HeartSaVioR edited a comment on issue #23612: [SPARK-26686][SS] Remove unnecessary KafkaSourceProvider parameter lowercase conversion URL: https://github.com/apache/spark/pull/23612#issuecomment-456610674 IMHO I'm a bit afraid (at least not 100% sure) that CaseInsensitiveMap will provide lowercase keys at all cases. (I'm not an expert of Scala honestly.) I just read MapLike and realize many of other methods rely on `get` and `iterator` so safe to assume they'll provide lowercase key as well, but not sure I'm not missing other methods. Moreover, we don't guarantee providing key-lowercased map in interface level (though Spark provides CaseInsensitiveMap for now). Suppose spark-sql-kafka is an external module outside of Spark codebase: then it would convert parameter to lowercase anyway to guarantee safety. If we would like to guarantee lowercase map in Spark side, then it would be ideal to change the type of parameter (if we are OK to break backward compatibility), or at least we may want to explain it to interface javadoc.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
