Skip to content

Latest commit

 

History

History
638 lines (480 loc) · 20.9 KB

README.md

File metadata and controls

638 lines (480 loc) · 20.9 KB

Jest

Jest is a Java HTTP Rest client for ElasticSearch.

ElasticSearch is an Open Source (Apache 2), Distributed, RESTful, Search Engine built on top of Apache Lucene.

ElasticSearch already has a Java API which is also used by ElasticSearch internally, but Jest fills a gap, it is the missing client for ElasticSearch Http Rest interface.

Read great introduction to ElasticSearch and Jest from IBM Developer works.

Installation

Stable

Jest maven repository is hosted on Sonatype which is then synced to Maven Central with a slight delay.

To get the latest stable version without waiting for Maven Central sync add Sonatype repository definition to your root pom.xml

<repositories>
.
.
 <repository>
   <id>sonatype</id>
   <name>Sonatype Groups</name>
   <url>https://oss.sonatype.org/content/groups/public/</url>
 </repository>
.
.
</repositories>

Add Jest as a dependency to your project.

<dependency>
  <groupId>io.searchbox</groupId>
  <artifactId>jest</artifactId>
  <version>5.3.3</version>
</dependency>

Snapshot

Jest also publishes a snapshot version on Sonatype Snapshot Repository after every push resulting in a successfull build.

To get the latest snapshot version add Sonatype Snapshot Repository definition to your root pom.xml

<repositories>
.
.
  <repository>
    <id>snapshots-repo</id>
    <url>https://oss.sonatype.org/content/repositories/snapshots</url>
    <releases><enabled>false</enabled></releases>
    <snapshots><enabled>true</enabled></snapshots>
  </repository>
.
.
</repositories>

Add Jest snapshot as a dependency to your project.

<dependency>
  <groupId>io.searchbox</groupId>
  <artifactId>jest</artifactId>
  <version>2.0.1-SNAPSHOT</version>
</dependency>

Ensure to check the Changelog.

Usage

Start using Jest by simply creating a JestClient instance:

 // Construct a new Jest client according to configuration via factory
 JestClientFactory factory = new JestClientFactory();
 factory.setHttpClientConfig(new HttpClientConfig
                        .Builder("http://localhost:9200")
                        .multiThreaded(true)
			//Per default this implementation will create no more than 2 concurrent connections per given route
			.defaultMaxTotalConnectionPerRoute(<YOUR_DESIRED_LEVEL_OF_CONCURRENCY_PER_ROUTE>)
			// and no more 20 connections in total
			.maxTotalConnection(<YOUR_DESIRED_LEVEL_OF_CONCURRENCY_TOTAL>)
                        .build());
 JestClient client = factory.getObject();

JestClient is designed to be singleton, don't construct it for each request!

Creating an Index

To create an index just pass the associated CreateIndex action to the client:

client.execute(new CreateIndex.Builder("articles").build());

Index settings can also be passed during the creation by

  • using a JSON formatted string:
String settings = "\"settings\" : {\n" +
                "        \"number_of_shards\" : 5,\n" +
                "        \"number_of_replicas\" : 1\n" +
                "    }\n";

client.execute(new CreateIndex.Builder("articles").settings(Settings.builder().loadFromSource(settings).build().getAsMap()).build());
  • using the SettingsBuilder helper class from Elasticsearch:
import org.elasticsearch.common.settings.Settings;
.
.

Settings.Builder settingsBuilder = Settings.builder();
settingsBuilder.put("number_of_shards",5);
settingsBuilder.put("number_of_replicas",1);

client.execute(new CreateIndex.Builder("articles").settings(settingsBuilder.build().getAsMap()).build());

Add ElasticSearch dependency to use Settings api

Creating an Index Mapping

An index mapping can be created via Jest with ease, just pass the mapping source as a JSON formatted string.

