Elasticsearch：RestClient+SearchSourceBuilder使用案例

1 前言

RestClient是較低層的API，這里使用基於其封裝的高層API，即RestHighLevelClient。

需要添加的依賴如下：

1.  
   <dependency>
2.  
   <groupId>org.elasticsearch</groupId>
3.  
   <artifactId>elasticsearch</artifactId>
4.  
   <version>5.6.10</version>
5.  
   </dependency>
6.  
   <dependency>
7.  
   <groupId>org.elasticsearch.client</groupId>
8.  
   <artifactId>elasticsearch-rest-high-level-client</artifactId>
9.  
   <version>5.6.10</version>
10.  
    </dependency>

較低版本的es可能不支持RestHighLevelClient。
查看一下其maven中央倉庫：https://mvnrepository.com/artifact/org.elasticsearch.client/elasticsearch-rest-high-level-client
可以看到，5.6以上的版本才有。

2 測試案例

測試代碼如下：

1.  
   package com.xpleaf.es.leaf;
2.  
    
3.  
   import org.apache.http.HttpHost;
4.  
   import org.elasticsearch.action.search.SearchRequest;
5.  
   import org.elasticsearch.action.search.SearchResponse;
6.  
   import org.elasticsearch.client.RestClient;
7.  
   import org.elasticsearch.client.RestHighLevelClient;
8.  
   import org.elasticsearch.common.unit.TimeValue;
9.  
   import org.elasticsearch.index.query.*;
10.  
    import org.elasticsearch.search.builder.SearchSourceBuilder;
11.  
    import org.elasticsearch.search.sort.SortOrder;
12.  
    import org.junit.After;
13.  
    import org.junit.Before;
14.  
    import org.junit.Test;
15.  
     
16.  
    /**
17.  
    * @author xpleaf
18.  
    * @GitHub https://github.com/xpleaf
19.  
    * @Blog https://blog.51cto.com/xpleaf
20.  
    * @date 2018/10/7 下午12:05
21.  
    */
22.  
    public class RestHighLevelClientTest {
23.  
     
24.  
    private HttpHost[] esHosts = new HttpHost[]{
25.  
    new HttpHost("localhost", 9200)
26.  
    };
27.  
    private RestClient restClient = null;
28.  
    private RestHighLevelClient client = null;
29.  
    private BoolQueryBuilder boolQueryBuilder = null;
30.  
     
31.  
    @Before
32.  
    public void init() throws Exception {
33.  
    // 1.創建RestClient對象
34.  
    restClient = RestClient.builder(esHosts).build();
35.  
    client = new RestHighLevelClient(restClient);
36.  
    // 2.創建BoolQueryBuilder對象
37.  
    boolQueryBuilder = new BoolQueryBuilder();
38.  
    // 3.設置boolQueryBuilder條件
39.  
    MatchPhraseQueryBuilder matchPhraseQueryBuilder = QueryBuilders
40.  
    .matchPhraseQuery( "key_word", "廣東");
41.  
    MatchPhraseQueryBuilder matchPhraseQueryBuilder2 = QueryBuilders
42.  
    .matchPhraseQuery( "key_word", "湖人");
43.  
    RangeQueryBuilder rangeQueryBuilder = QueryBuilders
44.  
    .rangeQuery( "postdate")
45.  
    .from( "2016-01-01 00:00:00");
46.  
    // 子boolQueryBuilder條件條件，用來表示查詢條件or的關系
47.  
    BoolQueryBuilder childBoolQueryBuilder = new BoolQueryBuilder()
48.  
    .should(matchPhraseQueryBuilder)
49.  
    .should(matchPhraseQueryBuilder2);
50.  
    // 4.添加查詢條件到boolQueryBuilder中
51.  
    boolQueryBuilder
52.  
    .must(childBoolQueryBuilder)
53.  
    .must(rangeQueryBuilder);
54.  
    }
55.  
     
56.  
    // 測試SearchSourceBuilder的搜索
57.  
    @Test
58.  
    public void test01() throws Exception {
59.  
     
60.  
    // 1.創建並設置SearchSourceBuilder對象
61.  
    SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
62.  
    // 查詢條件--->生成DSL查詢語句
63.  
    searchSourceBuilder.query(boolQueryBuilder);
64.  
    // 第幾頁
65.  
    searchSourceBuilder.from( 0);
66.  
    // 每頁多少條數據
67.  
    searchSourceBuilder.size( 100);
68.  
    // 獲取的字段（列）和不需要獲取的列
69.  
    searchSourceBuilder.fetchSource( new String[]{"postdate", "key_word"}, new String[]{});
70.  
    // 設置排序規則
71.  
    searchSourceBuilder.sort( "postdate", SortOrder.ASC);
72.  
    // 設置超時時間為2s
73.  
    searchSourceBuilder.timeout( new TimeValue(2000));
74.  
     
