ClSlaid commented on code in PR #9755:
URL: https://github.com/apache/arrow-rs/pull/9755#discussion_r3345743338
##########
arrow-select/src/coalesce.rs:
##########
@@ -145,6 +149,19 @@ pub struct BatchCoalescer {
completed: VecDeque<RecordBatch>,
/// Biggest coalesce batch size. See
[`Self::with_biggest_coalesce_batch_size`]
biggest_coalesce_batch_size: Option<usize>,
+ /// Inline view columns eligible for the fused filter path.
+ inline_view_columns: Option<InlineViewColumns>,
+}
+
+/// Inline Utf8View/BinaryView columns in a schema.
+///
+/// These columns can copy filtered views directly into the coalescer without
Review Comment:
This is an optimization I've done too early. I was afraid iterating over
columns' types would cost too much.
After benchmarking, I found it is not the bottleneck at all, and it will
draw the performance instead.
I'll remove it in the next commit.
--
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]
