rok commented on code in PR #14191:
URL: https://github.com/apache/arrow/pull/14191#discussion_r1025274541
##########
cpp/src/parquet/encoding.cc:
##########
@@ -2060,6 +2062,220 @@ class DictByteArrayDecoderImpl : public
DictDecoderImpl<ByteArrayType>,
}
};
+// ----------------------------------------------------------------------
+// DeltaBitPackEncoder
+
+constexpr uint32_t kValuesPerBlock = 128;
+constexpr uint32_t kMiniBlocksPerBlock = 4;
+
+template <typename DType>
+class DeltaBitPackEncoder : public EncoderImpl, virtual public
TypedEncoder<DType> {
+ public:
+ using T = typename DType::c_type;
+ using TypedEncoder<DType>::Put;
+
+ explicit DeltaBitPackEncoder(const ColumnDescriptor* descr, MemoryPool* pool,
+ const uint32_t values_per_block =
kValuesPerBlock,
+ const uint32_t mini_blocks_per_block =
kMiniBlocksPerBlock)
+ : EncoderImpl(descr, Encoding::DELTA_BINARY_PACKED, pool),
+ values_per_block_(values_per_block),
+ mini_blocks_per_block_(mini_blocks_per_block),
+ values_per_mini_block_(values_per_block / mini_blocks_per_block),
+ deltas_(values_per_block, ::arrow::stl::allocator<T>(pool)),
+ bits_buffer_(AllocateBuffer(pool, (values_per_block + 3) * sizeof(T))),
Review Comment:
I'm thinking of this as a upper bound of how long can the header be in bits.
I'll add a comment.
--
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]
