solr-scala-client

The simple Apache Solr client for Scala. This is based on the SolrJ and provides optimal interface for Scala.

Add the following dependency into your build.sbt to use solr-scala-client.

libraryDependencies += "com.github.takezoe" %% "solr-scala-client" % "0.0.18"

If you want to test SNAPSHOT version, add the following dependency instead of above:

resolvers += "sonatype-oss-snapshot" at "https://oss.sonatype.org/content/repositories/snapshots"

libraryDependencies += "com.github.takezoe" %% "solr-scala-client" % "x.x.x-SNAPSHOT"

This is a simplest example to show usage of solr-scala-client.

import com.github.takezoe.solr.scala._

val client = new SolrClient("http://localhost:8983/solr")

// register
client
  .add(Map("id"->"001", "manu" -> "Lenovo", "name" -> "ThinkPad X201s"))
  .add(Map("id"->"002", "manu" -> "Lenovo", "name" -> "ThinkPad X220"))
  .add(Map("id"->"003", "manu" -> "Lenovo", "name" -> "ThinkPad X121e"))
  .commit

// query
val result = client.query("name: %name%")
  .fields("id", "manu", "name")
  .sortBy("id", Order.asc)
  .getResultAsMap(Map("name" -> "ThinkPad"))

result.documents.foreach { doc: Map[String, Any] =>
  println("id: " + doc("id"))
  println("  manu: " + doc("manu"))
  println("  name: " + doc("name"))
}

It's also possible to use the case class as the search result and parameters instead of Map.

// query
val result = client.query("name: %name%")
  .fields("id", "manu", "name")
  .sortBy("id", Order.asc)
  .getResultAs[Product](Param(name = "ThinkPad"))

result.documents.foreach { product: Product =>
  println("id: " + product.id)
  println("  manu: " + product.manu)
  println("  name: " + product.name)
}

Query Syntax

Following notations are available to embed variables to the query:

• %VAR_NAME% : place holder to set a single word (parameter would be escaped)
• ?VAR_NAME? : place holder to set an expression (&, | and ! are available in an expression)
• $VAR_NAME$ : string replacement (parameter would be not escaped)

See examples of parameterized queries and assembled Solr queries.

// %VAR_NAME% (Single keyword)
client.query("name: %name%").getResultAsMap(Map("name" -> "ThinkPad X201s"))
  // => name:"ThinkPad X201s"

// $VAR_NAME$ (String replacement)
client.query("name: $name$").getResultAsMap(Map("name" -> "ThinkPad AND X201s"))
  // => name:ThinkPad AND X201s

// ?VAR_NAME? (Expression)
client.query("name: ?name?").getResultAsMap(Map("name" -> "ThinkPad & X201s"))
  // => name:("ThinkPad" AND "X201s")

Highlight

Configure the query to return the highlighted content by QueryBuilder#highlight(). The highlighted content is set as the "highlight" property to the Map or the case class.

val result = client.query("content: Scala")
  // NOTE: unique key field is required.
  .fields("id")
  // Specify the highlighted field, prefix and postfix (prefix and postfix are optional).
  .highlight("content", "<strong>", "</strong>")
  .getResultAsMap()

result.documents.foreach { doc: Map[String, Any] =>
  println("id: " + doc("id"))
  println(doc("highlight")) // highlighted content is set as the "highlight" property
}

solr-scala-client expects that the unique key is "id". If your schema has the different field as the unique key, you can specify the unique key name as following:

val result = client.query("content: Scala")
  .id("documentId") // Specify the unique key name
  .fields("documentId")
  .highlight("content", "<strong>", "</strong>")
  .getResultAsMap()

Asynchronous API

solr-scala-client has also asynchronous API based on AsyncHttpCleint.

val client = new AsyncSolrClient("http://localhost:8983/solr")

// Register
client
  .register(Map("id" -> "005", "name" -> "ThinkPad X1 Carbon", "manu" -> "Lenovo"))
  .onComplete{
    case Success(x) => println("registered!")
    case Failure(t) => t.printStackTrace()
  }

// Query
client.query("name:%name%")
  .fields("id", "manu", "name")
  .facetFields("manu")
  .sortBy("id", Order.asc)
  .getResultAsMap(Map("name" -> "ThinkPad X201s"))
  .onComplete {
    case Success(x) => println(x)
    case Failure(t) => t.printStackTrace()
  }

See more example at AsyncSolrClientSample.scala.

Release Notes

0.0.18 - 15 Feb 2018

• Fix response leaking bug

0.0.17 - 5 Dec 2017

• Upgrade to SolrJ-7.1.0
• Switch backend to OkHttp from async-http-client
• Allow specifying collection name when building query
• Add implementation of CloudSolrClient with and without authentication

0.0.16 - 18 Oct 2017

• Upgrade Scala and async-http-client

0.0.15 - 22 Nov 2016

• Scala 2.12 support and library updating

0.0.14 - 14 Aug 2016

• Small refactoring

0.0.13 - 13 Aug 2016

• Upgrade to SolrJ-6.1.0
• Change group id and package name to com.github.takezoe
• Publish to the Maven central repository

0.0.12 - 7 Feb 2015

• Add QueryBuilderBase#fq()
• Add QueryBuilderBase#setRequestHandler()
• Add date facet
• Support for streaming results

0.0.11 - 29 Mar 2014

• Add SolrClient#shutdown()
• QueryBuilder became immutable
• Upgrade solrj version to 4.5.1

0.0.10 - 08 Feb 2014

• Fix escaping in string literal.

0.0.9 - 18 Dec 2013

• Bug fix

0.0.8 - 2 Aug 2013

• Added recommendation search support.
• Added rollback and withTransaction to SolrScalaClient.
• Added Asynchronous API.

0.0.7 - 4 Apr 2013

• Add build for Scala 2.10
• Upgrade to SolrJ 4.2.0
• Support highlighting

0.0.6 - 22 Jan 2013

• Fixed some ExpressionParser bugs.

0.0.5 - 20 Nov 2012

• ExpressionParser became pluggable and added GoogleExpressionParser as an optional implementation of ExpressionParser.
• Converts the full-width space to the half-width space in the expression before calling ExpressionParser.
• Introduced the SolrServer factory. Auth.basic moved to SolrServerFactory.basicAuth and SolrServerFactory.dummy for unit testing.

0.0.4 - 16 Sep 2012

• Expanding expression to the Solr query by ?VAR_NAME? in SolrClient#query().
• Bug fix

0.0.3 - 16 Aug 2012

• Added case class support in update operations.
• Added commit() method to SolrClient.

0.0.2 - 27 May 2012

• Added initializer which configures SolrClient.
• Added basic authentication support as initializer.
• Added facet search support.
• Added case class support as query results and query parameters.

0.0.1 - 4 May 2012

• Initial public release.