PutMapping putMapping = new PutMapping.Builder(
        "my_index",
        "my_type",
        "{ \"my_type\" : { \"properties\" : { \"message\" : {\"type\" : \"string\", \"store\" : \"yes\"} } } }"
).build();
client.execute(putMapping);

The helper class DocumentMapper.Builder from Elasticsearch can also be used to create the mapping source.

import org.elasticsearch.index.mapper.DocumentMapper;
import org.elasticsearch.index.mapper.core.StringFieldMapper;
import org.elasticsearch.index.mapper.object.RootObjectMapper;
.
.
RootObjectMapper.Builder rootObjectMapperBuilder = new RootObjectMapper.Builder("my_mapping_name").add(
        new StringFieldMapper.Builder("message").store(true)
);
DocumentMapper documentMapper = new DocumentMapper.Builder("my_index", null, rootObjectMapperBuilder).build(null);
String expectedMappingSource = documentMapper.mappingSource().toString();
PutMapping putMapping = new PutMapping.Builder(
        "my_index",
        "my_type",
        expectedMappingSource
).build();
client.execute(putMapping);

Add ElasticSearch dependency to use DocumentMapper.Builder api

Indexing Documents

ElasticSearch requires index data as JSON. There are several ways to create documents to index via Jest. From now on, we will refer documents as source. Source objects can be String, Map or POJOs.

as JSON String;

String source = "{\"user\":\"kimchy\"}";

or creating JSON via ElasticSearch JSONBuilder;

String source = jsonBuilder()
.startObject()
.field("user", "kimchy")
.field("postDate", "date")
.field("message", "trying out Elastic Search")
.endObject().string();

as Map;

Map<String, String> source = new LinkedHashMap<String,String>();
source.put("user", "kimchy");

as POJO;

Article source = new Article();
source.setAuthor("John Ronald Reuel Tolkien");
source.setContent("The Lord of the Rings is an epic high fantasy novel");

An example of indexing given source to twitter index with type tweet;

Index index = new Index.Builder(source).index("twitter").type("tweet").build();
client.execute(index);

Index id can be typed explicitly;

Index index = new Index.Builder(source).index("twitter").type("tweet").id("1").build();
client.execute(index);

@JestId annotation can be used to mark a property of a bean as id;

class Article {

@JestId
private String documentId;

}

Now whenever an instance of Article is indexed, index id will be value of documentId.

If @JestId value is null, it will be set the value of ElasticSearch generated "_id".

Jest also supports using JestId annotation on fields with type other than String but the catch is then you will need to manually manage the document IDs.

class NumericArticle {

@JestId
private Long documentId;

}

It should be noted that when a non-String type is used for the documentId, conversion errors may occur unless you manually manage the documentId. For example if a NumericArticle instance without a documentId is indexed then Elasticsearch will assign an automatically generated id to that document but Jest will not be able to convert that id to Long since all id fields are of type String in Elasticsearch and the auto generated id contains non numeric characters. So the non-String type support for JestId annotation is purely for ease of use and should not be used if you plan to use the automatic id generation functionality of Elasticsearch.

Version field support (aka Optimistic Concurrency Control)

_version field for documents are supported via the @JestVersion annotation. Marking a property of your bean with @JestVersion will cause that field to be serialized and deserialized as _version in the document JSON communicated to Elasticsearch. This in turn lets you use the Optimistic Concurrency Control capabilities provided by Elasticsearch.

class Article {

@JestId
private String documentId;

@JestVersion
private Long documentVersion;

}

Searching Documents

Search queries can be either JSON String or created by ElasticSearch SourceBuilder Jest works with default ElasticSearch queries, it simply keeps things as is.

As JSON;

String query = "{\n" +
            "    \"query\": {\n" +
            "        \"filtered\" : {\n" +
            "            \"query\" : {\n" +
            "                \"query_string\" : {\n" +
            "                    \"query\" : \"test\"\n" +
            "                }\n" +
            "            },\n" +
            "            \"filter\" : {\n" +
            "                \"term\" : { \"user\" : \"kimchy\" }\n" +
            "            }\n" +
            "        }\n" +
            "    }\n" +
            "}";

Search search = new Search.Builder(query)
                // multiple index or types can be added.
                .addIndex("twitter")
                .addType("tweet")
                .build();

SearchResult result = client.execute(search);

By template;

String query = "{\n" +
            "    \"id\": \"myTemplateId\"," +
            "    \"params\": {\n" +
            "        \"query_string\" : \"search for this\"" +
            "    }\n" +
            "}";

Search search = new Search.TemplateBuilder(query)
                // multiple index or types can be added.
                .addIndex("twitter")
                .addType("tweet")
                .build();

SearchResult result = client.execute(search);

Also supports inline search templates and file-based templates.

