amol- commented on a change in pull request #11886:
URL: https://github.com/apache/arrow/pull/11886#discussion_r764329925
##########
File path: cpp/src/arrow/compute/kernels/scalar_validity.cc
##########
@@ -189,6 +191,72 @@ Status ConstBoolExec(KernelContext* ctx, const ExecBatch&
batch, Datum* out) {
return Status::OK();
}
+struct NonZeroVisitor {
+ UInt32Builder *builder;
+ const ArrayData& array;
+
+ NonZeroVisitor(UInt32Builder *builder, const ArrayData& array)
+ : builder(builder), array(array) {}
+
+ Status Visit(const DataType& type) {
+ return Status::TypeError("Unsupported type for nonzero: ",
type.ToString());
+ }
+
+ template <typename Type>
+ enable_if_t<has_c_type<Type>::value &&
+ !std::is_same<Type, MonthDayNanoIntervalType>::value &&
+ !std::is_same<Type, DayTimeIntervalType>::value,
+ Status>
+ Visit(const Type&) {
+ using T = typename GetViewType<Type>::T;
+ uint32_t index = 0;
+
+ return VisitArrayDataInline<Type>(
+ this->array,
+ [&](T v) {
+ if(v != 0) {
+ RETURN_NOT_OK(this->builder->Reserve(1));
+ this->builder->UnsafeAppend(index);
Review comment:
Fair one, it was actually the original implementation but I was getting
a
```
input.array()->length
Which is: 4
expected->length()
Which is: 2
```
error which I thought it was due to the fact that I was reserving 4 entries
even thought I was only willing to write 2. But in the end I found that it was
due to the `CheckScalarUnary` helper and got rid of it. Thus I can now put back
the single `Reserve` call that I had before
--
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]
