Data type map for indexers (Azure AI Search)
When you're building an index schema for indexer-based indexing, the data types in source data must map to an allowed data type for the fields in the target index.
This article provides data type comparisons between SQL Data Types, JSON data types, and Azure AI Search. It contains the following sections:
SQL Server Data Types to Azure AI Search Data Types
| SQL Server Data Type | Allowed target index field types | Notes |
|---|---|---|
| bit | Edm.Boolean, Edm.String | |
| int, smallint, tinyint | Edm.Int32, Edm.Int64, Edm.String | |
| bigint | Edm.Int64, Edm.String | |
| real, float | Edm.Double, Edm.String | |
| smallmoney, money decimal numeric |
Edm.String | Azure AI Search doesn't support converting decimal types into Edm.Double because doing so would lose precision. |
| char, nchar, varchar, nvarchar | Edm.String Collection(Edm.String). See Field Mapping Functions for details on how to transform a string column into a Collection(Edm.String) |
|
| smalldatetime, datetime, datetime2, date, datetimeoffset | Edm.DateTimeOffset, Edm.String | |
| uniqueidentifer | Edm.String | |
| rowversion | N/A | Row-version columns can't be stored in the search index, but they can be used for change tracking. |
| geography | Edm.GeographyPoint, Edm.String | If using geography data types, only geography instances of type POINT with SRID 4326 (which is the default) are supported. If using strings, only GeoJSON points in the following format are supported: {"type": "Point", "coordinates": [long, lat]} |
| time, timespan varbinary image xml geometry CLR types |
N/A | Not supported. |
JSON Data Types to Azure AI Search Data Types
| JSON data type | Allowed target index field types |
|---|---|
| bool | Edm.Boolean, Edm.String |
| Integral numbers | Edm.Int32, Edm.Int64, Edm.String |
| Floating-point numbers | Edm.Double, Edm.String |
| string | Edm.String |
| arrays of primitive types, for example [ "a", "b", "c" ] | Collection(Edm.String) |
| Strings that look like dates | Edm.DateTimeOffset, Edm.String |
| GeoJSON point objects | Edm.GeographyPoint GeoJSON points are JSON objects in the following format: {"type" : "Point", "coordinates": [long, lat]} |
| JSON objects | Edm.ComplexType Azure AI Search maps JSON objects to corresponding complex type schemas |