0AyanamiRei commented on issue #53987: URL: https://github.com/apache/doris/issues/53987#issuecomment-3128200031
I'm still quite unfamiliar with the Doris codebase, so I haven't decided whether to take on this issue yet. However, I have some suggestions regarding modifications to partition pruning. 1. For single-column partitions, we can simply use the column value itself as the partition identifier and apply the partitionPredicate to the partition identifier column during pruning. 2. For multi-column partitions, we can use a `List<a, b, ...>` as the partition identifier, and during pruning, pass a corresponding `List<Pred1, Pred2, ...>`, using two nested for loops for pruning: the outer loop iterates over the partition identifier columns, and the inner loop applies the predicates Pred to each element of the partition identifier list. Another question I have is whether the current design addresses this？When pruning for an SQL statement like `WHERE colA=xxx AND colB=xxx`, does the partitionPredicate passed in accurately match the columns used in partitioning, rather than passing all predicates from the SQL statement and comparing them one by one. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
