$_="eval eval"; s/a/i/;

The Reason

Several months ago, a co-worker asked me why someone would ever want to use multiple evals in a regular expression. At the time I had no good answer and even had difficulties producing a working example. Now (somewhat unfortunately), I can't seem to get those eee's out of my mind.

The Code

($_=join+qq e.e,map+ord,(join+qq ee,<DATA>)=~/[\d\.]/g)
&&s e.+e(36).qq+.+.(95).qq+.+.(32)eeeeeeeeeeeeeeeeeeeee

__DATA__
49.49.50.46.49.49.52.46.49.48.53.46.49.49.48.46.49.49.5
4.46.51.52.46.55.52.46.49.49.55.46.49.49.53.46.49.49.54
.46.51.50.46.57.55.46.49.49.48.46.49.49.49.46.49.49.54.
46.49.48.52.46.49.48.49.46.49.49.52.46.51.50.46.56.48.4
6.49.48.49.46.49.49.52.46.49.48.56.46.51.50.46.49.48.52
.46.57.55.46.57.57.46.49.48.55.46.49.48.49.46.49.49.52.
46.57.50.46.49.49.48.46.51.52.46.51.50.46.51.50.32.32.9
[download]

The Challange:

The above script ends with 21 e's, several of which are meaningless. Can you determine how many e's can be eliminated without effecting the output?

An Unrelated Question

Unlike most of my other obfus, this script should run under both strict and -w. What is the general consensus regarding this rule? Is it considered bad etiquette if obfuscated code does not run under these options?

Comment on $_="eval eval"; s/a/i/; Select or Download Code

Replies are listed 'Best First'.

Re: $_="eval eval"; s/a/i/;
by jdalbec (Deacon) on Aug 31, 2004 at 23:53 UTC

8 e's are required, so 13 e's may be omitted. I notice that 8+13=21 are all Fibonacci numbers. Is that deliberate?


Clear questions and runnable code get the best and fastest answer
	PerlMonks