DataSQRL / sqrl

Flexible development framework for building streaming data applications in SQL with Kafka, Flink, Postgres, GraphQL, and more.
https://www.datasqrl.com/
97 stars 14 forks source link

Consistent jar handling for Local and Remote UDFs #848

Closed nandorsoma closed 1 month ago

nandorsoma commented 1 month ago

The current build process for local UDFs generates a .function.json file in the build folder, but the JAR is not copied there and is instead used from the root folder. For remote UDFs, while the JARs are placed in the build folder, the jarPath in .function.json incorrectly points to the root folder, leading to inconsistent usage.

Objective:

henneberger commented 1 month ago

UDF jars have no path isolation so their path is unimportant, we can remove it from the function manifest. They should all be copied to the lib folder.