Search-as-you-type field type

edit

Search-as-you-type field type

edit

The search_as_you_type field type is a text-like field that is optimized to provide out-of-the-box support for queries that serve an as-you-type completion use case. It creates a series of subfields that are analyzed to index terms that can be efficiently matched by a query that partially matches the entire indexed text value. Both prefix completion (i.e matching terms starting at the beginning of the input) and infix completion (i.e. matching terms at any position within the input) are supported.

When adding a field of this type to a mapping

resp = client.indices.create(
    index="my-index-000001",
    mappings={
        "properties": {
            "my_field": {
                "type": "search_as_you_type"
            }
        }
    },
)
print(resp)
response = client.indices.create(
  index: 'my-index-000001',
  body: {
    mappings: {
      properties: {
        my_field: {
          type: 'search_as_you_type'
        }
      }
    }
  }
)
puts response
const response = await client.indices.create({
  index: "my-index-000001",
  mappings: {
    properties: {
      my_field: {
        type: "search_as_you_type",
      },
    },
  },
});
console.log(response);
PUT my-index-000001
{
  "mappings": {
    "properties": {
      "my_field": {
        "type": "search_as_you_type"
      }
    }
  }
}

This creates the following fields

my_field

Analyzed as configured in the mapping. If an analyzer is not configured, the default analyzer for the index is used

my_field._2gram

Wraps the analyzer of my_field with a shingle token filter of shingle size 2

my_field._3gram

Wraps the analyzer of my_field with a shingle token filter of shingle size 3

my_field._index_prefix

Wraps the analyzer of my_field._3gram with an edge ngram token filter

The size of shingles in subfields can be configured with the max_shingle_size mapping parameter. The default is 3, and valid values for this parameter are integer values 2 - 4 inclusive. Shingle subfields will be created for each shingle size from 2 up to and including the max_shingle_size. The my_field._index_prefix subfield will always use the analyzer from the shingle subfield with the max_shingle_size when constructing its own analyzer.

Increasing the max_shingle_size will improve matches for queries with more consecutive terms, at the cost of larger index size. The default max_shingle_size should usually be sufficient.

The same input text is indexed into each of these fields automatically, with their differing analysis chains, when an indexed document has a value for the root field my_field.

resp = client.index(
    index="my-index-000001",
    id="1",
    refresh=True,
    document={
        "my_field": "quick brown fox jump lazy dog"
    },
)
print(resp)
response = client.index(
  index: 'my-index-000001',
  id: 1,
  refresh: true,
  body: {
    my_field: 'quick brown fox jump lazy dog'
  }
)
puts response
const response = await client.index({
  index: "my-index-000001",
  id: 1,
  refresh: "true",
  document: {
    my_field: "quick brown fox jump lazy dog",
  },
});
console.log(response);
PUT my-index-000001/_doc/1?refresh
{
  "my_field": "quick brown fox jump lazy dog"
}

The most efficient way of querying to serve a search-as-you-type use case is usually a multi_match query of type bool_prefix that targets the root search_as_you_type field and its shingle subfields. This can match the query terms in any order, but will score documents higher if they contain the terms in order in a shingle subfield.

resp = client.search(
    index="my-index-000001",
    query={
        "multi_match": {
            "query": "brown f",
            "type": "bool_prefix",
            "fields": [
                "my_field",
                "my_field._2gram",
                "my_field._3gram"
            ]
        }
    },
)
print(resp)
response = client.search(
  index: 'my-index-000001',
  body: {
    query: {
      multi_match: {
        query: 'brown f',
        type: 'bool_prefix',
        fields: [
          'my_field',
          'my_field._2gram',
          'my_field._3gram'
        ]
      }
    }
  }
)
puts response
const response = await client.search({
  index: "my-index-000001",
  query: {
    multi_match: {
      query: "brown f",
      type: "bool_prefix",
      fields: ["my_field", "my_field._2gram", "my_field._3gram"],
    },
  },
});
console.log(response);
GET my-index-000001/_search
{
  "query": {
    "multi_match": {
      "query": "brown f",
      "type": "bool_prefix",
      "fields": [
        "my_field",
        "my_field._2gram",
        "my_field._3gram"
      ]
    }
  }
}
{
  "took" : 44,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 1,
      "relation" : "eq"
    },
    "max_score" : 0.8630463,
    "hits" : [
      {
        "_index" : "my-index-000001",
        "_id" : "1",
        "_score" : 0.8630463,
        "_source" : {
          "my_field" : "quick brown fox jump lazy dog"
        }
      }
    ]
  }
}

