09.span query 查询

1. 简介

这里介绍的是es对lucene的spanQuery的一些使用接口，主要有下面的几种查询

1. span_term query: 和term查询等效，一般用来在其他span query中作为组合查询
2. span_near query: 和intervals查询中的match 查询很像，约束多个查询term的order和gap
3. span_not query: 和boost query很像，但是包含的是 include和excludes两个条件（boost query 是positive和negative）
4. span_or query: 结合多个span query ,和bool 查询中的should很像，只要命中一个即可
5. span_first query: 要查询的term在对应的field中的位置距离起始位置不能超过N
6. span_within query: 这个挺特殊，是分为little和big两个span query, 而且little的query中的term必须在big的多个term之间（在source field中对应的text当中）
7. span_containing query: 这个感觉和span_within一样，就是查询顺序不一样，先查little再过滤big条件
8. field_masking_span query: 可以让span-near or span-or across different fields，具体使用没有看太明白
9. span_multi query: Wraps a term, range, prefix, wildcard, regexp, or fuzzy query.

2. 样例

原始数据

GET multi_test/_search

返回
{
  "hits" : {
    "total" : {
      "value" : 6,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_source" : {
          "shingle" : "If the calculations based on the specification determine that",
          "title" : "If the calculations based on the specification determine that",
          "age" : 80,
          "status" : true
        }
      },
      {
        "_source" : {
          "shingle" : "my favorite food hot water",
          "title" : "my favorite food hot water",
          "age" : 23,
          "status" : true
        }
      },
      {
        "_source" : {
          "shingle" : "that no optional clauses are needed, the usual rules about BooleanQueries ",
          "title" : "that no optional clauses are needed, the usual rules about BooleanQueries ",
          "age" : 39,
          "status" : false
        }
      },
      {
        "_source" : {
          "shingle" : "rules about BooleanQueries still apply at search ",
          "title" : "rules about BooleanQueries still apply at search ",
          "age" : 42,
          "status" : false
        }
      },
      {
        "_source" : {
          "shingle" : "If the calculations based",
          "title" : " suggester requires special mapping up  ",
          "age" : 38,
          "status" : false
        }
      },
      {
        "_source" : {
          "shingle" : "my favorite food is pisa",
          "title" : "the my favorite  that is pisa",
          "age" : 35,
          "status" : false
        }
      }
    ]
  }
}

1. span_term query

GET multi_test/_search
{
  "query": {
    "span_term": {
      "title": {
        "value": "that"
      }
    }
  }
}

2. span_near query

GET multi_test/_search
{
  "query": {
    "span_near": {
      "clauses": [
        {
          "span_term": {
            "title": {
              "value": "up"
            }
          }
        },
        {
          "span_term": {
            "title": {
              "value": "suggester"
            }
          }
        }
      ],
      "slop": 3,
      "in_order": false
    }
  }
}

3. span_not query

GET multi_test/_search
{
  "query": {
    "span_not": {
      "include": {
        "span_term": {
          "title": {
            "value": "favorite"
          }
        }
      },
      "exclude": {
        "span_near": {
          "clauses": [
            {
              "span_term": {
                "title": {
                  "value": "my"
                }
              }
            },
            {
              "span_term": {
                "title": {
                  "value": "favorite"
                }
              }
            }
          ],
          "slop": 0,
          "in_order": false
        }
      }
    }
  }
}

我在测试的时候，如果exclude当中使用span term的话那么会出现exclude失效的情况，具体原因不太清楚

4. span_or query

GET multi_test/_search
{
  "query": {
    "span_or": {
      "clauses": [
        {
          "span_term": {
            "title": {
              "value": "food"
            }
          }
        },
        {
          "span_term": {
            "title": {
              "value": "favorite"
            }
          }
        }
      ]
    }
  }
}

5. span_first query

GET multi_test/_search
{
  "query": {
    "span_first": {
      "match": {
        "span_term": {
          "title": "hot"
        }
      },
      "end": 4
    }
  }
}

返回

"_source" : {
          "shingle" : "my favorite food hot water",
          "title" : "my favorite food hot water",
          "age" : 23,
          "status" : true
        }

6. span_within query

GET multi_test/_search
{
  "query": {
    "span_within": {
      "little": {
        "span_term": {
          "title": {
            "value": "special"
          }
        }
      },
      "big": {
        "span_near": {
          "clauses": [
            {
              "span_term": {
                "title": {
                  "value": "suggester"
                }
              }
            },
            {
              "span_term": {
                "title": {
                  "value": "mapping"
                }
              }
            }
          ],
          "slop": 5,
          "in_order": false
        }
      }
    }
  }
}


返回
"_source" : {
          "shingle" : "If the calculations based",
          "title" : " suggester requires special mapping up  ",
          "age" : 38,
          "status" : false
        }

这里的litte中的term special在源文档中必须在big查询的两个term suggester, mapping中间才行，否则查不出来

7. span_containing query

GET multi_test/_search
{
  "query": {
    "span_containing": {
      "little": {
        "span_term": {
          "title": {
            "value": "special"
          }
        }
      },
      "big": {
        "span_near": {
          "clauses": [
            {
              "span_term": {
                "title": {
                  "value": "suggesterzh"
                }
              }
            },
            {
              "span_term": {
                "title": {
                  "value": "mapping"
                }
              }
            }
          ],
          "slop": 5,
          "in_order": false
        }
      }
    }
  }
}

要求同span within 查询

8. field_masking_span query

这个没有用明白。。。

GET multi_test02/_search
{
  "query": {
    "span_near": {
      "clauses": [
        {
          "span_term": {
            "title": "requires"
          }
        },
        {
          "field_masking_span": {
            "query": {
              "span_term": {
                "shingle": "based"
              }
            },
            "field": "title"
          }
        }
      ],
      "slop": 1,
      "in_order": false
    }
  }
}

9. span_multi query

GET multi_test/_search
{
  "query": {
   "span_multi": {
     "match": {
       "prefix": {
         "title": {
           "value": "sug"
         }
       }
     }
   }
  }
}

返回

        "_source" : {
          "shingle" : "If the calculations based",
          "title" : " suggester requires special mapping up  ",
          "age" : 38,
          "status" : false
        }