谷粒商城实战笔记-127-全文检索-ElasticSearch-整合-测试复杂检索

一，使用Elasticsearch的Java RESTHighLevel Client完成复杂的查询请求

前面es进阶学习中，我们学习过复杂的DSL查询。

POST bank/_search
{
  "query": {
    "match": {
      "address": {
        "query": "Mill"
      }
    }
  },
  "aggregations": {
    "ageAgg": {
      "terms": {
        "field": "age",
        "size": 10
      }
    },
    "ageAvg": {
      "avg": {
        "field": "age"
      }
    },
    "balanceAvg": {
      "avg": {
        "field": "balance"
      }
    }
  }
}

如何使用Java客户端执行复杂的查询呢？

使用Elasticsearch的Java REST High-Level Client执行一个复杂的带有聚合的搜索请求。

1. 创建检索请求 (SearchRequest)

• 创建 SearchRequest 对象:

  • SearchRequest searchRequest = new SearchRequest();

• 指定索引:

  • searchRequest.indices("bank");

2. 构造检索条件 (SearchSourceBuilder)

• 创建 SearchSourceBuilder 对象:

  • SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();

• 设置查询条件:

  • sourceBuilder.query(QueryBuilders.matchQuery("address", "Mill"));

• 添加聚合:

  • 按年龄分组的聚合:

    • TermsAggregationBuilder ageAgg = AggregationBuilders.terms("ageAgg").field("age").size(10);
    • sourceBuilder.aggregation(ageAgg);

  • 计算平均年龄:

    • AvgAggregationBuilder ageAvg = AggregationBuilders.avg("ageAvg").field("age");
    • sourceBuilder.aggregation(ageAvg);

  • 计算平均薪资:

    • AvgAggregationBuilder balanceAvg = AggregationBuilders.avg("balanceAvg").field("balance");
    • sourceBuilder.aggregation(balanceAvg);

• 打印检索条件:

  • System.out.println("检索条件：" + sourceBuilder);

• 将检索条件添加到 SearchRequest 中:

  • searchRequest.source(sourceBuilder);

3. 执行检索 (SearchResponse)

• 执行搜索请求:

  • SearchResponse searchResponse = client.search(searchRequest, RequestOptions.DEFAULT);

• 打印检索结果:

  • System.out.println("检索结果：" + searchResponse);

4. 处理解析结果

• 获取搜索命中的文档:

  • SearchHits hits = searchResponse.getHits();
  • SearchHit[] searchHits = hits.getHits();

• 遍历并处理每个文档:

  • for (SearchHit searchHit : searchHits) {
        String sourceAsString = searchHit.getSourceAsString();
        Account account = JSON.parseObject(sourceAsString, Account.class);
        System.out.println(account);
    }
    

5. 获取聚合信息

• 获取聚合结果:

  • Aggregations aggregations = searchResponse.getAggregations();

• 处理年龄分布的聚合:

  • Terms ageAgg1 = aggregations.get("ageAgg");
    for (Terms.Bucket bucket : ageAgg1.getBuckets()) {
        String keyAsString = bucket.getKeyAsString();
        System.out.println("年龄：" + keyAsString + " ==> " + bucket.getDocCount());
    }
    

• 处理平均年龄的聚合:

  • Avg ageAvg1 = aggregations.get("ageAvg");
    System.out.println("平均年龄：" + ageAvg1.getValue());
    

• 处理平均薪资的聚合:

  • Avg balanceAvg1 = aggregations.get("balanceAvg");
    System.out.println("平均薪资：" + balanceAvg1.getValue());
    

完整代码如下：

	/**
     * 复杂检索
     */
    @Test
    public void searchData() throws IOException {
        //1. 创建检索请求
        SearchRequest searchRequest = new SearchRequest();

        //1.1）指定索引
        searchRequest.indices("bank");
        //1.2）构造检索条件
        SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
        sourceBuilder.query(QueryBuilders.matchQuery("address", "Mill"));

        //1.2.1)按照年龄分布进行聚合
        TermsAggregationBuilder ageAgg = AggregationBuilders.terms("ageAgg").field("age").size(10);
        sourceBuilder.aggregation(ageAgg);

        //1.2.2)计算平均年龄
        AvgAggregationBuilder ageAvg = AggregationBuilders.avg("ageAvg").field("age");
        sourceBuilder.aggregation(ageAvg);
        //1.2.3)计算平均薪资
        AvgAggregationBuilder balanceAvg = AggregationBuilders.avg("balanceAvg").field("balance");
        sourceBuilder.aggregation(balanceAvg);

        System.out.println("检索条件：" + sourceBuilder);
        searchRequest.source(sourceBuilder);
        
        //2. 执行检索
        SearchResponse searchResponse = client.search(searchRequest, RequestOptions.DEFAULT);
        System.out.println("检索结果：" + searchResponse);

        //3. 将检索结果封装为Bean
        SearchHits hits = searchResponse.getHits();
        SearchHit[] searchHits = hits.getHits();
        for (SearchHit searchHit : searchHits) {
            String sourceAsString = searchHit.getSourceAsString();
            Account account = JSON.parseObject(sourceAsString, Account.class);
            System.out.println(account);

        }

        //4. 获取聚合信息
        Aggregations aggregations = searchResponse.getAggregations();

        Terms ageAgg1 = aggregations.get("ageAgg");

        for (Terms.Bucket bucket : ageAgg1.getBuckets()) {
            String keyAsString = bucket.getKeyAsString();
            System.out.println("年龄：" + keyAsString + " ==> " + bucket.getDocCount());
        }
        Avg ageAvg1 = aggregations.get("ageAvg");
        System.out.println("平均年龄：" + ageAvg1.getValue());

        Avg balanceAvg1 = aggregations.get("balanceAvg");
        System.out.println("平均薪资：" + balanceAvg1.getValue());
    }

二，AI时代的效率提升

相对于DSL，使用Java客户端来完成复杂的请求，代码是比较复杂不好理解的。

DSL相对清晰、容易理解。

所以，我们可以先根据需求，写好DSL，然后用大模型工具比如通义千问、Kimi、ChatGPT等将DSL转换为Java代码，这样我们就无需逐行编写复杂难懂的Java代码了，只需要在测试过程中进行微调即可。