Elasticsearch 7 父/子关系建立及查询(含JAVA操作)

旧城等待， 2023-05-29 03:36 5阅读 0赞

Elasticsearch 7 由于取消了type的定义，因此没有以前表的概念，定义父/子关系与以前不同，

下面是一个例子

`Elasticsearch 7使用join`数据类型来创建相同的索引文件中的父/子关系。

下面的relation\_field定义了父子关系，question为父，answer为子

>     PUT my_index
>     {
>       "mappings": {
>         "properties": {
>           "relation_field": { 
>             "type": "join",
>             "relations": {
>               "question": "answer" 
>             }
>           }
>         }
>       }
>     }

如果存在多个或者多级父子关系，则按下面的方法定义

*  question是answer和comment的父
 *  answer 是 vote的父

>     PUT my_index
>     {
>       "mappings": {
>         "properties": {
>           "my_join_field": {
>             "type": "join",
>             "relations": {
>               "question": ["answer", "comment"],  
>               "answer": "vote" 
>             }
>           }
>         }
>       }
>     }

增加父记录：

>     PUT my_index/_doc/1?refresh
>     {
>       "text": "This is a question",
>       "relation_field": "question" 
>     }
>     
>     PUT my_index/_doc/2?refresh
>     {
>       "text": "This is another question",
>       "relation_field": "question"
>     }

增加子记录

*  这里使用routing=1，1为父文档的id，routing决定了文档的分片
 *  需要在同一分片中索引父项，因此必须始终使用其更大的父项ID路由子文档

>     PUT my_index/_doc/3?routing=1&refresh 
>     {
>       "text": "This is an answer",
>       "relation_field": {
>         "name": "answer", 
>         "parent": "1" 
>       }
>     }
>     
>     PUT my_index/_doc/4?routing=1&refresh
>     {
>       "text": "This is another answer",
>       "relation_field": {
>         "name": "answer",
>         "parent": "1"
>       }
>     }
> 
>  
> 
> PUT my\_index1/\_doc/5?routing=1&refresh  
> \{  
>   "text": "This is an answer",  
>   "relation\_field": \{  
>     "name": "answer",   
>     "parent": "1"   
>       
>   \},  
>   "info":"This is an answer"  
> \}
> 
> PUT my\_index1/\_doc/7?routing=1&refresh  
> \{  
>   "text": "This is another answer",  
>   "relation\_field": \{  
>     "name": "answer",  
>     "parent": "1"  
>   \},  
>     "info":"This is an answer nxnnn"  
> \}
> 
> //id 31子记录，包含自定义的obj对象
> 
> PUT my\_index1/\_doc/31?routing=1&refresh  
> \{  
>   "text": "This is another answer 11",  
>   "relation\_field": \{  
>     "name": "answer",  
>     "parent": "1"  
>   \},  
>     "info":"This is an answer nxnnn",  
>   "age":30,  
>   "date":"2019-10-25",  
>   "obj": \{  
>     "name": "answer",  
>     "age": "1",  
>     "check":"i am testing"  
>   \}  
> \}
> 
> //针对上面obj对象的查询  
> GET my\_index1/\_search  
> \{  
>   "query": \{  
>     "has\_child": \{  
>       "type": "answer",  
>       "query": \{  
>         "match": \{  
>           "obj.name":"answer"  
>         \}  
>       \},"inner\_hits":\{\}  
>     \}  
>   \}  
> \}

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2hfc245OTk5_size_16_color_FFFFFF_t_70][]

*  `join`每个索引仅允许一个字段映射。
 *  父文档和子文档必须在同一分片上建立索引。这意味着`routing`在[获取][Link 1]，[删除][Link 2]或[更新][Link 3] 子文档时需要提供相同的值 。
 *  一个元素可以有多个子级，但只能有一个父级。
 *  可以向现有`join`字段添加新关系。
 *  也可以将子元素添加到现有元素中，但前提是该元素已经是父元素。

Parent-join 查询与聚合

>     GET my_index/_search
>     {
>       "query": {
>         "parent_id": { 
>           "type": "answer",
>           "id": "1"
>         }
>       },
>       "aggs": {
>         "parents": {
>           "terms": {
>             "field": "relation_field#question", 
>             "size": 10
>           }
>         }
>       },
>       "script_fields": {
>         "parent": {
>           "script": {
>              "source": "doc['relation_field#question']" 
>           }
>         }
>       }
>     }
> 
>

