wuchong commented on a change in pull request #11538: [FLINK-16813][jdbc]
JDBCInputFormat doesn't correctly map Short
URL: https://github.com/apache/flink/pull/11538#discussion_r401435369
##########
File path:
flink-connectors/flink-jdbc/src/main/java/org/apache/flink/api/java/io/jdbc/JDBCInputFormat.java
##########
@@ -127,7 +134,7 @@ public JDBCInputFormat() {
@Override
public RowTypeInfo getProducedType() {
- return rowTypeInfo;
+ return (RowTypeInfo)
fromDataTypeToLegacyInfo(fromLogicalToDataType(rowType));
Review comment:
Yes. They are orthogonal. The reason we should use `TypeInformation` instead
of `RowType` is `InputFormat` only works with `TypeInformation` and `RowType`
is a concept of table world.
The community is trying to reuse runtime implementation as much as possible.
So a better design is that `JDBCInputFormat` accepts result `TypeInformation`
and `JDBCConverter` as constructor parameters. So that, we can easily support
generic type for DataStream users in the futrue. We already refactored JDBC
sink function to support generic type (see `JdbcBatchingOutputFormat`), and
will refactor JDBC source soon.
----------------------------------------------------------------
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
