xiedeyantu commented on PR #4436: URL: https://github.com/apache/calcite/pull/4436#issuecomment-3003537069
> > I think this fix PR is very strange. In the rules, it is judged whether RelNode is registered, and then re-registered. I don't think this is what rewrite rules are supposed to do. > > hi @xiedeyantu, thanks for your attention. Maybe I couldn't explain it well in jira and confused you. Ideally, we should directly modify the config of `FILTER_SUB_QUERY_TO_CORRELATE`, which is straightforward. However, I noticed that not all nodes have parent nodes. Therefore, I added a new config `FILTER_WITH_PARENT_SUB_QUERY_TO_CORRELATE` in SubQueryRemoveRule. To ensure these two rules do not interfere with each other, I tested them both individually and together(test cases like `testEnumerableFilterRule2`, `testEnumerableFilterRule3`, `testEnumerableFilterRule4`). > > Any questions are welcome, if I haven't explained it clearly yet. You explained it very clearly, but I think what this pr does should be placed in the optimizer framework, not a rule. So I think it's strange. -- 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]