查询包含子文档信息：

>   
> GET my\_index1/\_search  
> \{  
>   "query": \{  
>     "parent\_id": \{   
>       "type": "answer",  
>       "id": "1"  
>     \}  
>   \},  
>   "aggs": \{  
>     "parents": \{  
>       "terms": \{  
>         "field": "relation\_field\#question",   
>         "size": 10  
>       \}  
>     \}  
>   \}  
> \}

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2hfc245OTk5_size_16_color_FFFFFF_t_70 1][]

> @Test  
>     public void testParentSonSearch() throws IOException \{
> 
>         QueryBuilder termQuery = QueryBuilders.termQuery("answer", "1");  
>           
>         //查询父级数据--包含answer子的父  ，加上innerHit(new InnerHitBuilder())可以获取子对象的数据  
>         HasChildQueryBuilder jb = JoinQueryBuilders.hasChildQuery("answer",QueryBuilders.matchAllQuery(),                          
>                 ScoreMode.None).innerHit(new InnerHitBuilder());  
>         //HasChildQueryBuilder jb = JoinQueryBuilders.hasChildQuery("answer",QueryBuilders.matchAllQuery(),                          
>         //ScoreMode.None);  
>           
>         SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();  
>         searchSourceBuilder.query(jb);  
>         searchSourceBuilder.from(0);  
>         searchSourceBuilder.size(5);  
>         searchSourceBuilder.timeout(new TimeValue(60, TimeUnit.SECONDS));  
>         SearchRequest searchRequest = new SearchRequest();  
>         searchRequest.source(searchSourceBuilder);  
>         searchRequest.indices("my\_index1");
> 
>         SearchResponse searchResponse = client.search(searchRequest, RequestOptions.DEFAULT);
> 
>         SearchHits hits = searchResponse.getHits();
> 
>         for (SearchHit searchHit : hits) \{  
>             System.out.println("---testParentSonSearch SearchHit---" + searchHit.getSourceAsString());  
>             //这里可以获取子对象的数据  
>             searchHit.getInnerHits().get("answer");  
>             searchHit.getId();  
>         \}  
>           
>           
>         //查询子级数据-- 父为question同时id=1  
>         QueryBuilder q = QueryBuilders.idsQuery().addIds("1");  
>           
>         //HasParentQueryBuilder jb1 = JoinQueryBuilders.hasParentQuery("question",QueryBuilders.matchAllQuery(),false);   
>         HasParentQueryBuilder jb1 = JoinQueryBuilders.hasParentQuery("question",q,false);          
>         searchSourceBuilder = new SearchSourceBuilder();  
>         searchSourceBuilder.query(jb1);  
>         searchSourceBuilder.from(0);  
>         searchSourceBuilder.size(5);  
>         searchSourceBuilder.timeout(new TimeValue(60, TimeUnit.SECONDS));  
>         searchRequest = new SearchRequest();  
>         searchRequest.source(searchSourceBuilder);  
>         searchRequest.indices("my\_index1");
> 
>         searchResponse = client.search(searchRequest, RequestOptions.DEFAULT);
> 
>         hits = searchResponse.getHits();
> 
>         for (SearchHit searchHit : hits) \{  
>             System.out.println("---SonSearch SearchHit---" + searchHit.getSourceAsString());  
>         \}  
>           
>           
>     \}

更多代码请参考：[https://github.com/hsn999/Elasticsearch\_7\_springboot\_demo][https_github.com_hsn999_Elasticsearch_7_springboot_demo]

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2hfc245OTk5_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20191103222607572.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2hfc245OTk5,size_16,color_FFFFFF,t_70
[Link 1]: https://www.elastic.co/guide/en/elasticsearch/reference/current/docs-get.html
[Link 2]: https://www.elastic.co/guide/en/elasticsearch/reference/current/docs-delete.html
[Link 3]: https://www.elastic.co/guide/en/elasticsearch/reference/current/docs-update.html
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2hfc245OTk5_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/20191103161846722.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2hfc245OTk5,size_16,color_FFFFFF,t_70
[https_github.com_hsn999_Elasticsearch_7_springboot_demo]: https://github.com/hsn999/Elasticsearch_7_springboot_demo