Written by omen23 -- David Schuster in 2012 (david [dot] schuster [at] kdemail [dot] net)
Open and free! (but please give me some credit =)

This is an interface to a sqlite3 database which stores its keys hashed or in plain
text. So it is designed for a login-procedure and was thought to be useful
for CGI login handling. The hashing capabilities can be turnt off in case you
are lazy enough to check your passwords or keys plain-text.

OK enough -- this is what this package provides:
login.c - login routine and sqlite3 callback (with an example main function)
hash.c - the hashing function which is used in the login program 
	 which can also be used on its own - demonstrated by the small "hashit" util
hashit.c - a small utility that can generate digests of all implemented gcrypt
	   hashing algorithms (e.g. for generation of your keys)
ftm.c - feature test macros
login.h - a header file where all the fun is defined :)

There is a Makefile for convenience - just type "make" and if you have a debian based
system with "libgcrypt-dev" and "libsqlite3-dev" installed everything should work OOB.
(and obviously make itself :P I guess this is called build-essentials)

The makefile provides a few targets:

"make" which is equivalent to "make debug" - builds everything 
(the main program is built with all debugging symbols)
"make pedantic" - showed a few warning on earlier gcc versions
all this is fixed now
"make feature" - to only generate the feature test program
"make hash" - to only generate the small hashing utility
"make release" - to generate the program with optimization on and strip from symbols
"make clean" - to delete all files that are (side)products of the compilation or
	       editing process


If you have problems setting this up - contact me! I cannot promise that I will answer
because my inbox is always really full. The string macros supplied use GNU extensions
or BSD versions of the standard implementation. (strlen, strcat, strcpy)

-Wno-pointer-sign -funsigned-char were used tho compile this with gcc before version
4.5 I guess - the code was fixed or the compiler or both.
(This code is under development since quite a while)