IMPORTANT: No additional bug fixes or documentation updates will be released for this version. For the latest information, see the current release documentation.

› › ›

Dynamic field mapping

edit

Dynamic field mapping

edit

When Elasticsearch detects a new field in a document, it dynamically adds the field to the type mapping by default. The dynamic parameter controls this behavior.

You can explicitly instruct Elasticsearch to dynamically create fields based on incoming documents by setting the dynamic parameter to true or runtime. When dynamic field mapping is enabled, Elasticsearch uses the rules in the following table to determine how to map data types for each field.

The field data types in the following table are the only field data types that Elasticsearch detects dynamically. You must explicitly map all other data types.

	Elasticsearch data type
JSON data type	`"dynamic":"true"`	`"dynamic":"runtime"`
`null`	No field added	No field added
`true` or `false`	`boolean`	`boolean`
`double`	`float`	`double`
`long`	`long`	`long`
`object`	`object`	No field added
`array`	Depends on the first non-`null` value in the array	Depends on the first non-`null` value in the array
`string` that passes date detection	`date`	`date`
`string` that passes numeric detection	`float` or `long`	`double` or `long`
`string` that doesn’t pass `date` detection or `numeric` detection	`text` with a `.keyword` sub-field	`keyword`

You can disable dynamic mapping, both at the document and at the object level. Setting the dynamic parameter to false ignores new fields, and strict rejects the document if Elasticsearch encounters an unknown field.

Use the update mapping API to update the dynamic setting on existing fields.

You can customize dynamic field mapping rules for date detection and numeric detection. To define custom mappings rules that you can apply to additional dynamic fields, use dynamic_templates.

Date detection

edit

If date_detection is enabled (default), then new string fields are checked to see whether their contents match any of the date patterns specified in dynamic_date_formats. If a match is found, a new date field is added with the corresponding format.

The default value for dynamic_date_formats is:

[ "strict_date_optional_time","yyyy/MM/dd HH:mm:ss Z||yyyy/MM/dd Z"]

For example:

response = client.index(
  index: 'my-index-000001',
  id: 1,
  body: {
    create_date: '2015/09/02'
  }
)
puts response

response = client.indices.get_mapping(
  index: 'my-index-000001'
)
puts response

PUT my-index-000001/_doc/1
{
  "create_date": "2015/09/02"
}

GET my-index-000001/_mapping

The create_date field has been added as a date field with the format:
"yyyy/MM/dd HH:mm:ss Z||yyyy/MM/dd Z".

Disabling date detection

edit

Dynamic date detection can be disabled by setting date_detection to false:

response = client.indices.create(
  index: 'my-index-000001',
  body: {
    mappings: {
      date_detection: false
    }
  }
)
puts response

response = client.index(
  index: 'my-index-000001',
  id: 1,
  body: {
    create_date: '2015/09/02'
  }
)
puts response

PUT my-index-000001
{
  "mappings": {
    "date_detection": false
  }
}

PUT my-index-000001/_doc/1 
{
  "create_date": "2015/09/02"
}

The create_date field has been added as a text field.

Customizing detected date formats

edit

Alternatively, the dynamic_date_formats can be customized to support your own date formats:

response = client.indices.create(
  index: 'my-index-000001',
  body: {
    mappings: {
      dynamic_date_formats: [
        'MM/dd/yyyy'
      ]
    }
  }
)
puts response

response = client.index(
  index: 'my-index-000001',
  id: 1,
  body: {
    create_date: '09/25/2015'
  }
)
puts response

PUT my-index-000001
{
  "mappings": {
    "dynamic_date_formats": ["MM/dd/yyyy"]
  }
}

PUT my-index-000001/_doc/1
{
  "create_date": "09/25/2015"
}

There is a difference between configuring an array of date patterns and configuring multiple patterns in a single string separated by ||. When you configure an array of date patterns, the pattern that matches the date in the first document with an unmapped date field will determine the mapping of that field:

response = client.indices.create(
  index: 'my-index-000001',
  body: {
    mappings: {
      dynamic_date_formats: [
        'yyyy/MM',
        'MM/dd/yyyy'
      ]
    }
  }
)
puts response

response = client.index(
  index: 'my-index-000001',
  id: 1,
  body: {
    create_date: '09/25/2015'
  }
)
puts response

PUT my-index-000001
{
  "mappings": {
    "dynamic_date_formats": [ "yyyy/MM", "MM/dd/yyyy"]
  }
}

PUT my-index-000001/_doc/1
{
  "create_date": "09/25/2015"
}

The resulting mapping will be:

{
  "my-index-000001": {
    "mappings": {
      "dynamic_date_formats": [
        "yyyy/MM",
        "MM/dd/yyyy"
      ],
      "properties": {
        "create_date": {
          "type": "date",
          "format": "MM/dd/yyyy"
        }
      }
    }
  }
}

Configuring multiple patterns in a single string separated by || results in a mapping that supports any of the date formats. This enables you to index documents that use different formats:

response = client.indices.create(
  index: 'my-index-000001',
  body: {
    mappings: {
      dynamic_date_formats: [
        'yyyy/MM||MM/dd/yyyy'
      ]
    }
  }
)
puts response

response = client.index(
  index: 'my-index-000001',
  id: 1,
  body: {
    create_date: '09/25/2015'
  }
)
puts response

PUT my-index-000001
{
  "mappings": {
    "dynamic_date_formats": [ "yyyy/MM||MM/dd/yyyy"]
  }
}

PUT my-index-000001/_doc/1
{
  "create_date": "09/25/2015"
}

The resulting mapping will be:

{
  "my-index-000001": {
    "mappings": {
      "dynamic_date_formats": [
        "yyyy/MM||MM/dd/yyyy"
      ],
      "properties": {
        "create_date": {
          "type": "date",
          "format": "yyyy/MM||MM/dd/yyyy"
        }
      }
    }
  }
}

Epoch formats (epoch_millis and epoch_second) are not supported as dynamic date formats.

Numeric detection

edit

While JSON has support for native floating point and integer data types, some applications or languages may sometimes render numbers as strings. Usually the correct solution is to map these fields explicitly, but numeric detection (which is disabled by default) can be enabled to do this automatically:

response = client.indices.create(
  index: 'my-index-000001',
  body: {
    mappings: {
      numeric_detection: true
    }
  }
)
puts response

response = client.index(
  index: 'my-index-000001',
  id: 1,
  body: {
    my_float: '1.0',
    my_integer: '1'
  }
)
puts response

PUT my-index-000001
{
  "mappings": {
    "numeric_detection": true
  }
}

PUT my-index-000001/_doc/1
{
  "my_float":   "1.0", 
  "my_integer": "1" 
}

	The `my_float` field is added as a `float` field.
	The `my_integer` field is added as a `long` field.

« Dynamic mapping Dynamic templates »