|
Peter Cawley wrote:
The code is accessing (upto) 32 bytes in total, not every 32nd byte. e.g. for a 1MB string, just 32 bytes are visited for the hash rather than 32768 bytes (1024*1024/32).
And what happens if there are two 1 MB strings that differ by 1 byte (at arbitrary position)?
-- Shmuel