elasticsearch笔记(7) 复合查询_boolean查询 boolsting查询 filter查询

6 复合查询

6.1 bool查询

将多个查询条件以一定的逻辑组合在一起

• must：表示and的意思，所有的条件都符合才能找到
• must_not：把满足条件的都去掉的结果
• should：表示or的意思

# 查询省份是上海或者河南
# 运营商不是联通
# smsContent中包含中国和移动
# bool查询

POST /sms-logs-index/sms-logs-type/_search
{
  "query": {
    "bool": {
      "should": [
        {
          "term": {
            "province": {
              "value": "河南"
            }
          }
        },  
        {
          "term": {
            "province": {
              "value": "北京"
             }
            }
         }
      ],
      "must_not": [
        {
          "term": {
            "operatorId": {
              "value": "2"
            }
          }
        }
      ],
      "must": [
        {
          "match": {
            "smsContent": "中国"
          }
        },
        {
          "match": {
            "smsContent": "移动"
          }
        }
      ]
    }
  }
}

java中的聚合查询

@Test
    public void bolleanTest() throws IOException {
        //        1. 创建查询对象
        String index = "sms-logs-index";
        String type = "sms-logs-type";
        SearchRequest searchRequest = new SearchRequest(index);//指定索引
        searchRequest.types(type);//指定类型
//    2. 封装查询条件
        SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
10         BoolQueryBuilder boolQueryBuilder = QueryBuilders.boolQuery();
        boolQueryBuilder.should(QueryBuilders.termQuery("province","河南"));
        boolQueryBuilder.should(QueryBuilders.termQuery("province","北京"));
        boolQueryBuilder.mustNot(QueryBuilders.termQuery("operatorId","2"));
        boolQueryBuilder.must(QueryBuilders.matchQuery("smsContent","中国"));
        boolQueryBuilder.must(QueryBuilders.matchQuery("smsContent","移动"));
        searchSourceBuilder.query(boolQueryBuilder);
        searchRequest.source(searchSourceBuilder);



        //        3.执行查询
        // client执行
        HttpHost httpHost = new HttpHost("192.168.43.30", 9200);
        RestClientBuilder restClientBuilder = RestClient.builder(httpHost);
        RestHighLevelClient restHighLevelClient = new RestHighLevelClient(restClientBuilder);
        SearchResponse search = restHighLevelClient.search(searchRequest, RequestOptions.DEFAULT);

//        4.获取数据(source中的数据)
        SearchHit[] hits = search.getHits().getHits();
        for(SearchHit searchHit : hits){
            System.out.println(searchHit);
        }
    }

6.2 boolsting

分数查询，比如math查询,结果有一个匹配度分数.  boolsting查询可以针对内容，让其分数大，或者小，达到排前，排后的效果

• positive： 指定查询条件,  只有匹配到positive的内容，才会放到结果集，也就是放查询条件的地方
• negative：如果匹配到的positive和negative，就会降低文档的分数
• negative_boost：对匹配到的positive和negative的内容, 指定降低分数的系数. 这个系数必须小于1.0，比如：10分 这个系数为0.5就会变为5分

关于分数的计算：

• 关键字在文档出现的频次越高，分数越高
• 文档的内容越短，分数越高
• 搜索时候，指定的关键字会被分词，分词内容匹配分词库，匹配的个数越多，分数就越高

# boosting查询
POST /sms-logs-index/sms-logs-type/_search
{
  "query": {
    "boosting": {
      "positive": { #这三个postitive. negative 和 negative_boost是在一级的.
        "match": {
          "smsContent": {
             "query": "电话玩家活物",
               "operator": "or"
          }
         }
      },
      "negative": { //匹配到这个的就按negative_boost在分数中减少
        "match": {
          "smsContent": "订单"
        }
      },
      "negative_boost": 0.1
    }
  }
}

java中用boosting查询

    @Test
    public void boostingTest() throws IOException {
        //        1. 创建查询对象
        String index = "sms-logs-index";
        String type = "sms-logs-type";
        SearchRequest searchRequest = new SearchRequest(index);//指定索引
        searchRequest.types(type);//指定类型
//    2. 封装查询条件
        SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();

