ElasticSearch返回数据精简：查询语句及结果技术咨询

Hey there! Looking at your current query and the response you're getting, there are a few effective ways to trim down the returned data even more. Let's go through each option based on what you need:

1. Use filter_path to strip out unwanted top-level metadata #

ElasticSearch has a handy URL parameter called filter_path that lets you explicitly define which parts of the response you want to keep. If you only care about the actual EAN values from the matching documents, you can append this parameter to your request to eliminate all extra fields like took, _shards, max_score, etc.:

GET sales/_search?filter_path=hits.hits._source.ean
{
  "query": {
    "terms": {
      "ean": ["8719092410766", "8719092444716"]
    }
  },
  "_source": ["ean"],
  "size": 10000
}

This will return only the _source.ean entries for each hit, cutting out all the unnecessary metadata you don't need.

2. Replace _source with fields for cleaner value output #

If you want to get rid of the _source wrapper entirely, switch to using the fields parameter instead, and disable _source to save extra space. This will return the EAN values directly under a fields key:

GET sales/_search?filter_path=hits.hits.fields.ean
{
  "query": {
    "terms": {
      "ean": ["8719092410766", "8719092444716"]
    }
  },
  "fields": ["ean"],
  "_source": false,
  "size": 10000
}

The response will look super streamlined, like this:

{
  "hits": {
    "hits": [
      {
        "fields": {
          "ean": ["8719092444716"]
        }
      },
      // Additional matching entries...
    ]
  }
}

3. Use a terms aggregation for a flat list of unique EANs #

If your end goal is to get a unique list of matching EANs (instead of individual document hits), using an aggregation is even more efficient. This way, you don't return any document-level data at all—just the unique values and their counts:

GET sales/_search
{
  "query": {
    "terms": {
      "ean": ["8719092410766", "8719092444716"]
    }
  },
  "size": 0, // Disable hit results entirely
  "aggs": {
    "unique_eans": {
      "terms": {
        "field": "ean",
        "size": 10000
      }
    }
  }
}

The response will be a clean bucket of unique EANs:

{
  "aggregations": {
    "unique_eans": {
      "buckets": [
        {
          "key": "8719092444716",
          "doc_count": 2
        },
        {
          "key": "8719092410766",
          "doc_count": 1
        }
      ]
    }
  }
}

Choose the method that best fits your use case! If you need individual document EANs, the first two options work great. If you just need unique values with counts, the aggregation is the most efficient approach.

内容的提问来源于stack exchange，提问作者physicalattraction