对嵌套对象的所有字段进行匹配的弹性搜索嵌套查询

Question

基本问题如下：有没有一种方便的方法可以为嵌套查询的所有字段指定多字段匹配？对于普通查询{匹配：{_all："查询字符串"}}有效。这在嵌套查询中不起作用，可能是因为嵌套对象没有_all？

下面是更详细的问题:

我有一个名为“Parent”的嵌套文档，如下所示:

{
    "children" : [
        { 
            "field_a": "value_a_1",
            "field_b" : "value_b_1",
            "field_c" : [ {
                "field_c_a" : "value_c_a_1",
                "field_c_b" : "value_c_b_1"
            } ]
        },
        { 
            "field_a": "value_a_2",
            "field_b" : "value_b_2",
            "field_c" : [ {
                "field_c_a" : "value_c_a_2",
                "field_c_b" : "value_c_b_2"
            } ]
        }
    ]
}

这是我用于制作儿童嵌套对象的映射：

"Parent" : {
    "properties" : {
        "children" : {
            "type" : "nested",
            "include_in_parent" : true
                }
    }
}

这是一个查询，我想使用所有子字段的匹配来选择几个术语查询，以及一个术语查询：

"query" : {
    "nested": { 
        "path" : "children",
        "query" : {
            "bool" : {
                "must" : [
                        {"multi_match" : {"query": "value_c_a_1", "fields" : ["children.*"]}},
                        {"term" : {children.field_a : "value_a_1" }}
                    ]
                }
            }
    } 
}

上述查询不起作用，因为我无法为嵌套对象选择多匹配查询中的所有字段。

"query" : {
    "nested": { 
        "path" : "children",
        "query" : {
            "bool" : {
                "must" : [
                        {"multi_match" : {"query": "value_c_a_1", "fields" : ["*_c_a"]}}
                    ]
                }
            }
    } 
}

上面的查询之所以有效，是因为模式匹配允许将*放在字符串之前，但出于某种原因（？）

有没有一种很好的速记方法来选择嵌套对象的所有字段？

如果能知道为什么预期的子项 *通配符不能按预期工作，那也会很好。

Answer 1

对我来说工作正常（注：Elasticsearch 1.7）。注意不同的名字（parent，somechildren）——我还没有用原来的名字测试过它，但它可能与它有关。

图式：

curl -XPUT localhost:9200/test -d '{
  "mappings": {
    "aparent": {
      "properties": {
        "somechildren": {
          "type": "nested",
          "include_in_parent": true
        }
      }
    }
  }
}'

文档:

curl -XPUT localhost:9200/test/aparent/1 -d '{
  "somechildren" : [
    {
      "field_a": "value_a_1",
      "field_b" : "value_b_1",
      "field_c" : [ {
        "field_c_a" : "value_c_a_1",
        "field_c_b" : "value_c_b_1"
      } ]
    },
    {
      "field_a": "value_a_2",
      "field_b" : "value_b_2",
      "field_c" : [ {
        "field_c_a" : "value_c_a_2",
        "field_c_b" : "value_c_b_2"
      } ]
    }
  ]
}'

查询：

GET test/_search
{
  "query": {
    "nested": {
      "path": "somechildren",
      "query": {
        "bool": {
          "must": [
            {
              "multi_match": {
                "query": "value_c_a_1",
                "fields": [
                  "somechildren.*"
                ]
              }
            },
            {
              "term": {
                "somechildren.field_a": {
                  "value": "value_a_1"
                }
              }
            }
          ]
        }
      }
    }
  }
}

结果：

  {
     "took": 2,
     "timed_out": false,
     "_shards": {
        "total": 5,
        "successful": 5,
        "failed": 0
     },
     "hits": {
        "total": 1,
        "max_score": 0.8603305,
        "hits": [
           {
              "_index": "test",
              "_type": "aparent",
              "_id": "1",
              "_score": 0.8603305,
              "_source": {
                 "somechildren": [
                    {
                       "field_a": "value_a_1",
                       "field_b": "value_b_1",
                       "field_c": [
                          {
                             "field_c_a": "value_c_a_1",
                             "field_c_b": "value_c_b_1"
                          }
                       ]
                    },
                    {
                       "field_a": "value_a_2",
                       "field_b": "value_b_2",
                       "field_c": [
                          {
                             "field_c_a": "value_c_a_2",
                             "field_c_b": "value_c_b_2"
                          }
                       ]
                    }
                 ]
              }
           }
        ]
     }
  }