Elasticsearch集成高级和低级API

Question

我构建了Elasticsearch查询（DSL），如下所示。我想从Java应用程序中集成它。当我阅读时，高水平和低水平的Rest选项由Elastic提供。是否可以使用高级API构建复杂的查询？或者我将使用低级API，通过程序构建DSL？

GET myindex/_search
{
    "query": {
        "bool": {
        "must": [
          {
            "nested": {
              "path": "blog_types",
              "query": {
                "match": {
                  "blog_types.blog_type": 45649
                }
              }
            }
          },
          {
            "nested": {
              "path": "blog_types",
              "query": {
                "match": {
                  "blog_types.blog_type": 45647
                }
              }
            }
          }
        ],
        "filter" : {
            "geo_bounding_box" : {
                "request_location.location" : {
                    "top_left" : {
                        "lat" : 50.925996,
                        "lon" : 4.584309
                    },
                    "bottom_right" : {
                        "lat" : 50.845996,
                        "lon" : 4.494309
                    }
                }
            }
        }
        }
    }
}

Answer 1

对于未来的用户：假设我们想要使用ES Client来复制此查询：

  {
  "query": {
    "bool": {
      "must": [
        {
          "bool": {
            "should": [
              {
                "match": {
                  "table_name": {
                    "query": "google",
                    "operator": "and"
                  }
                }
              },
              {
                "match": {
                  "table_name": {
                    "query": "yahoo",
                    "operator": "and"
                  }
                }
              }
            ]
          }
        },
        {
          "bool": {
            "must_not": [
              {
                "match": {
                  "vendor_type": {
                    "query": "9",
                    "operator": "and"
                  }
                }
              }
            ]
          }
        }
      ]
    }
  }
}

使用此代码可获得以下信息：

    //Sub Query of MustBool
    BoolQueryBuilder mustNotBoolQueryBuilder = QueryBuilders.boolQuery();
    mustNotBoolQueryBuilder.mustNot(QueryBuilders.matchQuery("vendor_type",  String.valueOf(DBVendor.Collibra.getValue())).operator(Operator.AND));

    //Sub Query of MustBool
    BoolQueryBuilder shouldBoolQueryBuilder = QueryBuilders.boolQuery();
    assetNames.forEach(assetName -> {
        shouldBoolQueryBuilder.should(QueryBuilders.matchQuery("table_name",assetName).operator(Operator.AND));
    });

    //Root Bool
    BoolQueryBuilder mustBoolQueryBuilder = QueryBuilders.boolQuery();
    mustBoolQueryBuilder.must(shouldBoolQueryBuilder);
    mustBoolQueryBuilder.must(mustNotBoolQueryBuilder);

Answer 2

好的，下面是：

QueryBuilder blogType1 = QueryBuilders.nestedQuery("blog_types", 
    QueryBuilders.matchQuery("blog_types.blog_type", 45649), ScoreMode.Avg);

QueryBuilder blogType2 = QueryBuilders.nestedQuery("blog_types", 
    QueryBuilders.matchQuery("blog_types.blog_type", 45647), ScoreMode.Avg); 

QueryBuilder geo = QueryBuilders.geoBoundingBoxQueryBuilder("request_location.location")
    .setCorners(50.925996, 4.584309, 50.845996, 4.494309);

QueryBuilder mainQuery = QueryBuilders.boolQuery()
    .filter(blogType1)
    .filter(blogType2)
    .filter(geo);

如果您需要或需要条件，可以使用should：

QueryBuilder blogType1 = QueryBuilders.nestedQuery("blog_types", 
    QueryBuilders.matchQuery("blog_types.blog_type", 45649), ScoreMode.Avg);

QueryBuilder blogType2 = QueryBuilders.nestedQuery("blog_types", 
    QueryBuilders.matchQuery("blog_types.blog_type", 45647), ScoreMode.Avg); 

QueryBuilder geo = QueryBuilders.geoBoundingBoxQueryBuilder("request_location.location")
    .setCorners(50.925996, 4.584309, 50.845996, 4.494309);

QueryBuilder mainQuery = QueryBuilders.boolQuery()
    .should(blogType1)
    .should(blogType2)
    .filter(geo)
    .minimumShouldMatch(1);