Fastest Hash Algorithm for Text Data

Question

I'm trying to choose a hash algorithm for comparing about max 20 different text data.

Which hash is better for these requirements?

• Less CPU Consumption
• Small footprint (<=32 bytes)
• Collision is not a big deal
• Can be generated from .NET Framework 2 (shouldn't be a 3rd party library)

I'm using hash for less memory footprint and comparison performance

Solution 1

If collision is not a big deal you can take the first letter of each document. Or you can use the length of the text or the string with the text.

Solution 2

Paul Hsieh has a decent, simple, fast, 32-bit SuperFastHash that performs better than most existing hash functions, is easier to understand/implement, and sounds like it meets your criteria.

Solution 3

The FNV hash is a well-known fast hashing algorithm. It is not cryptographically secure, but it sounds like you don't need a secure hash.

Solution 4

A very quick check would be to take the length of a text and XOR it with the first 4 bytes of it and use that as a hash. If this is good enough it is extremely fast because independent of the number of bytes of the file.

View more solutions