mathyingzhou commented on a change in pull request #8648:
URL: https://github.com/apache/arrow/pull/8648#discussion_r560634894
##########
File path: cpp/src/arrow/adapters/orc/adapter_util.cc
##########
@@ -316,10 +326,482 @@ Status AppendBatch(const liborc::Type* type,
liborc::ColumnVectorBatch* batch,
}
}
+template <class array_type, class batch_type>
+Status FillNumericBatch(const DataType* type, liborc::ColumnVectorBatch*
cbatch,
+ int64_t& arrowOffset, int64_t& orcOffset, int64_t
length,
+ Array* parray, std::vector<bool>* incomingMask) {
+ auto array = checked_cast<array_type*>(parray);
+ auto batch = checked_cast<batch_type*>(cbatch);
+ int64_t arrowLength = array->length();
+ if (!arrowLength) return Status::OK();
+ if (array->null_count() || incomingMask) batch->hasNulls = true;
+ for (; orcOffset < length && arrowOffset < arrowLength; orcOffset++,
arrowOffset++) {
+ if (array->IsNull(arrowOffset) || (incomingMask &&
!(*incomingMask)[orcOffset])) {
+ batch->notNull[orcOffset] = false;
+ } else {
+ batch->data[orcOffset] = array->Value(arrowOffset);
+ batch->notNull[orcOffset] = true;
+ }
+ }
+ batch->numElements = orcOffset;
+ return Status::OK();
+}
+
+template <class array_type, class batch_type, class target_type>
+Status FillNumericBatchCast(const DataType* type, liborc::ColumnVectorBatch*
cbatch,
+ int64_t& arrowOffset, int64_t& orcOffset, int64_t
length,
+ Array* parray, std::vector<bool>* incomingMask) {
+ auto array = checked_cast<array_type*>(parray);
+ auto batch = checked_cast<batch_type*>(cbatch);
+ int64_t arrowLength = array->length();
+ if (!arrowLength) return Status::OK();
+ if (array->null_count() || incomingMask) batch->hasNulls = true;
+ for (; orcOffset < length && arrowOffset < arrowLength; orcOffset++,
arrowOffset++) {
+ if (array->IsNull(arrowOffset) || (incomingMask &&
!(*incomingMask)[orcOffset])) {
+ batch->notNull[orcOffset] = false;
+ } else {
+ batch->data[orcOffset] =
static_cast<target_type>(array->Value(arrowOffset));
+ batch->notNull[orcOffset] = true;
+ }
+ }
+ batch->numElements = orcOffset;
+ return Status::OK();
+}
+
+Status FillDate64Batch(const DataType* type, liborc::ColumnVectorBatch* cbatch,
+ int64_t& arrowOffset, int64_t& orcOffset, int64_t
length,
+ Array* parray, std::vector<bool>* incomingMask) {
+ auto array = checked_cast<Date64Array*>(parray);
+ auto batch = checked_cast<liborc::TimestampVectorBatch*>(cbatch);
+ int64_t arrowLength = array->length();
+ if (!arrowLength) return Status::OK();
+ if (array->null_count() || incomingMask) batch->hasNulls = true;
+ for (; orcOffset < length && arrowOffset < arrowLength; orcOffset++,
arrowOffset++) {
+ if (array->IsNull(arrowOffset) || (incomingMask &&
!(*incomingMask)[orcOffset])) {
+ batch->notNull[orcOffset] = false;
+ } else {
+ int64_t miliseconds = array->Value(arrowOffset);
+ batch->data[orcOffset] =
+ static_cast<int64_t>(std::floor(miliseconds / kOneSecondMillis));
+ batch->nanoseconds[orcOffset] =
+ (miliseconds - kOneSecondMillis * batch->data[orcOffset]) *
kOneMilliNanos;
+ batch->notNull[orcOffset] = true;
+ }
+ }
+ batch->numElements = orcOffset;
+ return Status::OK();
+}
+
+Status FillTimestampBatch(const DataType* type, liborc::ColumnVectorBatch*
cbatch,
+ int64_t& arrowOffset, int64_t& orcOffset, int64_t
length,
+ Array* parray, std::vector<bool>* incomingMask) {
+ auto array = checked_cast<TimestampArray*>(parray);
+ auto batch = checked_cast<liborc::TimestampVectorBatch*>(cbatch);
+ int64_t arrowLength = array->length();
+ if (!arrowLength) return Status::OK();
+ if (array->null_count() || incomingMask) batch->hasNulls = true;
+ for (; orcOffset < length && arrowOffset < arrowLength; orcOffset++,
arrowOffset++) {
+ if (array->IsNull(arrowOffset) || (incomingMask &&
!(*incomingMask)[orcOffset])) {
+ batch->notNull[orcOffset] = false;
+ } else {
+ int64_t data = array->Value(arrowOffset);
+ batch->notNull[orcOffset] = true;
+ switch (std::static_pointer_cast<TimestampType>(array->type())->unit()) {
+ case TimeUnit::type::SECOND: {
+ batch->data[orcOffset] = data;
+ batch->nanoseconds[orcOffset] = 0;
+ break;
+ }
+ case TimeUnit::type::MILLI: {
+ batch->data[orcOffset] =
+ static_cast<int64_t>(std::floor(data / kOneSecondMillis));
+ batch->nanoseconds[orcOffset] =
+ (data - kOneSecondMillis * batch->data[orcOffset]) *
kOneMilliNanos;
+ break;
+ }
+ case TimeUnit::type::MICRO: {
+ batch->data[orcOffset] =
+ static_cast<int64_t>(std::floor(data / kOneSecondMicros));
+ batch->nanoseconds[orcOffset] =
+ (data - kOneSecondMicros * batch->data[orcOffset]) *
kOneMicroNanos;
+ break;
+ }
+ default: {
+ batch->data[orcOffset] =
+ static_cast<int64_t>(std::floor(data / kOneSecondNanos));
+ batch->nanoseconds[orcOffset] = data - kOneSecondNanos *
batch->data[orcOffset];
+ }
+ }
+ }
+ }
+ batch->numElements = orcOffset;
+ return Status::OK();
+}
+
+template <class array_type>
+Status FillStringBatch(const DataType* type, liborc::ColumnVectorBatch* cbatch,
Review comment:
Thanks! These functions are actually not that similar though.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
