RussellSpitzer opened a new pull request #2579: URL: https://github.com/apache/iceberg/pull/2579
I removed the previous changes to build.gradle for this demonstration, basically what I've done is create a second source set which uses the exact same directories as the normal test configuration. The difference is that the test runtime declares a dependecy on the newer Spark version while still using classes generated by the original iceberg-spark3:compile task. This task will compile the test clases using Spark 3.1 but use the library classes compiled against Spark 3.0. We can use a similar approach for Spark3-Extensions and Spark3-Runtime Jar. One thing we may want to fix here is that I just lazily update the Spark Dependecy by overriding the value in the base configuration. We may want to exclude and then reintroduce it, but that seemed a bit more difficult in my testing. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
