Implementation for 5 > x > 4 / RSpec fuzz testing

Hi,

before going to sleep I had an idea for how to implement 5 > x > 4.
It’s pretty simple an implementation. I remember reading on ruby-talk
years ago that this was impossible to do without changing Ruby’s
parser.

So, nothing big. I just don’t want this to decay on my hard disk. It
demonstrates a fair bit of meta programming and that Ruby is powerful
as well.

It’s available at http://flgr.0x42.net/code/comp_chain.rb

I’m also wondering if somebody has already thought about how to best
do fuzz testing in RSpec? I know about RFuzz, but it seems to be web
specific. Is there any official best practise for using random input
in your specs?

Kind regards,
Florian G.

Florian Groß wrote:

So, nothing big. I just don’t want this to decay on my hard disk. It
demonstrates a fair bit of meta programming and that Ruby is powerful
as well.

It’s available at http://flgr.0x42.net/code/comp_chain.rb
Ha! Coming from the guy who wrote Binding.of_caller, I was expecting a
challenge. That’s ridiculously simple! Cool beans.

I’m also wondering if somebody has already thought about how to best
do fuzz testing in RSpec?
The only thing that comes to mind is a presentation Jim W. did on a
prototype Contract Driven Development library.
http://video.google.com/videoplay?docid=-3883668083321892812

Devin