Design Guidelines and Constraints for Functions and Procedures¶
This topic describes constraints and guidelines to keep in mind when writing UDFs and stored procedures.
- Keeping handler code in-line or on a stage
Choose whether to have your handler code in-line or packaged in a separate file.
- Designing Handlers that Stay Within Snowflake-Imposed Constraints
Ensure stability within the Snowflake environment by developing within constraints described in this topic.
- Naming and overloading procedures and UDFs
Learn the rules for naming and overloading procedures and UDFs.
- Defining arguments for UDFs and stored procedures
Specify the arguments for your procedures and UDFs.
- Data Type Mappings Between SQL and Handler Languages
Choose the best data types for argument and return values in handler code.
- Making dependencies available to your code
Make your handler or its dependencies available for use at runtime on Snowflake.
Security¶
- Security Practices for UDFs and Procedures
Help your handler code execute securely using these best practices.
- Protecting Sensitive Information with Secure UDFs and Stored Procedures
Ensure that sensitive information is concealed from users who should not have access to it.
- Pushdown Optimization and Data Visibility
Learn about the pushdown optimization that makes queries more efficient, but which can also expose data that you might not want to be visible.