To search for documents that strictly match the query terms in order, or to search using other properties of phrase queries, use a match_phrase_prefix query on the root field. A match_phrase query can also be used if the last term should be matched exactly, and not as a prefix. Using phrase queries may be less efficient than using the match_bool_prefix query.

resp = client.search(
    index="my-index-000001",
    query={
        "match_phrase_prefix": {
            "my_field": "brown f"
        }
    },
)
print(resp)
response = client.search(
  index: 'my-index-000001',
  body: {
    query: {
      match_phrase_prefix: {
        my_field: 'brown f'
      }
    }
  }
)
puts response
const response = await client.search({
  index: "my-index-000001",
  query: {
    match_phrase_prefix: {
      my_field: "brown f",
    },
  },
});
console.log(response);
GET my-index-000001/_search
{
  "query": {
    "match_phrase_prefix": {
      "my_field": "brown f"
    }
  }
}

Parameters specific to the search_as_you_type field

edit

The following parameters are accepted in a mapping for the search_as_you_type field and are specific to this field type

max_shingle_size

(Optional, integer) Largest shingle size to create. Valid values are 2 (inclusive) to 4 (inclusive). Defaults to 3.

A subfield is created for each integer between 2 and this value. For example, a value of 3 creates two subfields: my_field._2gram and my_field._3gram

More subfields enables more specific queries but increases index size.

Parameters of the field type as a text field

edit

The following parameters are accepted in a mapping for the search_as_you_type field due to its nature as a text-like field, and behave similarly to their behavior when configuring a field of the text data type. Unless otherwise noted, these options configure the root fields subfields in the same way.

analyzer
The analyzer which should be used for text fields, both at index-time and at search-time (unless overridden by the search_analyzer). Defaults to the default index analyzer, or the standard analyzer.
index
Should the field be searchable? Accepts true (default) or false.
index_options
What information should be stored in the index, for search and highlighting purposes. Defaults to positions.
norms
Whether field-length should be taken into account when scoring queries. Accepts true or false. This option configures the root field and shingle subfields, where its default is true. It does not configure the prefix subfield, where it is false.
store
Whether the field value should be stored and retrievable separately from the _source field. Accepts true or false (default). This option only configures the root field, and does not configure any subfields.
search_analyzer
The analyzer that should be used at search time on text fields. Defaults to the analyzer setting.
search_quote_analyzer
The analyzer that should be used at search time when a phrase is encountered. Defaults to the search_analyzer setting.
similarity
Which scoring algorithm or similarity should be used. Defaults to BM25.
term_vector
Whether term vectors should be stored for the field. Defaults to no. This option configures the root field and shingle subfields, but not the prefix subfield.

Optimization of prefix queries

edit

When making a prefix query to the root field or any of its subfields, the query will be rewritten to a term query on the ._index_prefix subfield. This matches more efficiently than is typical of prefix queries on text fields, as prefixes up to a certain length of each shingle are indexed directly as terms in the ._index_prefix subfield.

The analyzer of the ._index_prefix subfield slightly modifies the shingle-building behavior to also index prefixes of the terms at the end of the field’s value that normally would not be produced as shingles. For example, if the value quick brown fox is indexed into a search_as_you_type field with max_shingle_size of 3, prefixes for brown fox and fox are also indexed into the ._index_prefix subfield even though they do not appear as terms in the ._3gram subfield. This allows for completion of all the terms in the field’s input.

Synthetic _source

edit

Synthetic _source is Generally Available only for TSDB indices (indices that have index.mode set to time_series). For other indices synthetic _source is in technical preview. Features in technical preview may be changed or removed in a future release. Elastic will work to fix any issues, but features in technical preview are not subject to the support SLA of official GA features.

search_as_you_type fields support synthetic _source in their default configuration. Synthetic _source cannot be used together with copy_to.