第九讲—ES Java API（一）

一、Client

        使用Java API进行操作必须要有一个Client。

Client client = null ; Settings settings = ImmutableSettings.settingsBuilder()                      .put( "client.transport.sniff" , true )                      .put( "cluster.name" , "Minutch" )                      .build(); TransportClient transportClient = new TransportClient(settings); transportClient.addTransportAddress( new InetSocketTransportAddress( "127.0.0.1" , 9300 )); client = transportClient;

二、Index

        搜索引擎一个核心就是建索引，索引我们看下使用Java API如何建索引。

        1.使用Elasticsearch提供的JSON生成工具

//jsonBuilder XContentBuilder builder = jsonBuilder()      .startObject()          .field( "user" , "kimchy" )          .field( "postDate" , new Date())          .field( "message" , "trying out Elasticsearch" )      .endObject()   //如果primaryKeyValue这个参数不要，Elasticsearch会自动生成一个id IndexRequestBuilder indexRequestBuilder = client.prepareIndex(indexName,typeName,primaryKeyValue).setSource(builder); //创建索引 IndexResponse indexResponse = indexRequestBuilder.execute().actionGet(); /******indexResponse对象会提供很多信息**************/ // Index name String _index = indexResponse.getIndex(); // Type name String _type = indexResponse.getType(); // Document ID (generated or not) String _id = indexResponse.getId(); // Version (if it's the first time you index this document, you will get: 1) long _version = indexResponse.getVersion(); // isCreated() is true if the document is a new one, false if it has been updated boolean created = indexResponse.isCreated();

         2.使用JSON字符串

String json = "{" +          "\"user\":\"kimchy\"," +          "\"postDate\":\"2013-01-30\"," +          "\"message\":\"trying out Elasticsearch\"" +      "}" ;   IndexResponse response = client.prepareIndex( "twitter" , "tweet" )          .setSource(json)          .execute()          .actionGet();

      3.使用Map

Map<String,Object> map = new HashMap<String,Object>(); IndexResponse response = client.prepareIndex( "twitter" , "tweet" )          .setSource(map)          .execute()          .actionGet();

三、Get

        创建完索引，我们可以通过ID把那条纪录搜索出来，prepareGet方式的查询是基于ID的。

GetResponse response = client.prepareGet( "twitter" , "tweet" , "1" )          .execute()          .actionGet();

        setOperationThreaded：默认为true，表示该查询在多个线程上执行。

GetResponse response = client.prepareGet( "twitter" , "tweet" , "1" )          .setOperationThreaded( false )          .execute()          .actionGet();

四、DELETE 

        删除索引，prepareDelete的删除方式是基于ID的删除

DeleteResponse response = client.prepareDelete( "twitter" , "tweet" , "1" )          .execute()          .actionGet();

        setOperationThreaded：默认为true，表示该查询在多个线程上执行。

DeleteResponse response = client.prepareDelete( "twitter" , "tweet" , "1" )          .setOperationThreaded( false )          .execute()          .actionGet();

五 、Update

        更新索引，实际上他还是先删掉了原来的索引，然后重新建了所有，但是版本号会增加。

UpdateRequest updateRequest = new UpdateRequest(); updateRequest.index( "index" ); updateRequest.type( "type" ); updateRequest.id( "1" ); updateRequest.doc(jsonBuilder()          .startObject()              .field( "gender" , "male" )          .endObject()); client.update(updateRequest).get();

client.prepareUpdate( "ttl" , "doc" , "1" )          .setDoc(jsonBuilder()                            .startObject()                  .field( "gender" , "male" )              .endObject())          .get();

        3.使用script，我们可以写脚本文件，在使用时指定ScriptService.ScriptType.FILE

//直接使用脚本 client.prepareUpdate( "ttl" , "doc" , "1" )          .setScript( "ctx._source.gender = \"male\""  , ScriptService.ScriptType.INLINE)          .get();

UpdateRequest updateRequest = new UpdateRequest( "ttl" , "doc" , "1" )          .script( "ctx._source.gender = \"male\"" ); client.update(updateRequest).get();

        4.如果我们要更新的数据不存在时，我们新增一条指定数据。

IndexRequest indexRequest = new IndexRequest( "index" , "type" , "1" )          .source(jsonBuilder()              .startObject()                  .field( "name" , "Joe Smith" )                  .field( "gender" , "male" )              .endObject()); UpdateRequest updateRequest = new UpdateRequest( "index" , "type" , "1" )          .doc(jsonBuilder()              .startObject()                  .field( "gender" , "male" )              .endObject())          .upsert(indexRequest);              client.update(updateRequest).get();

六、Bulk

       批量操作，一次请求处理多个文档的操作.

BulkRequestBuilder bulkRequest = client.prepareBulk();   // either use client#prepare, or use Requests# to directly build index/delete requests bulkRequest.add(client.prepareIndex( "twitter" , "tweet" , "1" )          .setSource(jsonBuilder()                      .startObject()                          .field( "user" , "kimchy" )                          .field( "postDate" , new Date())                          .field( "message" , "trying out Elasticsearch" )                      .endObject()                    )          );   bulkRequest.add(client.prepareIndex( "twitter" , "tweet" , "2" )          .setSource(jsonBuilder()                      .startObject()                          .field( "user" , "kimchy" )                          .field( "postDate" , new Date())                          .field( "message" , "another post" )                      .endObject()                    )          );   BulkResponse bulkResponse = bulkRequest.execute().actionGet(); if (bulkResponse.hasFailures()) {      // process failures by iterating through each bulk response item }

        bulkProcessor:当满足一定条件后，自动发送批量的请求，比如：已经有100个request了，或者每分钟发送一次，或者大小超过5mb

BulkProcessor bulkProcessor = BulkProcessor.builder(          client,           new BulkProcessor.Listener() {                            //1.bulk执行前被触发，比如你可以看看当前bulk有多少个action：request.numberOfActions()              @Override              public void beforeBulk( long executionId,                                     BulkRequest request) { ... }                //2.bulk执行后被触发，你可以看看是否有Actions失败：esponse.hasFailures()              @Override              public void afterBulk( long executionId,                                    BulkRequest request,                                    BulkResponse response) { ... }              //3.当bulk抛出异常时被触发              @Override              public void afterBulk( long executionId,                                    BulkRequest request,                                    Throwable failure) { ... }          })          //4.每1000个actions就执行bulk          .setBulkActions( 10000 )           //5.满1GB就执行bulk          .setBulkSize( new ByteSizeValue( 1 , ByteSizeUnit.GB))           //6.每5秒发送一次bulk          .setFlushInterval(TimeValue.timeValueSeconds( 5 ))           //7.设置的并发请求数。值为0意味着只有一个请求可以被执行。值为1表示1个并发请求可以同时积累新的批量要求执行。          .setConcurrentRequests( 1 )          .build();   //你可以向下面一样往bulkProcessor里添加action bulkProcessor.add( new IndexRequest( "twitter" , "tweet" , "1" ).source( /* your doc here */ )); bulkProcessor.add( new DeleteRequest( "twitter" , "tweet" , "2" ));

        默认的BulkProcessor配置如下：

sets bulkActions to 1000 sets bulkSize to 5mb does not set flushInterval sets concurrentRequests to 1

        关闭bulkProcessor的两种方式

bulkProcessor.awaitClose( 10 , TimeUnit.MINUTES);   bulkProcessor.close();

        两个方法关闭后会将剩余的请求发送出去，并且不在接收新增的action.

        如果并发请求可用，并且使用的是awaitClose，在时间内完成了所有的请求，则返回true。