Skip to content

monstar-lab/elasticsearch-service

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Elasticsearch Service for Vapor

Swift Vapor

This library allows you to connect to a Elasticsearch server from your Vapor application, perform searches and get back Codable/Content results. Currently only searching is supported, not indices nor data manipulation.

Setup

Add Elasticsearch Service to your Package.swift file:

.package(url: "https://github.com/monstar-lab/elasticsearch-service.git", from: "0.9.0")

Register the configuration object and the provider inside your configure.swift file:

let elasticURL = URL(string: "http://localhost:9200")

if let elasticURL = elasticURL {
  let elasticConfig = ElasticsearchConfig(serverURL: elasticURL)
  services.register(elasticConfig)

  try services.register(ElasticsearchProvider())
}

Using the service

After importing ElasticsearchService and obtaining instance of ElasticsearchClient from your container, you can send your search request (either wrapped in QueryContainer, or a simple [AnyHashable: Any] in case of complicated queries) and receive search results (again either as custom Codable structure, or as an [AnyHashable: Any]).

QueryContainer and structs that conform to the Query protocol allow you to model your queries in a type-safe, Swift-y way. However, not the entirety of Elasticsearch’s query language has been converted, only the most commonly used queries. In case a query you require is missing, feel free to submit a pull request.

func getSearchHandler(_ req: Request) throws -> Future<[Product]> {
  let elasticClient = try req.make(ElasticsearchClient.self)
  let query = QueryContainer(
    Query(
      Fuzzy(key: "name", value: "bolster", transpositions: true)
    )
  )

  return try elasticClient
    .search(index: "product", query, decodeTo: Product.self)
    .map(to: [Product].self) { result in
      return result.hits.hits.map { $0.source }
    }
}

For more examples, please feel free to look into the test suite or comments.

Known issues

  • BoolQuery’s usefulness is significantly hampered by the fact that only one type of queries can be passed as an argument, for example if you pass type [MatchPhrase] as an must argument, other arguments (filter, mustNot, etc.) must also be of type [MatchPhrase]. If you know how to solve this (type erasure?), a pull request is most appreciated!
  • Indices manipulation is not supported.
  • Data manipulation is not supported.
  • Aggregation is not supported.