如何在 spring 引导中编写以下弹性搜索通配符查询？

Question

Database: Elasticsearch

Field Name: "tag"

Datatype: String

Question: How to perform a search in the tag field with a wildcard character?

I tried the following (in Kibana):

User Request: { "tag": [ "Attendance", "Employee" ] }

POST test/_search
{
    "query":{
        "bool" : {
          "should" : [
            {"wildcard":{"tag.keyword": "*Attendance*" }},
            {"wildcard":{"tag.keyword": "*Employee*" }}
          ]
        }
    }
}

This worked successfully but I don't have an idea how to perform the same thing in spring boot.

I worked with following:

.should(wildcardQuery("tag.keyword", "Attendance"))

But this is just for a single field, I have a requirement for a dynamic field where user input should be different in size and value.

Can someone please guide me?

Answer 1

You could use something like this:

Rest Highlevel client Query:

SearchRequest searchRequest = new SearchRequest(<your-index-name>);
SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
 QueryBuilder query =  QueryBuilders.boolQuery().should(new WildcardQueryBuilder("tag.keyword", "*Attendance*"))
                    .should(new WildcardQueryBuilder("tag.keyword", "*Employee*"));
searchSourceBuilder.query(query);
searchRequest.source(searchSourceBuilder);

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

Here client is the RestHighLevelClient bean that I have autowired in my class as follows:

@Autowired
private RestHighLevelClient client;

and this bean I have defined in my config class as:

@Bean(destroyMethod = "close")
public RestHighLevelClient client() {

        RestHighLevelClient client = new RestHighLevelClient(
                RestClient.builder(new HttpHost("localhost", 9200, "http")));

        return client;

}

This would construct query in this form:

Query

    {
  "bool" : {
    "should" : [
      {
        "wildcard" : {
          "tag.keyword" : {
            "wildcard" : "*Attendance*",
            "boost" : 1.0
          }
        }
      },
      {
        "wildcard" : {
          "tag.keyword" : {
            "wildcard" : "*Employee*",
            "boost" : 1.0
          }
        }
      }
    ],
    "adjust_pure_negative" : true,
    "boost" : 1.0
  }
}

I have tested by creating index with same mapping as mentioned by you:

Documents indexed:

  "_source": {
            "tag": "Attendance1"
    }


    "_source": {
            "tag": "1Employee"
    }



      "_source": { "tag": "*Employee*" }


    { "tag" : [ "Attendance", "Employee" ] }

And when I search using above rest query, I got the following response:

Response

"hits": [
        {
            "_index": "test",
            "_type": "_doc",
            "_id": "4",
            "_score": 2.0,
            "_source": {
                "tag": [
                    "Attendance",
                    "Employee"
                ]
            }
        },
        {
            "_index": "test",
            "_type": "_doc",
            "_id": "1",
            "_score": 1.0,
            "_source": {
                "tag": "Attendance1"
            }
        },
        {
            "_index": "test",
            "_type": "_doc",
            "_id": "2",
            "_score": 1.0,
            "_source": {
                "tag": "*Employee*"
            }
        },
        {
            "_index": "test",
            "_type": "_doc",
            "_id": "3",
            "_score": 1.0,
            "_source": {
                "tag": "1Employee"
            }
        }
    ]

Not sure what is meant by this part of your question:

I have requirement for dynamic field where user input should be different in size and value .

But, I am assuming it means from user you may get different tag list and based on that you want to create query.

You can do something like this to achieve the same:

        // { "tag" : [ "Attendance", "Employee" ] } you can read this tag list though converter and pass it into list kind of data structure. Currently, I have created dummy list `tags` which has Attendance and Employee as items.
            List<String> tags = new ArrayList<>();
            tags.add("Attendance");
            tags.add("Employee");

            SearchRequest searchRequest = new SearchRequest(<your-index-name>);
            SearchSourceBuilder searchSourceBuilder = new SearchSourceBuilder();
            List<WildcardQueryBuilder> wildcards = new ArrayList<WildcardQueryBuilder>();

            for(String tag :  tags) {
                WildcardQueryBuilder wildcard =  new WildcardQueryBuilder("tag.keyword", "*" + tag + "*");
                wildcards.add(wildcard);

            }
            BoolQueryBuilder boolQuery = new BoolQueryBuilder();
            for(WildcardQueryBuilder wildcard : wildcards) {
                boolQuery.should(wildcard);
            }

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

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