terrymanu commented on PR #39131:
URL: https://github.com/apache/shardingsphere/pull/39131#issuecomment-5099857414
Thanks for clarifying. You may remove the current HintManager-specific
plumbing from `GeneralDQLE2EIT` if it is replaced by a focused JDBC E2E
regression test. However, please do not remove the actual HintManager scenario
or leave only the mocked unit test.
The public reproduction confirms that the query succeeds without HintManager
and fails when `HintManager.setDataSourceName(...)` activates the
derived-projection path. Therefore, the committed regression test should:
- Use the actual `MySQL + JDBC + tbl` scenario.
- Call `HintManager.setDataSourceName("tbl")`.
- Execute:
```sql
SELECT T.*, T.status status_new
FROM t_order T
WHERE T.order_id = 1000
```
- Verify the values returned through bare labels such as `order_id`,
`user_id`, `status`, and `status_new`.
- Fail on the merge base and pass with this fix.
Please also ensure that non-target E2E matrices produce zero invocations
instead of failing. The current provider’s `checkState(!matched.isEmpty())`
would fail other database, adapter, and scenario combinations. The test-case
SQL selected by the provider should match the SQL actually executed; the
non-reproducing JOIN case should not be used merely as a parameter carrier.
One additional implementation point: `expandedMap.size() ==
resultSetMetaData.getColumnCount()` does not prove that both sides contain the
same labels. Please cover the equal-size-but-different-label case and
supplement missing metadata labels as well.
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