By using SearchSourceBuilder;

SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
searchSourceBuilder.query(QueryBuilders.matchQuery("user", "kimchy"));

Search search = new Search.Builder(searchSourceBuilder.toString())
                                // multiple index or types can be added.
                                .addIndex("twitter")
                                .addType("tweet")
                                .build();

SearchResult result = client.execute(search);

Add ElasticSearch dependency to use SearchSourceBuilder

Result can be cast to List of domain object;

SearchResult result = client.execute(search);
List<SearchResult.Hit<Article, Void>> hits = result.getHits(Article.class);
// or
List<Article> articles = result.getSourceAsObjectList(Article.class);

Please refer ElasticSearch Query DSL documentation to work with complex queries.

Getting Documents

Get get = new Get.Builder("twitter", "1").type("tweet").build();

JestResult result = client.execute(get);

Result can be cast to domain object;

Get get = new Get.Builder("twitter", "1").type("tweet").build();

JestResult result = client.execute(get);

Article article = result.getSourceAsObject(Article.class);

Updating Documents

String script = "{\n" +
                "    \"script\" : \"ctx._source.tags += tag\",\n" +
                "    \"params\" : {\n" +
                "        \"tag\" : \"blue\"\n" +
                "    }\n" +
                "}";

client.execute(new Update.Builder(script).index("twitter").type("tweet").id("1").build());

Deleting Documents

client.execute(new Delete.Builder("1")
                .index("twitter")
                .type("tweet")
                .build());

Bulk Operations

ElasticSearch's bulk API makes it possible to perform many index/delete operations in a single API call. This can greatly increase the indexing speed.

Bulk bulk = new Bulk.Builder()
    .defaultIndex("twitter")
    .defaultType("tweet")
    .addAction(new Index.Builder(article1).build())
    .addAction(new Index.Builder(article2).build())
    .addAction(new Delete.Builder("1").index("twitter").type("tweet").build())
    .build();

client.execute(bulk);

List of objects can be indexed via bulk api

String article1 = "tweet1";
String article2 = "tweet2";

Bulk bulk = new Bulk.Builder()
                .defaultIndex("twitter")
                .defaultType("tweet")
                .addAction(Arrays.asList(
                    new Index.Builder(article1).build(),
                    new Index.Builder(article2).build()))
                .build();

client.execute(bulk);

Action Parameters

ElasticSearch offers request parameters to set properties like routing, versioning, operation type etc.

For example refresh parameter can be set to true while indexing a document as below:

Index index = new Index.Builder("{\"user\":\"kimchy\"}")
    .index("cvbank")
    .type("candidate")
    .id("1")
    .setParameter(Parameters.REFRESH, true),
    .build();
client.execute(index);

Any request parameter that is passed through URL can be set this way. Only the commonly used parameters are enumarated in Parameters class, raw strings can be used in place of the non-enumarated parameters.

Asynchronous Execution

Jest http client supports execution of any action asynchronously with non blocking IO.

Following example illustrates how to execute an action with Jest asynchronous call.

client.executeAsync(action,new JestResultHandler<JestResult>() {
    @Override
    public void completed(JestResult result) {
        ... do process result ....
    }
    @Override
    public void failed(Exception ex) {
       ... catch exception ...
    }
});

Node Discovery through Nodes API

Enabling node discovery will (poll) and update the list of servers in the client periodically. Configuration of the discovery process can be done in the client config as follows:

//enable host discovery
ClientConfig clientConfig = new ClientConfig.Builder("http://localhost:9200")
    .discoveryEnabled(true)
    .discoveryFrequency(1l, TimeUnit.MINUTES)
    .build();

Setting a filter on the nodes to discover will allow you specify they types of nodes to discover, with the same syntax as outlined in Node Specification for Elasticsearch. For example:

//enable host discovery
ClientConfig clientConfig = new ClientConfig.Builder("http://localhost:9200")
    .discoveryEnabled(true)
    .discoveryFrequency(1l, TimeUnit.MINUTES)
    .discoveryFilter("type:arbitrary")
    .build();

Authentication

Basic username and password authentication can be configured when constructing the client; it should be noted that these credentials will be used for all servers provided and discovered.

JestClientFactory factory = new JestClientFactory();
factory.setHttpClientConfig(
    new HttpClientConfig.Builder("http://localhost:9200")
        .defaultCredentials("global_user", "global_password")
        .build()
);

If your authentication needs are more complicated than above (e.g.: different credentials for different servers, Kerberos etc.) then you can also provide a CredentialsProvider instance.

