将SQL转换为Elasticsearch DSL中的AND案例。

Question

我有一个SQL查询，像这样

sql_1 = "SELECT * FROM dd_s3data WHERE (yelp_address = '370 Barren Rd' OR yelp_businessname ILIKE '%Computer%') AND (yelp_state = 'CT' OR yelp_category ILIKE '%flooring%');"

我正试图将其转换为Elasticsearch查询。这是我尝试的查询。它给出的是 OR 结果 AND

es_query1 = {
    "query": {
        "constant_score": {
            "filter": {
                "bool": {
                    "should": [
                        {"match_phrase": {"yelp_address": "370 Barren Rd"}},
                        {"match": {"yelp_businessname": "Computer"}}
                    ],
                    "should": [
                        {"match": {"yelp_state": "CT"}},
                        {"match_phrase": {"yelp_category": "flooring"}}
                    ]
                }
            }
        }
    }
}

我有一个偶大的查询，我的第一个查询正确后，我必须转换。

sql_2 = "SELECT * FROM dd_s3data WHERE (yelp_address = '370 Barren Rd' OR yelp_businessname ILIKE '%Computer%') AND (yelp_state = 'CT' OR yelp_category ILIKE '%flooring%') AND yelp_noofreviews < 3.0 AND yelp_noofrating > 3.0;"

如何转换我的SQL查询，使我得到一个新的查询。AND 结果 OR?

Answer 1

对于 "OR"，你可以使用 "should"，最小_should_match:1。

对于 "和"，你可以使用 "必须"

如果你不想计算搜索结果的分数，那么就使用filter.constant_score -返回每个匹配的文档，其相关性分数等于boost参数值。

在你的例子中，单单使用filter就足够了，如果你想使用constant_score，那么就用constant_score包装filter查询，然后使用boost。

Query1.查询1：查询2：查询3：查询4：查询5

{
  "query": {
    "bool": {
      "filter": {
        "bool": {
          "must": [
            {
              "bool": {
                "should": [
                  {
                    "match_phrase": {
                      "yelp_address": "370 Barren Rd"
                    }
                  },
                  {
                    "match": {
                      "yelp_businessname": "Computer"
                    }
                  }
                ],
                "minimum_should_match":1
              }
            },
            {
              "bool": {
                "should": [
                  {
                    "match": {
                      "yelp_state": "CT"
                    }
                  },
                  {
                    "match_phrase": {
                      "yelp_category": "flooring"
                    }
                  }
                ],
                "minimum_should_match":1
              }
            }
          ]
        }
      }
    }
  }
}

查询2：

{
  "query": {
    "bool": {
      "filter": {
        "bool": {
          "must": [
            {
              "bool": {
                "should": [
                  {
                    "match_phrase": {
                      "yelp_address": "370 Barren Rd"
                    }
                  },
                  {
                    "match": {
                      "yelp_businessname": "Computer"
                    }
                  }
                ]

              }
            },
            {
              "bool": {
                "should": [
                  {
                    "match": {
                      "yelp_state": "CT"
                    }
                  },
                  {
                    "match_phrase": {
                      "yelp_category": "flooring"
                    }
                  }
                ],
                "minimum_should_match":1
              }
            },
            {
              "range": {
                "yelp_noofreviews": {
                  "lt": 3
                }
              }
            },
            {
              "range": {
                "yelp_noofrating": {
                  "gt": 3
                }
              }
            }
          ]
        }
      }
    }
  }
}