I made a simple implementation of the md5 algorithm based on the original RFC 1321 from 1992!
It was quite interesting to see the original code for the reference implementation, as it is using the old-style K&R syntax. For instance, the types for function arguments are defined separately:
/* Decodes input (unsigned char) into output (UINT4). Assumes len is
a multiple of 4.
*/
static void Decode (output, input, len)
UINT4 *output;
unsigned char *input;
unsigned int len;
{
I tried to stay as close as possible to the RFC, without considerations for speed. The resulting program is around 15% slower than the md5sum implementation from coreutils, which is not that bad.
Even if I spent quite some time on a test harness (testtools), including a python script to generate the test data, it is not production-ready: the error-handling system is very primitive and has not been tested thoroughly. It was only a weekend project to have a bit of fun learning about this venerable algorithm.