如何在多分析器字段上实现模糊的以字段为中心（cross_fields）查询？

Got it, let's walk through how to query your Elasticsearch index properly, given the mapping and sample data you shared. Each field has its own analyzer or type, so we'll handle them one by one:

Since these are text fields with different analyzers, Elasticsearch will automatically use the field's configured analyzer when processing your query terms—this ensures consistency with how the data was indexed.

For the author field (standard analyzer) #

The standard analyzer splits text into words and lowercases them. If you want to find any document where the author includes "John" or "Smith", use a basic match query:

{
  "query": {
    "match": {
      "author": "John Smith"
    }
  }
}

If you need an exact phrase match (only documents where "John Smith" appears as a continuous phrase), use match_phrase:

{
  "query": {
    "match_phrase": {
      "author": "John Smith"
    }
  }
}

For title and content (english analyzer) #

The english analyzer does extra processing like stemming (reducing words to their root form) and stopword removal. For example, searching for "examples" would match documents with "example" in the field.

To find documents with "hello world" in the title:

{
  "query": {
    "match": {
      "title": "hello world"
    }
  }
}

This will match your sample data, since the english analyzer lowercases the query terms and matches against the indexed "hello world".

To search the content field for articles mentioning "example article":

{
  "query": {
    "match": {
      "content": "example article"
    }
  }
}

Keyword fields are indexed as exact values—no analysis is done. So you need to use exact match queries here.

Match a single tag #

{
  "query": {
    "term": {
      "tags": "programming"
    }
  }
}

Match multiple tags (all must be present) #

{
  "query": {
    "terms": {
      "tags": ["programming", "life"]
    }
  }
}

Match any of multiple tags #

Use a bool query with should clauses to get documents that have at least one of the tags:

{
  "query": {
    "bool": {
      "should": [
        {"term": {"tags": "programming"}},
        {"term": {"tags": "life"}}
      ],
      "minimum_should_match": 1
    }
  }
}

You can use a bool query to mix different field queries. For example, find articles where the title contains "hello" AND the tags include "programming":

{
  "query": {
    "bool": {
      "must": [
        {"match": {"title": "hello"}},
        {"term": {"tags": "programming"}}
      ]
    }
  }
}

If you ever need to bypass the field's default analyzer (e.g., use standard instead of english for the title), you can specify the analyzer in the query:

{
  "query": {
    "match": {
      "title": {
        "query": "Hello World",
        "analyzer": "standard"
      }
    }
  }
}

内容的提问来源于stack exchange，提问作者Piotr