Question

I want to implement a hashing technique in C where all the permutation of a string have same hash keys.
e.g. abc & cab both should have same keys.

I have thought of adding the ascii values & then checking frequency of characters[important otherwise both abc & aad would have same keys which we do not want].
But, it doesn’t seem to be much efficient.

Is there any better hashing function which resolves collisions well & also doesn’t result into sparse hash table?

Which hashing technique is used internally by Java [for strings] which not only minimizes the collisions but also the operations[insertion ,deletion, search] are fast enough?

Answer 1

The obvious technique is to simply sort the string. You could simply use the sorted string as the lookup key, or you can hash it with any algorithm deemed appropriate. Or you could use a run-length encoded (RLE) representation of your string (so the RLE of banana would be a3bn2), and optionally hash that.

A lot depends on what you’re going to do with the hashes, and how resistant they must be to collisions. A simple CRC (cylic redundancy checksum) might be adequate, or it might be that cryptographic checksums such as MD5 or SHA1 are not secure enough for you.