westonpace commented on a change in pull request #9945:
URL: https://github.com/apache/arrow/pull/9945#discussion_r609964135
##########
File path: cpp/src/arrow/util/async_generator.h
##########
@@ -1063,6 +1063,86 @@ AsyncGenerator<T>
MakeConcatenatedGenerator(AsyncGenerator<AsyncGenerator<T>> so
return MergedGenerator<T>(std::move(source), 1);
}
+template <typename T>
+struct Enumerated {
+ util::optional<T> value;
+ int index;
+ bool last;
+};
+
+template <typename T>
+struct IterationTraits<Enumerated<T>> {
+ static Enumerated<T> End() { return Enumerated<T>{{}, -1, false}; }
+ static bool IsEnd(const Enumerated<T>& val) { return !val.value.has_value();
}
+};
+
+/// \see MakeEnumeratedGenerator
+template <typename T>
+class EnumeratingGenerator {
+ public:
+ EnumeratingGenerator(AsyncGenerator<T> source, T initial_value)
+ : state_(std::make_shared<State>(std::move(source),
std::move(initial_value))) {}
+
+ Future<Enumerated<T>> operator()() {
+ if (state_->finished) {
+ return AsyncGeneratorEnd<Enumerated<T>>();
+ } else {
+ auto state = state_;
+ return state->source().Then([state](const T& next) {
+ auto finished = IsIterationEnd<T>(next);
+ auto prev = Enumerated<T>{state->prev_value, state->prev_index,
finished};
Review comment:
When all is said and done what I have in the stream is something that
looks like
fragment 2 batch 1 false // Buffer for resequencing
fragment 1 batch 1 false // Send downstream, first batch
fragment 1 batch 2 true // Send downstream, this is the next batch, then
send f2b1 because I see this is the last batch in f1
fragment 2 batch 2 false // Send downstream, this follows the last batch we
sent (f2b1)
--
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.
For queries about this service, please contact Infrastructure at:
[email protected]
