From: Florian Gilcher <flo@...>
Date: 2008-03-01T04:44:51+09:00
Subject: Re: Need a regex searching html code

>>
>> ab
>> aabb
>> aaabbb
>> aaaabbbb
>> etc.
>
> "ab
> xx
> aabb
> aaabbb
> aaabb
> aaaabbbb".split.each{|s|
>  if s.match(/^(a+)/) and s.match(/^a+b{#{$1.size}}$/)
>    puts s
>  else
>    puts '-'
>  end
> }
>
> Or one can use regular expression + code:
>
> "ab
> xx
> aabb
> aaabbb
> aaabb
> aaaabbbb".split.each{|s|
>  if s.match(/^(a+)(b+)$/) and $1.size == $2.size
>    puts s
>  else
>    puts '-'
>  end
> }
>
> What makes anyone think that a single regular expression
> has to do all the work?
>
>


I don't know. But many think one fits. Thats why i wrote this  
explanation, as it is something i see almost everyday and to give some  
insight to those that are pondering on why this is so.
So: your solution does not fit the problem, but thanks for showing  
that another problem (parsing "a*nb*n" with a touring-complete  
language) can indeed be solved.

I also stated this in my last paragraph: you can solve the problem by  
using regular expressions. But the language of regular expressions by  
itself is not mighty enough to solve it alone.

Greetings
Florian