MurmurHash3: a pure Julia implementation of the MurmurHash3 functions

Julia Version	Unit Tests	Coverage
Julia Latest

Note: this can be used to replace the C MurmurHash library used by base Julia, to implement a hash function that gives compatible results.

This provides the following functions (written in pure Julia): mmhash128 which hashes a UTF-8 (or ASCII, which is compatible) string mmhash128_c which can be used to hash an generic abstract string (by converting to UTF-8 on the fly, without having to allocate the string, working a few characters at a time) Note that the hashes on 64-bit systems are not the same as on 32-bit systems (this is true for base Julia hash as well)

mmhash32 which creates a 32 bit hash from a UTF-8/ASCII string (this works for String, as well as some of the Str types, such as ASCIIStr, UTF8Str, Binary, Text1Str can all be hashed directly) (note, there is no mmhash32_c yet, so other strings have to be converted to String type before hashing)

Julia uses the following code to create a hash using the C MurmurHash library:

const memhash = UInt === UInt64 ? :memhash_seed : :memhash32_seed
const memhash_seed = UInt === UInt64 ? 0x71e729fd56419c81 : 0x56419c81

function hash(s::String, h::UInt)
    h += memhash_seed
    ccall(memhash, UInt, (Ptr{UInt8}, Csize_t, UInt32), s, sizeof(s), h % UInt32) + h
end

similar code, such as that in JuliaString/StrBase.jl/src/hash.jl, implements the hash function for all of the Str types, but is specialized for performance based on the type of the string and whether or not it is aligned in memory.