[SPARK-29635] Deduplicate test suites between Kafka micro-batch sink and Kafka continuous sink

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 3.0.0
• Fix Version/s: 3.0.0
• Component/s: Structured Streaming
• Assignee: Jungtaek Lim
• Reporter: Jungtaek Lim
• Created: 29/Oct/19 08:59
• Updated: 07/Nov/19 03:08
• Resolved: 07/Nov/19 01:09

Description

There's a comment in KafkaContinuousSinkSuite which is most likely explaining TODO:

https://github.com/apache/spark/blob/37690dea107623ebca1e47c64db59196ee388f2f/external/kafka-0-10-sql/src/test/scala/org/apache/spark/sql/kafka010/KafkaContinuousSinkSuite.scala#L35-L39

/**
 * This is a temporary port of KafkaSinkSuite, since we do not yet have a V2 memory stream.
 * Once we have one, this will be changed to a specialization of KafkaSinkSuite and we won't have
 * to duplicate all the code.
 */

Given latest master branch has V2 memory stream now, it is a good time to deduplicate two suites into one, via having base class and let these suites override necessary things.

Issue Links

• links to: GitHub Pull Request #26292