75.  
    // 2.創建並設置SearchRequest對象
76.  
    SearchRequest searchRequest = new SearchRequest();
77.  
    // 設置request要搜索的索引和類型
78.  
    searchRequest.indices( "spnews").types("news");
79.  
    // 設置SearchSourceBuilder查詢屬性
80.  
    searchRequest.source(searchSourceBuilder);
81.  
     
82.  
    // 3.查詢
83.  
    SearchResponse searchResponse = client.search(searchRequest);
84.  
    System.out.println(searchResponse.toString());
85.  
    }
86.  
     
87.  
    @After
88.  
    public void after() throws Exception {
89.  
    restClient.close();
90.  
    }
91.  
     
92.  
    }

3 分析

3.1 Rest Json

上面測試案例的查詢條件：

1.  
   // 2.創建BoolQueryBuilder對象
2.  
   boolQueryBuilder = new BoolQueryBuilder();
3.  
   // 3.設置boolQueryBuilder條件
4.  
   MatchPhraseQueryBuilder matchPhraseQueryBuilder = QueryBuilders
5.  
   .matchPhraseQuery( "key_word", "廣東");
6.  
   MatchPhraseQueryBuilder matchPhraseQueryBuilder2 = QueryBuilders
7.  
   .matchPhraseQuery( "key_word", "湖人");
8.  
   RangeQueryBuilder rangeQueryBuilder = QueryBuilders
9.  
   .rangeQuery( "postdate")
10.  
    .from( "2016-01-01 00:00:00");
11.  
    // 子boolQueryBuilder條件條件，用來表示查詢條件or的關系
12.  
    BoolQueryBuilder childBoolQueryBuilder = new BoolQueryBuilder()
13.  
    .should(matchPhraseQueryBuilder)
14.  
    .should(matchPhraseQueryBuilder2);
15.  
    // 4.添加查詢條件到boolQueryBuilder中
16.  
    boolQueryBuilder
17.  
    .must(childBoolQueryBuilder)
18.  
    .must(rangeQueryBuilder);

實際上會轉化為如下的es查詢語句（可以debug一下，使用searchSourceBuilder就是用來做這種轉換）：

1.  
   {
2.  
   "from" : 0,
3.  
   "size" : 100,
4.  
   "timeout" : "2000ms",
5.  
   "query" : {
6.  
   "bool" : {
7.  
   "must" : [
8.  
   {
9.  
   "bool" : {
10.  
    "should" : [
11.  
    {
12.  
    "match_phrase" : {
13.  
    "key_word" : {
14.  
    "query" : "廣東",
15.  
    "slop" : 0,
16.  
    "boost" : 1.0
17.  
    }
18.  
    }
19.  
    },
20.  
    {
21.  
    "match_phrase" : {
22.  
    "key_word" : {
23.  
    "query" : "湖人",
24.  
    "slop" : 0,
25.  
    "boost" : 1.0
26.  
    }
27.  
    }
28.  
    }
29.  
    ],
30.  
    "disable_coord" : false,
31.  
    "adjust_pure_negative" : true,
32.  
    "boost" : 1.0
33.  
    }
34.  
    },
35.  
    {
36.  
    "range" : {
37.  
    "postdate" : {
38.  
    "from" : "2016-01-01 00:00:00",
39.  
    "to" : null,
40.  
    "include_lower" : true,
41.  
    "include_upper" : true,
42.  
    "boost" : 1.0
43.  
    }
44.  
    }
45.  
    }
46.  
    ],
47.  
    "disable_coord" : false,
48.  
    "adjust_pure_negative" : true,
49.  
    "boost" : 1.0
50.  
    }
51.  
    },
52.  
    "_source" : {
53.  
    "includes" : [
54.  
    "postdate",
55.  
    "key_word"
56.  
    ],
57.  
    "excludes" : [ ]
58.  
    },
59.  
    "sort" : [
60.  
    {
61.  
    "postdate" : {
62.  
    "order" : "asc"
63.  
    }
64.  
    }
65.  
    ]
66.  
    }

3.2 match query VS match_phrase query

注意其差別：

• match query：會對查詢語句進行分詞，分詞后查詢語句中的任何一個詞項被匹配，文檔就會被搜索到。如果想查詢匹配所有關鍵詞的文檔，可以用and操作符連接；
• match_phrase query：滿足下面兩個條件才會被搜索到
  • （1）分詞后所有詞項都要出現在該字段中
  • （2）字段中的詞項順序要一致