BasicCredentialsProvider customCredentialsProvider = new BasicCredentialsProvider();
customCredentialsProvider.setCredentials(
        new AuthScope("192.168.0.88", 9200),
        new UsernamePasswordCredentials("eu_user", "123")
);
customCredentialsProvider.setCredentials(
        new AuthScope("192.168.0.172", 9200),
        new UsernamePasswordCredentials("us_user", "456")
);

JestClientFactory factory = new JestClientFactory();
factory.setHttpClientConfig(
    new HttpClientConfig.Builder(Arrays.asList("http://192.168.0.88:9200", "http://192.168.0.172:9200"))
        .credentialsProvider(customCredentialsProvider)
        .build()
);

HTTPS / SSL

HTTPS or SSL (or TLS) connections can be configured by passing your own instance of LayeredConnectionSocketFactory to the builder.

// trust ALL certificates
SSLContext sslContext = new SSLContextBuilder().loadTrustMaterial(null, new TrustStrategy() {
    public boolean isTrusted(X509Certificate[] arg0, String arg1) throws CertificateException {
        return true;
    }
}).build();

// skip hostname checks
HostnameVerifier hostnameVerifier = NoopHostnameVerifier.INSTANCE;

SSLConnectionSocketFactory sslSocketFactory = new SSLConnectionSocketFactory(sslContext, hostnameVerifier);
SchemeIOSessionStrategy httpsIOSessionStrategy = new SSLIOSessionStrategy(sslContext, hostnameVerifier);

JestClientFactory factory = new JestClientFactory();
factory.setHttpClientConfig(new HttpClientConfig.Builder("https://localhost:9200")
                .defaultSchemeForDiscoveredNodes("https") // required, otherwise uses http
                .sslSocketFactory(sslSocketFactory) // this only affects sync calls
                .httpsIOSessionStrategy(httpsIOSessionStrategy) // this only affects async calls
                .build()
);

Keep in mind that (the SSLContext and HostnameVerifier in) above example is just for example and very insecure as it is.

Proxy

Any system-wide proxy setting will be used by default; so if the proxy is set on the system level (e.g.: through OS or environment variables) then you don't need to do any further configuration on the Jest side.

Configuring proxy settings exclusively for Jest can also be done through the builder.

String proxyHost = "proxy.company.com";
int proxyPort = 7788;

JestClientFactory factory = new JestClientFactory();
factory.setHttpClientConfig(
    new HttpClientConfig.Builder("http://remote.server.com:9200")
        .proxy(new HttpHost(poxyHost, proxyPort))
        .build()
);

Further Reading

Integration Tests are best places to see things in action.

Logging

Jest uses slf4j for logging purposes and expects an implementation to be provided at runtime, therefore the log4j dependency is in provided scope.

For instance to use log4j implementation, add below dependency to your pom.xml

<dependency>
 <groupId>org.slf4j</groupId>
	<artifactId>slf4j-log4j12</artifactId>
	<version>1.6.1</version>
</dependency>

Please read slf4j manual here.

ElasticSearch Optional Dependency

If you want to use ElasticSearch's QueryBuilder or Settings classes, ensure to add ElasticSearch dependency.

<dependency>
    <groupId>org.elasticsearch</groupId>
    <artifactId>elasticsearch</artifactId>
    <version>${elasticsearch.version}</version>
</dependency>

Comparison to native API

There are several alternative clients available when working with ElasticSearch from Java, like Jest that provides a POJO marshalling mechanism on indexing and for the search results. In this example we are using the Client that is included in ElasticSearch. By default the client doesn't use the REST API but connects to the cluster as a normal node that just doesn't store any data. It knows about the state of the cluster and can route requests to the correct node but supposedly consumes more memory. For our application this doesn't make a huge difference but for production systems that's something to think about. -- Florian Hopf

So if you have several ES clusters running different versions, then using the native (or transport) client will be a problem, and you will need to go HTTP (and Jest is the main option I think). If versioning is not an issue, the native client will be your best option as it is cluster aware (thus knows how to route your queries and does not need another hop), and also moves some computation away from your ES cluster (like merging search results that will be done locally instead of on the data node). -- Rotem Hermon

ElasticSearch does not have Java rest client. It has only native client comes built in. That is the gap. You can add security layer to HTTP but native API. That is why none of SAAS offerings can be used with native api. -- Searchly

Thanks

Thanks to JetBrains for providing a license for IntelliJ IDEA to develop this project.

Copyright and License

Copyright 2013 www.searchly.com

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this work except in compliance with the License. You may obtain a copy of the License in the LICENSE file, or at:

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.