Question

I’m using Boost for C++ at the moment, and am trying to implement the unordered map (aka hash table) using CRC32. To my knowledge it will take a string as an initial key, hash it, and apply another operation so that it will fit into the number of buckets.

Though in my situation, I would like to hash the string key beforehand (using a separate CRC function in Boost), then use that ID to index the table. The problem that I need help with is that a CRC32 hash has 2^32 potential values, and I doubt I’ll ever need a table with 2^32 elements. What should I do in this situation?

Thanks for any help here!

Answer 1

Use the modulus operator — % in the C-based languages:

int hashtableIndex = hashValue % hashtableSize;

But note that the sign of the result, in C++, is “implementation defined”, and can be negative if hashValue is negative. So one may want to turn off the sign bit in hashValue before doing the % operation.

Also note that, if hashtableSize is known to be a power of two, one can simply mask hashValue to get the index:

int hashtableIndex = hashValue & (hashtableSize - 1);