ElasticSearch：constant_score 查询 vs function_score 查询

Question

我最近将我的 ElasticSearch 版本从 5.3 版升级到了 5.6 版

 "query" : {
    "constant_score" : {
      "query" : {
        "bool" : {
          "must" : {
            "terms" : {
              "customerId" : [ "ASERFE", "7004567457" ]
            }
          },
          "must_not" : {
            "terms" : {
              "useCase" : [ "PAY", "COLLECT" ]
            }
          }
        },
        "bool" : {
          "must" : {
            "match" : {
               "cardProductGroupName" : {
                  "query" : "Pre-fill Test birthday Present",
                  "type" : "phrase"
                }
             }
          }
        }
      }
    }
  }

执行上面提到的查询给了我以下错误-

{"root_cause":[{"type":"parsing_exception","re​​ason":"[constant_score] 查询不支持 [query]","line":1,"col":37}]," type":"parsing_exception","re​​ason":"[constant_score] 查询不支持 [query]","line":1,"col":37}

所以，我搜索了解决方案并找到了这个 function_score 查询。在执行下面提到的查询时，我得到的结果与使用 constant_score 得到的结果相同。

"query" : {
    "function_score" : {
      "query" : {
        "bool" : {
          "must" : {
            "terms" : {
              "customerId" : [ "ASERFE", "7004567457" ]
            }
          },
          "must_not" : {
            "terms" : {
              "useCase" : [ "PAY", "COLLECT" ]
            }
          }
        },
         "bool" : {
          "must" : {
            "match" : {
               "groupName" : {
                  "query" : "Pre-fill Test birthday Present",
                  "type" : "phrase"
                }
             }
          }
        }
      },
      "functions" : [ {
        "script_score" : {
          "script" : "1"
        }
      } ],
      "boost_mode" : "replace"
    }
  }

所以我的问题是，这是否意味着带有 script 的 function_score : "1" 会给出与 constant_function 相同的结果？

Answer 1

它确实会给出相同的结果，但如果它仍然为每个匹配的文档运行“脚本”，性能可能会更差。

另一方面，constant_score still exists in 5.6 虽然您必须使用 filter+boost 而不是 query。