function warehouse

This article introduces the usage scenarios, operation processes and key configurations of the function warehouse function to help users complete daily management and business operations.

Query function warehouse

Custom function reference: https://github.com/ispong/spark-function-template

Add function

Click the Add Function button

• Name: Required. This name will be used as the function name in SQL. The name is unique within the tenant. It is recommended to use all lowercase letters.
• Type: required, supports Scalar User-Defined Functions (UDF), User-Defined Aggregate Functions (UDAF)
• Resource File: Required, Function type file in Resource Center
• Class name: required, the entry point of the custom function, for example: com.isxcode.zhiqingyun.udf.Func
• Result type: required, including string, int, long, double, boolean, date, timestamp
• Remarks: Optional

Supports use in Spark SQL query jobs Note: If the custom function requires dependencies, you need to manually add dependency configuration during use.

Supports use in data synchronization field transformation

Supports use in real-time calculation field transformation

Supports use in Excel import job field conversion

Create a new custom function

Name: to_chinese_date Type: UDF Class name: com.isxcode.zhiqingyun.udf.Func Return type: string Note: Convert English time format to Chinese time

Use custom functions

Custom functions need to first upload the compiled Jar package to the Resource Center Official custom function template: https://github.com/ispong/spark-function-template

• Demo function file download: https://zhiqingyun-saas.isxcode.com/tools/open/file/spark-custom-func.jar
• Demo dependency file download: https://zhiqingyun-saas.isxcode.com/tools/open/file/hutool-all-5.8.27.jar