lidavidm commented on a change in pull request #12162:
URL: https://github.com/apache/arrow/pull/12162#discussion_r786082875
##########
File path: cpp/src/arrow/compute/kernels/scalar_nested.cc
##########
@@ -429,6 +430,97 @@ const FunctionDoc make_struct_doc{"Wrap Arrays into a
StructArray",
{"*args"},
"MakeStructOptions"};
+struct MapArrayLookupFunctor {
+ static Status ExecMapArray(KernelContext* ctx, const ExecBatch& batch,
Datum* out) {
+ const auto& options = OptionsWrapper<MapArrayLookupOptions>::Get(ctx);
+
+ MapArray map_array(batch[0].array());
+
+ // Offset differences will tell the number of Strcut = {key, value} pairs
+ // present in the current list.
+ // const std::shared_ptr<arrow::Buffer> offsets =
map_array.value_offsets();
+
+ std::shared_ptr<arrow::Array> keys = map_array.keys();
+ std::shared_ptr<arrow::Array> items = map_array.items();
+
+ const auto& query_key = options.query_key;
+ const auto& occurence = options.occurence;
+
+ std::unique_ptr<ArrayBuilder> builder;
+ RETURN_NOT_OK(
+ MakeBuilder(ctx->memory_pool(), map_array.map_type()->item_type(),
&builder));
+
+ int32_t last_key_idx_checked = 0;
+
+ // aka, number of {key, value} pairs in the current map
+ int32_t list_struct_len;
+ bool found_one_key = false;
+ for (int32_t map_array_idx = 0; map_array_idx < map_array.length();
++map_array_idx) {
+ // Number of Struct('s) = {key, value} in the list at the current index
+ list_struct_len = map_array.value_length(map_array_idx);
+ for (int32_t key_idx_to_check = last_key_idx_checked;
+ key_idx_to_check < last_key_idx_checked + list_struct_len;
+ ++key_idx_to_check) {
+ ARROW_ASSIGN_OR_RAISE(std::shared_ptr<Scalar> key,
+ keys->GetScalar(key_idx_to_check));
+ if (key->Equals(*query_key)) {
+ std::cout << "Key being checked: " << key->ToString() << "\n";
+ ARROW_ASSIGN_OR_RAISE(std::shared_ptr<Scalar> item,
+ items->GetScalar(key_idx_to_check));
+ std::cout << "Value at key: " << item->ToString() << "\n";
+ ARROW_ASSIGN_OR_RAISE(auto value,
+
item->CastTo(map_array.map_type()->item_type()));
+
+ std::cout << "Item being appended: " << value->ToString() << "\n";
+ RETURN_NOT_OK(builder->AppendScalar(*value));
+
+ if (occurence == MapArrayLookupOptions::First) {
+ found_one_key = true;
+ break;
+ }
+ }
+ }
+ if (found_one_key && occurence == MapArrayLookupOptions::First) break;
Review comment:
Ignore `All` for now. Let's get `First` to work and then we can see how
`All` works.
The logic here looks reasonable, though maybe I'm missing something as you
should have one output per input row and the test cases below don't reflect
that.
--
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: github-unsubscr...@arrow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