1 前言

RestClient是較低層的API，這里使用基於其封裝的高層API，即RestHighLevelClient。

需要添加的依賴如下：

1.  
   <dependency>
2.  
   <groupId>org.elasticsearch </groupId>
3.  
   <artifactId>elasticsearch </artifactId>
4.  
   <version>5.6.10 </version>
5.  
   </dependency>
6.  
   <dependency>
7.  
   <groupId>org.elasticsearch.client </groupId>
8.  
   <artifactId>elasticsearch-rest-high-level-client </artifactId>
9.  
   <version>5.6.10 </version>
10.  
    </dependency>

較低版本的es可能不支持RestHighLevelClient。
查看一下其maven中央倉庫：https://mvnrepository.com/artifact/org.elasticsearch.client/elasticsearch-rest-high-level-client
可以看到，5.6以上的版本才有。

2 測試案例

測試代碼如下：

1.  
   package com.xpleaf.es.leaf;
2.  
    
3.  
   import org.apache.http.HttpHost;
4.  
   import org.elasticsearch.action.search.SearchRequest;
5.  
   import org.elasticsearch.action.search.SearchResponse;
6.  
   import org.elasticsearch.client.RestClient;
7.  
   import org.elasticsearch.client.RestHighLevelClient;
8.  
   import org.elasticsearch.common.unit.TimeValue;
9.  
   import org.elasticsearch.index.query.*;
10.  
    import org.elasticsearch.search.builder.SearchSourceBuilder;
11.  
    import org.elasticsearch.search.sort.SortOrder;
12.  
    import org.junit.After;
13.  
    import org.junit.Before;
14.  
    import org.junit.Test;
15.  
     
16.  
    /**
17.  
    * @author xpleaf
18.  
    * @GitHub https://github.com/xpleaf
19.  
    * @Blog https://blog.51cto.com/xpleaf
20.  
    * @date 2018/10/7 下午12:05
21.  
    */
22.  
    public class RestHighLevelClientTest {
23.  
     
24.  
    private HttpHost[] esHosts = new HttpHost[]{
25.  
    new HttpHost( "localhost", 9200)
26.  
    };
27.  
    private RestClient restClient = null;
28.  
    private RestHighLevelClient client = null;
29.  
    private BoolQueryBuilder boolQueryBuilder = null;
30.  
     
31.  
    @Before
32.  
    public void init() throws Exception {
33.  
    // 1.創建RestClient對象
34.  
    restClient = RestClient.builder(esHosts).build();
35.  
    client = new RestHighLevelClient(restClient);
36.  
    // 2.創建BoolQueryBuilder對象
37.  
    boolQueryBuilder = new BoolQueryBuilder();
38.  
    // 3.設置boolQueryBuilder條件
39.  
    MatchPhraseQueryBuilder matchPhraseQueryBuilder = QueryBuilders
40.  
    .matchPhraseQuery( "key_word", "廣東");
41.  
    MatchPhraseQueryBuilder matchPhraseQueryBuilder2 = QueryBuilders
42.  
    .matchPhraseQuery( "key_word", "湖人");
43.  
    RangeQueryBuilder rangeQueryBuilder = QueryBuilders
44.  
    .rangeQuery( "postdate")
45.  
    .from( "2016-01-01 00:00:00");
46.  
    // 子boolQueryBuilder條件條件，用來表示查詢條件or的關系
47.  
    BoolQueryBuilder childBoolQueryBuilder = new BoolQueryBuilder()
48.  
    .should(matchPhraseQueryBuilder)
49.  
    .should(matchPhraseQueryBuilder2);
50.  
    // 4.添加查詢條件到boolQueryBuilder中
51.  
    boolQueryBuilder
52.  
    .must(childBoolQueryBuilder)
53.  
    .must(rangeQueryBuilder);
54.  
    }
55.  
     
56.  
    // 測試SearchSourceBuilder的搜索
57.  
    @Test
58.  
    public void test01() throws Exception {
59.  
     
60.  
    // 1.創建並設置SearchSourceBuilder對象
61.  
    SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
62.  
    // 查詢條件--->生成DSL查詢語句
63.  
    searchSourceBuilder.query(boolQueryBuilder);
64.  
    // 第幾頁
65.  
    searchSourceBuilder.from( 0);
66.  
    // 每頁多少條數據
67.  
    searchSourceBuilder.size( 100);
68.  
    // 獲取的字段（列）和不需要獲取的列
69.  
    searchSourceBuilder.fetchSource( new String[]{ "postdate", "key_word"}, new String[]{});
70.  
    // 設置排序規則
71.  
    searchSourceBuilder.sort( "postdate", SortOrder.ASC);
72.  
    // 設置超時時間為2s
73.  
    searchSourceBuilder.timeout( new TimeValue( 2000));
74.  
     
