Hattonuri opened a new issue, #38886: URL: https://github.com/apache/arrow/issues/38886
### Describe the enhancement requested When arrow parses decimal from big endian (and it happens on every element read from parquet for example), it does this in a very complex way https://github.com/apache/arrow/blob/eb5de184a7e5d02f98526332ace54250417bd232/cpp/src/arrow/util/decimal.cc#L745-L796 I assume that it's because FromBigEndian function uses ByteSwap which lacks fast version(those are in switch case) https://github.com/apache/arrow/blob/eb5de184a7e5d02f98526332ace54250417bd232/cpp/src/arrow/util/endian.h#L96-L119 But ByteSwap on 16 byte length can be written using two ByteSwaps of 8 length as here https://stackoverflow.com/questions/8004790/how-to-change-byte-order-of-128-bit-number ### Component(s) C++ -- 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]
