-
Notifications
You must be signed in to change notification settings - Fork 559
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
correct extract column methods (#6133)
- Loading branch information
1 parent
48a2519
commit 6e81fbe
Showing
21 changed files
with
246 additions
and
221 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,141 @@ | ||
#include "process_columns.h" | ||
#include "common/adapter.h" | ||
|
||
#include <util/string/join.h> | ||
|
||
namespace NKikimr::NArrow { | ||
|
||
namespace { | ||
template <class TDataContainer, class TStringImpl> | ||
std::shared_ptr<TDataContainer> ExtractColumnsValidateImpl(const std::shared_ptr<TDataContainer>& srcBatch, | ||
const std::vector<TStringImpl>& columnNames) { | ||
std::vector<std::shared_ptr<arrow::Field>> fields; | ||
fields.reserve(columnNames.size()); | ||
std::vector<std::shared_ptr<typename NAdapter::TDataBuilderPolicy<TDataContainer>::TColumn>> columns; | ||
columns.reserve(columnNames.size()); | ||
|
||
auto srcSchema = srcBatch->schema(); | ||
for (auto& name : columnNames) { | ||
const int pos = srcSchema->GetFieldIndex(name); | ||
if (Y_LIKELY(pos > -1)) { | ||
fields.push_back(srcSchema->field(pos)); | ||
columns.push_back(srcBatch->column(pos)); | ||
} | ||
} | ||
|
||
return NAdapter::TDataBuilderPolicy<TDataContainer>::Build(std::move(fields), std::move(columns), srcBatch->num_rows()); | ||
} | ||
|
||
template <class TDataContainer> | ||
TConclusion<std::shared_ptr<TDataContainer>> AdaptColumnsImpl(const std::shared_ptr<TDataContainer>& srcBatch, | ||
const std::shared_ptr<arrow::Schema>& dstSchema) { | ||
AFL_VERIFY(srcBatch); | ||
AFL_VERIFY(dstSchema); | ||
std::vector<std::shared_ptr<typename NAdapter::TDataBuilderPolicy<TDataContainer>::TColumn>> columns; | ||
columns.reserve(dstSchema->num_fields()); | ||
|
||
for (auto& field : dstSchema->fields()) { | ||
const int index = srcBatch->schema()->GetFieldIndex(field->name()); | ||
if (index > -1) { | ||
columns.push_back(srcBatch->column(index)); | ||
auto srcField = srcBatch->schema()->field(index); | ||
if (field->Equals(srcField)) { | ||
AFL_VERIFY(columns.back()->type()->Equals(field->type()))("event", "cannot_use_incoming_batch")("reason", "invalid_column_type")("column", field->name()) | ||
("column_type", field->type()->ToString())("incoming_type", columns.back()->type()->ToString()); | ||
} else { | ||
AFL_ERROR(NKikimrServices::ARROW_HELPER)("event", "cannot_use_incoming_batch")("reason", "invalid_column_type")("column", field->name()) | ||
("column_type", field->ToString(true))("incoming_type", srcField->ToString(true)); | ||
return TConclusionStatus::Fail("incompatible column types"); | ||
} | ||
} else { | ||
AFL_ERROR(NKikimrServices::ARROW_HELPER)("event", "not_found_column")("column", field->name()) | ||
("column_type", field->type()->ToString())("columns", JoinSeq(",", srcBatch->schema()->field_names())); | ||
return TConclusionStatus::Fail("not found column '" + field->name() + "'"); | ||
} | ||
} | ||
|
||
return NAdapter::TDataBuilderPolicy<TDataContainer>::Build(dstSchema, std::move(columns), srcBatch->num_rows()); | ||
} | ||
|
||
template <class TDataContainer, class TStringType> | ||
std::shared_ptr<TDataContainer> ExtractImpl(const TColumnOperator::EExtractProblemsPolicy& policy, | ||
const std::shared_ptr<TDataContainer>& incoming, const std::vector<TStringType>& columnNames) { | ||
AFL_VERIFY(incoming); | ||
AFL_VERIFY(columnNames.size()); | ||
auto result = ExtractColumnsValidateImpl(incoming, columnNames); | ||
switch (policy) { | ||
case TColumnOperator::EExtractProblemsPolicy::Verify: | ||
