rtpsw commented on code in PR #14352:
URL: https://github.com/apache/arrow/pull/14352#discussion_r1027275017
##########
cpp/src/arrow/compute/row/grouper.cc:
##########
@@ -44,7 +45,407 @@ namespace compute {
namespace {
-struct GrouperImpl : Grouper {
+inline const uint8_t* GetValuesAsBytes(const ArrayData& data, int64_t offset =
0) {
+ int64_t absolute_byte_offset = (data.offset + offset) *
data.type->byte_width();
+ return data.GetValues<uint8_t>(1, absolute_byte_offset);
+}
+
+inline const uint8_t* GetValuesAsBytes(const ArraySpan& data, int64_t offset =
0) {
+ int64_t absolute_byte_offset = (data.offset + offset) *
data.type->byte_width();
+ return data.GetValues<uint8_t>(1, absolute_byte_offset);
+}
+
+template <typename Value>
+Status CheckForGetNextSegment(const std::vector<Value>& values, int64_t length,
+ int64_t offset, const std::vector<TypeHolder>&
key_types) {
+ if (offset < 0 || offset > length) {
+ return Status::Invalid("invalid grouping segmenter offset: ", offset);
+ }
+ if (values.size() != key_types.size()) {
+ return Status::Invalid("expected batch size ", key_types.size(), " but got
",
+ values.size());
+ }
+ for (size_t i = 0; i < key_types.size(); i++) {
+ const auto& value = values[i];
+ const auto& key_type = key_types[i];
+ if (*value.type() != *key_type.type) {
+ return Status::Invalid("expected batch value ", i, " of type ",
*key_type.type,
+ " but got ", *value.type());
+ }
+ }
+ return Status::OK();
+}
+
+Status CheckForGetNextSegment(const ExecSpan& batch, int64_t offset,
+ const std::vector<TypeHolder>& key_types) {
+ return CheckForGetNextSegment(batch.values, batch.length, offset, key_types);
+}
+
+Status CheckForGetNextSegment(const ExecBatch& batch, int64_t offset,
+ const std::vector<TypeHolder>& key_types) {
+ return CheckForGetNextSegment(batch.values, batch.length, offset, key_types);
+}
+
+struct StatelessGroupingSegmenter : public GroupingSegmenter {
+ Status Reset() override { return Status::OK(); }
+};
+
+GroupingSegment MakeSegment(int64_t batch_length, int64_t offset, int64_t
length) {
+ return GroupingSegment{offset, length, offset + length >= batch_length};
+}
+
+int64_t GetMatchLength(const uint8_t* match_bytes, int64_t match_width,
+ const uint8_t* array_bytes, int64_t offset, int64_t
length) {
+ int64_t cursor, byte_cursor;
+ for (cursor = offset, byte_cursor = match_width * cursor; cursor < length;
+ cursor++, byte_cursor += match_width) {
+ if (memcmp(match_bytes, array_bytes + byte_cursor,
+ static_cast<size_t>(match_width)) != 0) {
+ break;
Review Comment:
I believe binary-equality is correct because Arrow uses `Hashing32` and
`Hashing64` elsewhere for defining equivalence-classes of keys, and these
hashing methods are sensitive to binary-level differences.
--
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]
