哈希和取模字符串在C＃和Python之间具有等效的结果

Question

I have need to group strings into ten different groups in a deterministic fashion, with some level of uniformity. The strings are identifiers that come from different sources, all with different (basically unknown) formats.

To accomplish this I decided to hash the strings and mod by 10. However I am going to be doing this in two different locations and I need their results to be consistent, one is a C# app and the other is a python one.

To ensure consistent hashing I have decided to go with MD5 (reasonably fast and consistent). Python already has this in the hashlib library and C# has one as in Cryptography

However I need to int and modulo these numbers with consistency. In python this is easy

md5 = hashlib.md5()
md5.update(my_string)
int(md5.hexdigest(), 16) % 10

But I can't just do this in C# as I only have 64 bit integers. So my thought is to just grab the last 16 characters from the hex. In python

int(md5.hexdigest()[-16:]) % 10

Then in C#

// hashString filled via MD5 code in the C# link above
string subHash = hashString.Substring(hashString.Length - 16);
Convert.ToUInt64(subHash, 16) % 10;

Now my questions are these. Are these two methods guaranteed to be equivalent? Is MD5 a good choice here? It's certainly consistent but if there is something faster that would be ideal. Is grabbing the last 16 characters the best way to prevent overflow?

Answer 1

The answer to this question Where can I find source or algorithm of Python's hash() function? includes the source code for the Python hash function (in C). Couldn't you implement that in C#. I'm guessing it would be much faster than MD5.

Python hash function for strings:

static long string_hash(PyStringObject *a)
{
    register Py_ssize_t len;
    register unsigned char *p;
    register long x;

    if (a->ob_shash != -1)
        return a->ob_shash;
    len = Py_SIZE(a);
    p = (unsigned char *) a->ob_sval;
    x = *p << 7;
    while (--len >= 0)
        x = (1000003*x) ^ *p++;
    x ^= Py_SIZE(a);
    if (x == -1)
        x = -2;
    a->ob_shash = x;
    return x;
}