Improve this page

Dynamic Typing

One of the most important features of Xapiand is that it tries to get out of your way and let you start exploring your data as quickly as possible. To index a document, you don’t have to first create an index, define a schema, and define your fields; you can just index a document and the index, schema, and fields will spring to life automatically:

PUT /my_index/1
{
  "name": "John Doe",
  "age": 20
}

Creates the /my_index/ index, a schema with two fields: a field called name with datatype text and a field called count with datatype integer.

This automatic detection and addition of new fields is called dynamic typing.

Datatype Detection

By default, when a previously unseen field is found in a document, Xapiand will add the new field to the schema. This behaviour can be disabled globally by running the server in strict mode.

Assuming dynamic typing is enabled, some simple rules are used to determine which datatype the field should have:

JSON datatype	Xapiand datatype
null	No type
Boolean (true or false)	boolean
Floating point number	float
Integer	integer
Object	object
Array	array
String	A datetime if the value passes datetime detection, a float or integer if the value passes numeric detection a UUID if the value passes uuid detection, a text if the value passes text detection or a keyword.

These are the only field datatypes that are dynamically detected. Any other datatypes must be mapped explicitly.

Datetime Detection

If datetime_detection is enabled (the default), then new string fields are checked to see whether their contents match any of the date patterns. If a match is found, a new datetime field is added.

Disabling datetime detection

Dynamic datetime detection can be disabled by setting datetime_detection to false in the index schema:

PUT /my_index/
{
  "_schema": {
    "datetime_detection": false
  }
}

Then, when indexing a new field that looks like a date will not be detected as datetime and instead it will end up being a keyword:

PUT /my_index/1
{
  "create": "2015/09/02"
}

Numeric detection

While JSON has support for native floating point and integer datatypes, 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:

PUT /my_index/
{
  "_schema": {
    "numeric_detection": true
  }
}

The following will detect my_float field to be float and my_integer to be integer:

PUT /my_index/1
{
  "my_float":   "1.0",
  "my_integer": "1"
}

Back

Next