elasticsearch - 如何在Elasticsearch中使用match_phrase查询使用正则表达式搜索子对象的值字段

Question

Elasticseach中的层次结构文档如下

"hierarchy":{
        "username":"Nothing on",
        "Location":"",
        "Owner":"unknown",
        "works": "IT"
        
    }

是否有任何方法可以用层次结构内的任何字段搜索值，层次结构内的字段是动态的，并尝试使用正则表达式模式匹配并出错。
我们无法指定 ，例如architecture.works，hierarchy.Owner 等。因为在层次结构内的 字段是动态的，还使用了转义字符串，但它也会产生错误。

{
        "from": 0,
        "query": {
            "multi_match": {
                "query": "IT",
                "type": "phrase_prefix",
                "fields": [
                    "number",
                    "name",
                    "address",
                    "hierarchy.*.",
                    "tags"
                ],
            }
        }
    }

Answer 1

If no fields are provided, the multi_match query defaults to theindex.query.default_field index settings, which in turn defaults to *.This extracts all fields in the mapping that are eligible to term queries and filters the metadata fields. All extracted fields are thencombined to build a query.

由于 hierarchy字段是动态的，因此您无法指定要查询的字段，因此查询查询的一种方法是在 field参数中不添加任何字段。
索引数据1:

{
  "hierarchy": {
    "username": "Nothing on",
    "Location": "",
    "Owner": "unknown",
    "works": "IT",
    "education":"btech"
  }
}

索引数据2:

{
  "hierarchy": {
    "username": "Nothing on",
    "Location": "",
    "Owner": "unknown",
    "works": "IT"
  }
}

搜索查询:

{
  "query": {
    "multi_match": {
      "query": "btech",
      "type": "phrase_prefix"
    }
  }
}

搜索结果:

 "hits": [
      {
        "_index": "64467424",
        "_type": "_doc",
        "_id": "2",
        "_score": 0.2876821,
        "_source": {
          "hierarchy": {
            "username": "Nothing on",
            "Location": "",
            "Owner": "unknown",
            "works": "IT",
            "education": "btech"
          }
        }
      }
    ]

编辑1:

You can use the query_string query to create a complex search thatincludes wildcard characters, searches across multiple fields, andmore. While versatile, the query is strict and returns an error if thequery string includes any invalid syntax.

{
  "query": {
    "query_string": {
      "fields": [
        "number",
        "name",
        "address",
        "hierarchy*",
        "tags"
      ],
      "query": "btech"
    }
  }
}