zanmato1984 commented on code in PR #43389:
URL: https://github.com/apache/arrow/pull/43389#discussion_r1718471071
##########
cpp/src/arrow/compute/row/compare_test.cc:
##########
@@ -166,138 +171,306 @@ TEST(KeyCompare, CompareColumnsToRowsTempStackUsage) {
}
}
+namespace {
+
+Result<RowTableImpl> MakeRowTableFromExecBatch(const ExecBatch& batch) {
+ RowTableImpl row_table;
+
+ std::vector<KeyColumnMetadata> column_metadatas;
+ RETURN_NOT_OK(ColumnMetadatasFromExecBatch(batch, &column_metadatas));
+ RowTableMetadata table_metadata;
+ table_metadata.FromColumnMetadataVector(column_metadatas, sizeof(uint64_t),
+ sizeof(uint64_t));
+ RETURN_NOT_OK(row_table.Init(default_memory_pool(), table_metadata));
+ std::vector<uint16_t> row_ids(batch.length);
+ std::iota(row_ids.begin(), row_ids.end(), 0);
+ RowTableEncoder row_encoder;
+ row_encoder.Init(column_metadatas, sizeof(uint64_t), sizeof(uint64_t));
+ std::vector<KeyColumnArray> column_arrays;
+ RETURN_NOT_OK(ColumnArraysFromExecBatch(batch, &column_arrays));
+ row_encoder.PrepareEncodeSelected(0, batch.length, column_arrays);
+ RETURN_NOT_OK(row_encoder.EncodeSelected(
+ &row_table, static_cast<uint32_t>(batch.length), row_ids.data()));
+
+ return row_table;
+}
+
+Result<RowTableImpl> RepeatRowTableUntil(const RowTableImpl& seed, int64_t
num_rows) {
+ RowTableImpl row_table;
+
+ RETURN_NOT_OK(row_table.Init(default_memory_pool(), seed.metadata()));
+ // Append the seed row table repeatedly to grow the row table to big enough.
+ while (row_table.length() < num_rows) {
+ RETURN_NOT_OK(row_table.AppendSelectionFrom(seed,
+
static_cast<uint32_t>(seed.length()),
+ /*source_row_ids=*/NULLPTR));
+ }
+
+ return row_table;
+}
+
+void AssertCompareColumnsToRowsAllMatch(const std::vector<KeyColumnArray>&
columns,
+ const RowTableImpl& row_table,
+ const std::vector<uint32_t>&
row_ids_to_compare) {
+ uint32_t num_rows_to_compare =
static_cast<uint32_t>(row_ids_to_compare.size());
+
+ TempVectorStack stack;
+ ASSERT_OK(
+ stack.Init(default_memory_pool(),
+
KeyCompare::CompareColumnsToRowsTempStackUsage(num_rows_to_compare)));
+ LightContext ctx{CpuInfo::GetInstance()->hardware_flags(), &stack};
+
+ {
+ // No selection, output no match row ids.
+ uint32_t num_rows_no_match;
+ std::vector<uint16_t> row_ids_out(num_rows_to_compare);
+ KeyCompare::CompareColumnsToRows(num_rows_to_compare,
/*sel_left_maybe_null=*/NULLPTR,
+ row_ids_to_compare.data(), &ctx,
&num_rows_no_match,
+ row_ids_out.data(), columns, row_table,
+ /*are_cols_in_encoding_order=*/true,
+
/*out_match_bitvector_maybe_null=*/NULLPTR);
+ ASSERT_EQ(num_rows_no_match, 0);
+ }
+
+ {
+ // No selection, output match bit vector.
+ std::vector<uint8_t> match_bitvector(BytesForBits(num_rows_to_compare));
+ KeyCompare::CompareColumnsToRows(
+ num_rows_to_compare, /*sel_left_maybe_null=*/NULLPTR,
row_ids_to_compare.data(),
+ &ctx,
+ /*out_num_rows=*/NULLPTR, /*out_sel_left_maybe_same=*/NULLPTR,
columns, row_table,
+ /*are_cols_in_encoding_order=*/true, match_bitvector.data());
+ ASSERT_EQ(CountSetBits(match_bitvector.data(), 0, num_rows_to_compare),
+ num_rows_to_compare);
Review Comment:
The doc string is misleading. It actually returns the result in either of
two ways: selection vector of the non-match row ids (as the doc string says),
and the match bit vector (1 for match and 0 otherwise, this is not mentioned in
the doc string), depending on if the output parameter of row ids is not null.
And this code is checking using the second way.
I can update the doc string to make it more clear.
--
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]
