lidavidm commented on a change in pull request #10664:
URL: https://github.com/apache/arrow/pull/10664#discussion_r664794191
##########
File path: cpp/src/arrow/dataset/scanner.cc
##########
@@ -729,20 +792,72 @@ Future<std::shared_ptr<Table>> AsyncScanner::ToTableAsync(
});
}
+namespace {
+Result<int64_t> GetSelectionSize(const Datum& selection, int64_t length) {
+ if (length == 0) return 0;
+
+ if (selection.is_scalar()) {
+ if (!selection.scalar()->is_valid) return 0;
+ if (!selection.scalar_as<BooleanScalar>().value) return 0;
+ return length;
+ }
+
+ ARROW_ASSIGN_OR_RAISE(auto count, compute::Sum(selection));
+ return static_cast<int64_t>(count.scalar_as<UInt64Scalar>().value);
+}
+} // namespace
+
Result<int64_t> AsyncScanner::CountRows() {
ARROW_ASSIGN_OR_RAISE(auto fragment_gen, GetFragments());
- ARROW_ASSIGN_OR_RAISE(auto count_gen_gen,
- FragmentsToRowCount(std::move(fragment_gen),
scan_options_));
- auto count_gen = MakeConcatenatedGenerator(std::move(count_gen_gen));
- int64_t total = 0;
- auto sum_fn = [&total](util::optional<int64_t> count) -> Status {
- if (count.has_value()) total += *count;
- return Status::OK();
- };
-
RETURN_NOT_OK(VisitAsyncGenerator<util::optional<int64_t>>(std::move(count_gen),
- std::move(sum_fn))
- .status());
- return total;
+ ARROW_ASSIGN_OR_RAISE(auto plan, compute::ExecPlan::Make());
+
+ std::atomic<int64_t> total{0};
+
+ fragment_gen = MakeMappedGenerator(
+ std::move(fragment_gen), [&](const std::shared_ptr<Fragment>& fragment) {
+ return fragment->CountRows(scan_options_->filter, scan_options_)
+ .Then([&, fragment](util::optional<int64_t> fast_count) mutable
+ -> std::shared_ptr<Fragment> {
+ if (fast_count) {
+ // fast path: got row count directly; skip scanning this
fragment
+ total += *fast_count;
+ return std::make_shared<OneShotFragment>(
+ scan_options_->dataset_schema,
+ MakeEmptyIterator<std::shared_ptr<RecordBatch>>());
+ }
+
+ // slow path: actually filter this fragment's batches
+ return std::move(fragment);
+ });
+ });
+
+ ARROW_ASSIGN_OR_RAISE(auto scan,
+ MakeScanNode(plan.get(), std::move(fragment_gen),
scan_options_));
+
+ ARROW_ASSIGN_OR_RAISE(
+ auto get_selection,
+ compute::MakeProjectNode(scan, "get_selection",
{scan_options_->filter}));
+
+ AsyncGenerator<util::optional<compute::ExecBatch>> sink_gen =
+ compute::MakeSinkNode(get_selection, "sink");
+
+ RETURN_NOT_OK(plan->StartProducing());
+
+ RETURN_NOT_OK(
+ VisitAsyncGenerator(std::move(sink_gen),
+ [&](const util::optional<compute::ExecBatch>& batch)
{
+ // TODO replace with scalar aggregation node
+ ARROW_ASSIGN_OR_RAISE(
+ int64_t slow_count,
+ GetSelectionSize(batch->values[0],
batch->length));
+ total += slow_count;
+ return Status::OK();
+ })
+ .status());
+
+ plan->finished().Wait();
Review comment:
Ah sorry, I didn't realize this can't fail.
--
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]