        BoostingQueryBuilder boostingQueryBuilder = QueryBuilders.boostingQuery(
                QueryBuilders.matchQuery("smsContent", "电话").operator(Operator.OR),//相当于positivequery
                QueryBuilders.matchQuery("smsContent", "订单")
        ).negativeBoost(0.1f);

        searchSourceBuilder.query(boostingQueryBuilder);
        searchRequest.source(searchSourceBuilder);



        //        3.执行查询
        // client执行
        HttpHost httpHost = new HttpHost("192.168.43.30", 9200);
        RestClientBuilder restClientBuilder = RestClient.builder(httpHost);
        RestHighLevelClient restHighLevelClient = new RestHighLevelClient(restClientBuilder);
        SearchResponse search = restHighLevelClient.search(searchRequest, RequestOptions.DEFAULT);

//        4.获取数据(source中的数据)
        SearchHit[] hits = search.getHits().getHits();
        for(SearchHit searchHit : hits){
            System.out.println(searchHit);
        }
    }

二, filter查询:  如果结果不需要匹配度

     与query查询相比, 会计算分数, 并对结果排序

     filter不计算分数, 结果也不排序

 filter查询在java中应用

    @Test
    public void filterTest() throws IOException {
        //        1. 创建查询对象
        String index = "sms-logs-index";
        String type = "sms-logs-type";
        SearchRequest searchRequest = new SearchRequest(index);//指定索引
        searchRequest.types(type);//指定类型
//    2. 封装查询条件
        SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
        BoolQueryBuilder boolQueryBuilder = QueryBuilders.boolQuery();//先创建Boolean查询对象
11         boolQueryBuilder.filter(QueryBuilders.termQuery("corpName", "网易"));


        searchSourceBuilder.query(boolQueryBuilder);
        searchRequest.source(searchSourceBuilder);



        //        3.执行查询
        // client执行
        HttpHost httpHost = new HttpHost("192.168.43.30", 9200);
        RestClientBuilder restClientBuilder = RestClient.builder(httpHost);
        RestHighLevelClient restHighLevelClient = new RestHighLevelClient(restClientBuilder);
        SearchResponse search = restHighLevelClient.search(searchRequest, RequestOptions.DEFAULT);

//        4.获取数据(source中的数据)
        SearchHit[] hits = search.getHits().getHits();
        for(SearchHit searchHit : hits){
            System.out.println(searchHit);
        }
    }

三. 高亮查询: 将查询到的内容自动加到高亮标签中.

将用户输入的内容，以高亮的样式展示出来，查询的结果会附带在hits下面以单独的形式返回，不会影响查询的结果

ES提供了一个hightlight的属性，和query同级别，其属性如下：

• fragment_size：指定要展示多少内容，可以看到百度的内容后面有...还有很长，默认100个
• pre_tags：指定前缀标签 比如：就是红色
• post_tags：指定后缀标签：
• fields：指定哪几个field以高亮形式返回

 java中用highlight查询

    @Test
    public void highlightQuery() throws IOException {
        //        1. 创建查询对象
        String index = "sms-logs-index";
        String type = "sms-logs-type";
        SearchRequest searchRequest = new SearchRequest(index);//指定索引
        searchRequest.types(type);//指定类型
//    2. 封装查询条件
        SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
        searchSourceBuilder.query(QueryBuilders.matchQuery("smsContent", "网易"));
        HighlightBuilder highlightBuilder = new HighlightBuilder();
        highlightBuilder.field("smsContent") .preTags("<font color='red'>").postTags("</font>").fragmentSize(20);
        searchSourceBuilder.highlighter(highlightBuilder);
        searchRequest.source(searchSourceBuilder);

        //        3.执行查询
        // client执行
        HttpHost httpHost = new HttpHost("192.168.43.30", 9200);
        RestClientBuilder restClientBuilder = RestClient.builder(httpHost);
        RestHighLevelClient restHighLevelClient = new RestHighLevelClient(restClientBuilder);
        SearchResponse search = restHighLevelClient.search(searchRequest, RequestOptions.DEFAULT);

//        4.获取数据(source中的数据)
        SearchHit[] hits = search.getHits().getHits();
        for(SearchHit searchHit : hits){
            System.out.println(searchHit);
        }
    }