JingsongLi commented on issue #8867: [FLINK-12956][jdbc] Introduce upsert table sink for JDBC URL: https://github.com/apache/flink/pull/8867#issuecomment-507106994 Thanks @wuchong 's great code and review, The reason why I don't introduce JdbcDialect first is: 1.The only difference between databases is just quoteIdentifier and the upsert mechanism now. 2.Spark JdbcDialect only has table-related sql interfaces, If we need DML-related statements, it is difficult to define the relationship between them and data columns. If we must have a upsert statement, but we are not sure whether the implementation of each database requires relevant parameters and whether the order of parameters is the same. But after looking at your prototype, I think we can do some lightweight abstraction. I'll refactor based on your code. And I think JdbcDialect is an interface to users, it should be simpler. We should expose it because there are so many different upsert mechanisms that we don't support.
---------------------------------------------------------------- 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] With regards, Apache Git Services
