Elasticsearch查询字符串多字段匹配实现方法咨询

Hey there! As someone new to Elasticsearch, let's walk through exactly how to set up multi-field matching using query strings for your dataset. Your documents have fields like person_name, name, and the nested office.name—so we'll cover how to target all of these in your queries.

Basic Multi-Field Match (Search Across Multiple Fields) #

If you want to search for a term across all your target fields at once, use the fields parameter in the query_string query. This tells Elasticsearch to check each specified field for the search term.

For example, to find all documents where any of person_name, name, or office.name contains "Abraham":

{
  "query": {
    "query_string": {
      "query": "Abraham",
      "fields": ["person_name", "name", "office.name"]
    }
  }
}

This will match your first, second, fourth, and fifth documents since they all have "Abraham" in one of those fields.

Target Specific Fields with Custom Conditions #

If you need more control—like matching different terms in different fields, or using logical operators—you can explicitly reference fields directly in your query string.

For instance, to find documents where person_name is "John Bradshaw" OR office.name is "Abraham deVilliers":

{
  "query": {
    "query_string": {
      "query": "person_name:\"John Bradshaw\" OR office.name:\"Abraham deVilliers\""
    }
  }
}

This will match your second and third documents.

Partial/Wildcard Matches #

If you need to match parts of a term (like finding any field containing "deVilliers"), you can use wildcards. Just note that leading wildcards (like *deVilliers) can impact performance, so use them sparingly.

Example:

{
  "query": {
    "query_string": {
      "query": "*deVilliers*",
      "fields": ["person_name", "name", "office.name"],
      "allow_leading_wildcard": true
    }
  }
}

Exact Matches & Default Operators #

To enforce exact phrase matches or require multiple terms to all be present, adjust the default_operator or use quotes around phrases.

For example, to find documents where name is exactly "Abraham" (no partial matches):

{
  "query": {
    "query_string": {
      "query": "name:\"Abraham\"",
      "default_operator": "AND"
    }
  }
}

Setting default_operator: "AND" ensures that if you search for multiple terms (like "Abraham deVilliers"), both terms must appear in the same field to match.

Key Notes for Nested Fields #

Since office is an object field, we access its nested name field using the dot notation (office.name)—this works seamlessly in query strings, so you don't need any special setup for nested objects here.

内容的提问来源于stack exchange，提问作者kgangadhar