lidavidm commented on a change in pull request #11452:
URL: https://github.com/apache/arrow/pull/11452#discussion_r741528772
##########
File path: cpp/src/arrow/compute/kernels/hash_aggregate.cc
##########
@@ -1677,6 +1703,177 @@ struct GroupedMinMaxImpl final : public
GroupedAggregator {
ScalarAggregateOptions options_;
};
+// For binary-like types
+// In principle, FixedSizeBinary could use base implementation
+template <typename Type>
+struct GroupedMinMaxImpl<Type,
+ enable_if_t<is_base_binary_type<Type>::value ||
+ std::is_same<Type,
FixedSizeBinaryType>::value>>
+ final : public GroupedAggregator {
+ Status Init(ExecContext* ctx, const FunctionOptions* options) override {
+ ctx_ = ctx;
+ options_ = *checked_cast<const ScalarAggregateOptions*>(options);
+ // type_ initialized by MinMaxInit
+ has_values_ = TypedBufferBuilder<bool>(ctx->memory_pool());
+ has_nulls_ = TypedBufferBuilder<bool>(ctx->memory_pool());
+ return Status::OK();
+ }
+
+ Status Resize(int64_t new_num_groups) override {
+ auto added_groups = new_num_groups - num_groups_;
Review comment:
We don't actually care about reducing capacity, so long as the vector is
at least big enough. (Also I don't think it's possible to reduce capacity; the
number of groups can only ever increase.)
--
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]
