random generates series of unique random integers in an arbitrary range.
More specifically it means that it generates an ordered set of unique integers.
It guarantees that no number occurs more than once.
It makes use of the GMP Library (The GNU Multiple Precision Arithmetic Library) for generating integers of any imaginable order, given the limitations on memory, processing power, the shell's ability to provide long enough arguments, the user's ability to type them et cetera.
By default the program generates one number between 1 and 100.
It can take a number of options to control the size and range of the generated set:
-l, --lower=<number> Lower bound (inclusive). Default is 1.
-u, --upper=<number> Upper bound (inclusive). Default is 100.
-c, --count=<number> Generate this many unique numbers.
The lower and upper bounds can accept negative numbers (although the lower bound must always be lower than the upper).
There is a limit on the size of the set (currently set to UINT_MAX from your
system's /usr/include/limits.h) which is primarily designed for simplicity
of implementation. Generating very large sets of random numbers is pretty
silly anyway, and takes a lot of time and memory. If you want to, you
can try to generate a set of 20 000 000 numbers, but don't come
crying and say I didn't warn you when your memory has run out.
There's nothing stopping you from shooting yourself in the foot. And there shouldn't be.
You may find that you want to generate extremely big random numbers, and that the argument list becomes too long for the system to handle. In that case you can read values from a file with the -f option. The file shall contain only two values, separated by a space: the lower and the upper bound. Again, these may be negative values. You can also read from stdin, of course.
This was written mostly for learning the GNU MP library and for the fun of it. Hopefully you will find this software useful and a maybe even slightly fun.
I used it's predecessor for generating image indexes in a shell script on a web server once. (It could only generate measly standard ints though).
And if you make a Lotto row with the help of this program I wish you the best of luck trying to win.
This program was first written in 2005. Updates regarding safe memory usage and updates to the code was done in 2017.
This software may contain bugs. If you find one, please notify me. Also, if you have suggestions, improvements or comments, I'd like to hear about it.
Also, the software lacks test cases and QA, so you might not want to rely on this particular piece of software to drive trains, planes or automobiles.
The good people writing and maintaining the GMP library.
The GNU people and Donald Knuth for the binary tree implementation used herein.
Mikael Magnusson (Mikachu) and Erik Waling (ddl) for suggestions and improvements (generate excluded numbers when the size of the set is more than half the range (Mikachu); reading parameters from a file (Mikachu); reading a random seed (ddl) ).
Pär Karlsson, September 2017