AFL_VERIFY((ui32)result->num_columns() == columnNames.size())("schema", incoming->schema()->ToString())("required", JoinSeq(",", columnNames)); | ||
break; | ||
case TColumnOperator::EExtractProblemsPolicy::Null: | ||
if ((ui32)result->num_columns() != columnNames.size()) { | ||
return nullptr; | ||
} | ||
break; | ||
case TColumnOperator::EExtractProblemsPolicy::Skip: | ||
break; | ||
} | ||
return result; | ||
} | ||
|
||
template <class TDataContainer, class TStringType> | ||
TConclusion<std::shared_ptr<TDataContainer>> ReorderImpl(const std::shared_ptr<TDataContainer>& incoming, const std::vector<TStringType>& columnNames) { | ||
AFL_VERIFY(!!incoming); | ||
AFL_VERIFY(columnNames.size()); | ||
if ((ui32)incoming->num_columns() < columnNames.size()) { | ||
return TConclusionStatus::Fail("not enough columns for exact reordering"); | ||
} | ||
if ((ui32)incoming->num_columns() > columnNames.size()) { | ||
return TConclusionStatus::Fail("need extraction before reorder call"); | ||
} | ||
auto result = ExtractColumnsValidateImpl(incoming, columnNames); | ||
AFL_VERIFY(result); | ||
if ((ui32)result->num_columns() != columnNames.size()) { | ||
return TConclusionStatus::Fail("not enough fields for exact reordering"); | ||
} | ||
return result; | ||
} | ||
|
||
} | ||
|
||
std::shared_ptr<arrow::RecordBatch> TColumnOperator::Extract(const std::shared_ptr<arrow::RecordBatch>& incoming, const std::vector<std::string>& columnNames) { | ||
return ExtractImpl(AbsentColumnPolicy, incoming, columnNames); | ||
} | ||
|
||
std::shared_ptr<arrow::Table> TColumnOperator::Extract(const std::shared_ptr<arrow::Table>& incoming, const std::vector<std::string>& columnNames) { | ||
return ExtractImpl(AbsentColumnPolicy, incoming, columnNames); | ||
} | ||
|
||
std::shared_ptr<arrow::RecordBatch> TColumnOperator::Extract(const std::shared_ptr<arrow::RecordBatch>& incoming, const std::vector<TString>& columnNames) { | ||
return ExtractImpl(AbsentColumnPolicy, incoming, columnNames); | ||
} | ||
|
||
std::shared_ptr<arrow::Table> TColumnOperator::Extract(const std::shared_ptr<arrow::Table>& incoming, const std::vector<TString>& columnNames) { | ||
return ExtractImpl(AbsentColumnPolicy, incoming, columnNames); | ||
} | ||
|
||
NKikimr::TConclusion<std::shared_ptr<arrow::RecordBatch>> TColumnOperator::Adapt(const std::shared_ptr<arrow::RecordBatch>& incoming, const std::shared_ptr<arrow::Schema>& dstSchema) { | ||
return AdaptColumnsImpl(incoming, dstSchema); | ||
} | ||
|
||
NKikimr::TConclusion<std::shared_ptr<arrow::Table>> TColumnOperator::Adapt(const std::shared_ptr<arrow::Table>& incoming, const std::shared_ptr<arrow::Schema>& dstSchema) { | ||
return AdaptColumnsImpl(incoming, dstSchema); | ||
} | ||
|
||
NKikimr::TConclusion<std::shared_ptr<arrow::RecordBatch>> TColumnOperator::Reorder(const std::shared_ptr<arrow::RecordBatch>& incoming, const std::vector<std::string>& columnNames) { | ||
return ReorderImpl(incoming, columnNames); | ||
} | ||
|
||
NKikimr::TConclusion<std::shared_ptr<arrow::Table>> TColumnOperator::Reorder(const std::shared_ptr<arrow::Table>& incoming, const std::vector<std::string>& columnNames) { | ||
return ReorderImpl(incoming, columnNames); | ||
} | ||
|
||
NKikimr::TConclusion<std::shared_ptr<arrow::RecordBatch>> TColumnOperator::Reorder(const std::shared_ptr<arrow::RecordBatch>& incoming, const std::vector<TString>& columnNames) { | ||
return ReorderImpl(incoming, columnNames); | ||
} | ||
|
||
NKikimr::TConclusion<std::shared_ptr<arrow::Table>> TColumnOperator::Reorder(const std::shared_ptr<arrow::Table>& incoming, const std::vector<TString>& columnNames) { | ||
return ReorderImpl(incoming, columnNames); | ||
} | ||
|
||
} |
Oops, something went wrong.