75.  
    // 2.創建並設置SearchRequest對象
76.  
    SearchRequest searchRequest = new SearchRequest();
77.  
    // 設置request要搜索的索引和類型
78.  
    searchRequest.indices( "spnews").types( "news");
79.  
    // 設置SearchSourceBuilder查詢屬性
80.  
    searchRequest.source(searchSourceBuilder);
81.  
     
82.  
    // 3.查詢
83.  
    SearchResponse searchResponse = client.search(searchRequest);
84.  
    System.out.println(searchResponse.toString());
85.  
    }
86.  
     
87.  
    @After
88.  
    public void after() throws Exception {
89.  
    restClient.close();
90.  
    }
91.  
     
92.  
    }

3 分析

3.1 Rest Json

上面測試案例的查詢條件：

1.  
   // 2.創建BoolQueryBuilder對象
2.  
   boolQueryBuilder = new BoolQueryBuilder();
3.  
   // 3.設置boolQueryBuilder條件
4.  
   MatchPhraseQueryBuilder matchPhraseQueryBuilder = QueryBuilders
5.  
   .matchPhraseQuery( "key_word", "廣東");
6.  
   MatchPhraseQueryBuilder matchPhraseQueryBuilder2 = QueryBuilders
7.  
   .matchPhraseQuery( "key_word", "湖人");
8.  
   RangeQueryBuilder rangeQueryBuilder = QueryBuilders
9.  
   .rangeQuery( "postdate")
10.  
    .from( "2016-01-01 00:00:00");
11.  
    // 子boolQueryBuilder條件條件，用來表示查詢條件or的關系
12.  
    BoolQueryBuilder childBoolQueryBuilder = new BoolQueryBuilder()
13.  
    .should(matchPhraseQueryBuilder)
14.  
    .should(matchPhraseQueryBuilder2);
15.  
    // 4.添加查詢條件到boolQueryBuilder中
16.  
    boolQueryBuilder
17.  
    .must(childBoolQueryBuilder)
18.  
    .must(rangeQueryBuilder);

實際上會轉化為如下的es查詢語句（可以debug一下，使用searchSourceBuilder就是用來做這種轉換）：

1.  
   {
2.  
   "from" : 0,
3.  
   "size" : 100,
4.  
   "timeout" : "2000ms",
5.  
   "query" : {
6.  
   "bool" : {
7.  
   "must" : [
8.  
   {
9.  
   "bool" : {
10.  
    "should" : [
11.  
    {
12.  
    "match_phrase" : {
13.  
    "key_word" : {
14.  
    "query" : "廣東",
15.  
    "slop" : 0,
16.  
    "boost" : 1.0
17.  
    }
18.  
    }
19.  
    },
20.  
    {
21.  
    "match_phrase" : {
22.  
    "key_word" : {
23.  
    "query" : "湖人",
24.  
    "slop" : 0,
25.  
    "boost" : 1.0
26.  
    }
27.  
    }
28.  
    }
29.  
    ],
30.  
    "disable_coord" : false,
31.  
    "adjust_pure_negative" : true,
32.  
    "boost" : 1.0
33.  
    }
34.  
    },
35.  
    {
36.  
    "range" : {
37.  
    "postdate" : {
38.  
    "from" : "2016-01-01 00:00:00",
39.  
    "to" : null,
40.  
    "include_lower" : true,
41.  
    "include_upper" : true,
42.  
    "boost" : 1.0
43.  
    }
44.  
    }
45.  
    }
46.  
    ],
47.  
    "disable_coord" : false,
48.  
    "adjust_pure_negative" : true,
49.  
    "boost" : 1.0
50.  
    }
51.  
    },
52.  
    "_source" : {
53.  
    "includes" : [
54.  
    "postdate",
55.  
    "key_word"
56.  
    ],
57.  
    "excludes" : [ ]
58.  
    },
59.  
    "sort" : [
60.  
    {
61.  
    "postdate" : {
62.  
    "order" : "asc"
63.  
    }
64.  
    }
65.  
    ]
66.  
    }

3.2 match query VS match_phrase query

注意其差別：

• match query：會對查詢語句進行分詞，分詞后查詢語句中的任何一個詞項被匹配，文檔就會被搜索到。如果想查詢匹配所有關鍵詞的文檔，可以用and操作符連接；
• match_phrase query：滿足下面兩個條件才會被搜索到
  • （1）分詞后所有詞項都要出現在該字段中
  • （2）字段中的詞項順序要一致