Mapping Columns in Snowflake as a Source
- 1 Minute to read
- Print
- DarkLight
- PDF
Mapping Columns in Snowflake as a Source
- 1 Minute to read
- Print
- DarkLight
- PDF
Article summary
Did you find this summary helpful?
Thank you for your feedback
Overview
This document outlines the mapping of Snowflake data types to Rivery data types. It also defines the behavior for handling unsupported data types in various scenarios.
Data Type Mapping
The following table describes the mapping of Snowflake data types to their corresponding Rivery data types:
| Snowflake as a Source Data Type | Rivery Data Type |
|---|---|
| NUMBER (up to 9 decimal places) | SHORTDECIMAL |
| NUMBER (more than 9 decimal places) | LONGDECIMAL |
| FLOAT | FLOAT |
| BOOLEAN | BOOLEAN |
| DATE | DATE |
| TIME | TIME |
| TIMESTAMP | TIMESTAMP |
| VARCHAR | VARCHAR |
| BINARY | VARCHAR |
| VARIANT | VARCHAR |
| OBJECT | VARCHAR |
| ARRAY | VARCHAR |
Expected Behavior for Unsupported Data Types
Multi-table Metadata:
- When returning metadata for tables in a multi-table query, map the data type to the closest available datatype in Rivery. If no close match is found, use "STRING".
Custom Queries:
- For unsupported data types encountered in custom queries, map them to the closest available datatype in Rivery. If no close match is available, use "STRING".
Failure Handling:
- Rivery should not fail due to unsupported data types. The system should handle these gracefully according to the mapping rules.
Manual Mapping Changes:
- If a user manually changes the mapping to an unsupported datatype in the target, Rivery should fail with a clear and appropriate error message.
Was this article helpful?