yaqi-zhao commented on code in PR #14585:
URL: https://github.com/apache/arrow/pull/14585#discussion_r1048177840
##########
cpp/src/arrow/util/bit_stream_utils.h:
##########
@@ -398,6 +412,27 @@ inline int BitReader::GetBatch(int num_bits, T* v, int
batch_size) {
return batch_size;
}
+#ifdef ARROW_WITH_QPL
+inline bool BitReader::GetBatchWithQpl(int batch_size, qpl_job* job) {
+ if (!job) {
+ return false;
+ }
+ job->param_low = value_offset_;
+ job->param_high = batch_size + value_offset_;
+ job->num_input_elements = batch_size + value_offset_;
+
+ job->next_in_ptr = const_cast<uint8_t*>(buffer_ - 1);
Review Comment:
> Oh, does it mean that we can use this feature only for Parquet's RLE
format? Could you tell me IAA's document for this?
Hi, Kou. You can download IAA specification from
https://www.intel.com/content/www/us/en/content-details/721858/intel-in-memory-analytics-accelerator-architecture-specification.html?wapkw=In-memory%20accelerator
And this PR is only for Parquet's RLE format.
> Or is "bit-width" included in general RLE format?
Yes, it's a general format, you can read this format from
https://parquet.apache.org/docs/file-format/data-pages/encodings/#a-namerlearun-length-encoding--bit-packing-hybrid-rle--3
--
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]
