Open viadea opened 1 year ago
Related cudf issue https://github.com/rapidsai/cudf/issues/8641
We need to be a bit careful here to be sure that our requirements match that of pandas/python. In Spark the output of all sha hashes is a string that is a lowercase HEX encoded representation of the binary hash result. This corresponds to the hexdigest()
method. Also https://github.com/rapidsai/cudf/issues/8641 calls out sha256 and sha512, not sha1. Spark supports sha1 and sha2. sha2 supports bit lengths of 224, 256, 384, and 512. sha2 matches sha256 and sha512.
The PR https://github.com/rapidsai/cudf/pull/9215 attached to https://github.com/rapidsai/cudf/issues/8641 appears to be adding support for SHA-1.
Depends on https://github.com/rapidsai/cudf/pull/14391
I wish we can support function sha1.
eg: