>>> READ THIS ALL! IF YOU DON'T WANT TO WASTE HOURS <<<

[0xdeadc0de]

It's been discussed in several other issue threads. But folks like me doesn't pay attention. Perhaps this will get your attention.

There's a bug in double hash calculation.

Pay attention to the issue #35

There's a case where hash_b will return the (bucket_size - 1) and then the calculation will ignore attemptNumber because of modulus operator.

(hash_a + (attempt * (hash_b + 1))) % num_buckets;

is basically:

(hash_a + (attempt * num_buckets) % num_buckets;

Anything multiplies with num_buckets will get to 0 because of modulus. Therefore it will stuck in infinite loop to get you a negative index.

Fix:

static int ht_hash(const char* s, const int num_buckets, const int attempt) {
    const int hash_a = ht_generic_hash(s, 151, num_buckets);
    int hash_b = ht_generic_hash(s, 163, num_buckets);
	if (hash_b == num_buckets - 1)
	{
		hash_b = 1;
	}
    return (hash_a + (attempt * (hash_b + 1))) % num_buckets;
}

Another fix (possibly more mathematically correct):
Reference: https://en.wikipedia.org/wiki/Double_hashing

We should map the domain of h2 to {1, num_buckets-1}.
One solution could be:

static int ht_hash(const char* s, const int num_buckets, const int attempt) {
    const int hash_a = ht_generic_hash(s, 151, num_buckets);
    const int hash_b = ht_generic_hash(s, 163, num_buckets-1);
    return (hash_a + (attempt * (hash_b + 1))) % num_buckets;
}