What are field data types in Elasticsearch mapping?

In Elasticsearch, mapping defines the data types for each field and how they are indexed and stored. Data types are a critical aspect, as they directly impact indexing methods and search performance. Elasticsearch supports various data types, which can be broadly categorized into the following types:

1. Core Datatypes:

   • String types: Such as text (for full-text search) and keyword (for exact value search, such as filtering and aggregation).
   • Numeric types: Including integer, long, short, byte, double, float, half_float, scaled_float, etc.
   • Date type: date, which stores dates and times.
   • Boolean type: boolean, representing true or false.
   • Binary type: binary, used for storing binary data.

2. Complex Datatypes:

   • Object type: object, used for a single JSON object.
   • Nested type: nested, used for JSON objects within an array, which can be indexed and searched.

3. Geo Datatypes:

   • Geo point type: geo_point, used for storing geographic coordinates (latitude and longitude).
   • Geo shape type: geo_shape, used for storing complex shapes such as polygons.

4. Specialised Datatypes:

   • IP type: ip, used for storing IP addresses.
   • Completion type: completion, used for autocomplete functionality.
   • Token count type: token_count, used for counting tokens in text.
   • Range types: Such as integer_range, float_range, etc., used for storing numerical ranges.

Example: Suppose we need to create an index for an e-commerce website to store product information. Product information includes product name, description, price, and release date. In Elasticsearch, we can design the mapping as follows:

json
{
  "mappings": {
    "properties": {
      "product_name": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 256
          }
        }
      },
      "description": {
        "type": "text"
      },
      "price": {
        "type": "double"
      },
      "release_date": {
        "type": "date",
        "format": "strict_date_optional_time||epoch_millis"
      }
    }
  }
}

Here, the product_name field is of text type for full-text search and additionally defines a keyword subfield for exact search. description is also of text type, suitable for full-text search. price uses double type to store product prices. release_date uses date type, suitable for storing date information.

By selecting appropriate data types, the index structure can meet search requirements while ensuring optimal performance.