Lightweight hash-table implementation written in C

SYNOPSIS:
    I really wanted to write my own hash table implementation for 
    a tool that I am writing, so I did. 

    Ive based the api primarily on the opensource ght_hash code,
    however I opted to go open address vs chained. Another difference
    is that I decided to keep it really simple and basic, whereas
    ght_hash is more flexible.

DESIGN:
    Open-address hash table which uses quadratic probing for collision
    resolution.

    hash_remove does not free buckets, instead it marks the buckets as
    empty so that they can be reused.

    Currently only uses fnv1a digests for hashing, however it works
    fairly well for my uses.(1)

FILES:
    hashtable.c:
    hashtable.h:
        Contains the hash table management code.

    digest.c:
    digest.h:
        Contains hashing functions for use by hashtable.

FUNCTIONS:
    hash_create:
        Create a new hash table object.

    hash_destroy:
        Free all memory allocated by the hash table.  (does not free
        user allocated data blocks).

    hash_insert:
        Insert a key/value pair into the table.

    hash_get:
        Return the value for the given key in the hash table.

    hash_remove:
        Mark an element as removed from the table and return a pointer
        to the data in the table.

    hash_first:
        Return the first element in the hash table.

    hash_next:
        Return next element in a table.

    hash_dump:
        Print out which index is filled.

    fnv1a_digest:
        Fnv1a hash implementation. 

TODO:
    * Add hash_rehash function to automagically resize the table.
    * Write libcheck test suite for the hash-table. 
    * Compare performance against ght_hash.
    * Compare performance against khash.
    * Compare performance against oml_hash.

NOTES:
    1. fnv1 digest works best where table size is a power of 2. 

LICENSE:
    Use it for whatever you want. No warranty expressed or implied..