Daedal::Filters module. You can construct filter components in the same way as queries:

term_filter = Daedal::Filters::TermFilter.new(field: 'characters', term: 'Pozzo')

term_filter.to_json # => "{\"term\":{\"characters\":\"Pozzo\"}}"

Currently, the following filters have been implemented:

Type checking and attribute coercion

When creating ElasticSearch queries via nested hashes, it is all too easy to assign an invalid value to a specific field, which would then result in an error response when sending the query to the server. For instance, the query:

constant_score_query = {'constant_score' => {'boost' => 'foo', 'query' => {'match_all' => {}}}}

would yield a server error, since the boost parameter must be a number.

Daedal uses Virtus to perform data-type coercions. Invalid query parameters are surfaced at runtime, making debugging much easier. The previous example in Daedal would raise an error:

match_all_query = Daedal::Queries::MatchAllQuery.new

constant_score_query = Daedal::Queries::ConstantScoreQuery.new(boost: 'foo', query: match_all_query)
# Virtus::CoercionError: Failed to coerce "foo" into Float

Similarly, trying to add non-queries to an aggregate query like the bool or dis max queries will result in an error:

dis_max_query = Daedal::Queries::DisMaxQuery.new

dis_max_query.queries << :foo
# Virtus::CoercionError: Failed to coerce :foo into "Daedal::Queries::Query"

Creating custom queries

Currently, I've only made it through a fraction of the entire Query DSL, but will be adding more as time goes by. If there's a component of the query DSL that you need that isn't implemented yet, it's easy to define it yourself within your project (or, feel free to contribute to Daedal!).

Creating custom filters or queries tailored specifically to your use case is also pretty straightforward. Here are some guidelines:

Make your class inherit from Daedal::Queries::Query or Daedal::Filters::Filter
Define the parameters for your query using Virtus attributes (note: strict coercion is being used)
Define the #to_hash method

Example of a custom query:

class PlayQuery < Daedal::Queries::Query

  # define the parameters that you want in your query
  # if the field is optional, make sure to set required to false
  attribute :author, String
  attribute :title, String

  attribute :characters, Array[String], required: false

  def construct_query
    author_query = Daedal::Queries::MatchQuery.new(field: 'author', query: author)
    title_query = Daedal::Queries::MatchQuery.new(field: 'title', query: title)

    full_query = Daedal::Queries::BoolQuery.new(must: [author_query], should: [title_query])
    characters.each do |character|
      full_query.should << Daedal::Queries::MatchQuery.new(field: 'characters', query: character)
    end

    full_query
  end

  # define the to_hash method to convert for use in ElasticSearch
  def to_hash
    construct_query.to_hash
  end
end

play_query = PlayQuery.new(author: 'Beckett', title: 'Waiting for Godot', characters: ['Estragon', 'Vladimir'])
puts play_query.to_json # => {"bool":{"should":[{"match":{"title":{"query":"Waiting for Godot"}}},{"match":{"characters":{"query":"Estragon"}}},{"match":{"characters":{"query":"Vladimir"}}}],"must":[{"match":{"author":{"query":"Beckett"}}}],"must_not":[]}}

Contributing

The ElasticSearch Query DSL is pretty big and includes a ton of nuance. I'm starting with the most basic parts of the DSL (and the parts I use for work), so if you want to help out with the project to meet your needs please feel free to contribute! I just ask that you:

Fork the project
Make your changes or additions
Add tests! My goal is to keep Daedal a thoroughly tested project
Send me a pull request

Feedback or suggestions are also always welcome!

License

The MIT License (MIT)

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.