From: Nikolai Weibull <mailing-lists.ruby-talk@...>
Date: 2005-07-22T06:21:18+09:00
Subject: Re: Rubys Regular Expression Engine

Wolfgang N�dasi-Donner wrote:
           ^-- Sorry, but your missing proper encoding information.

> Some weeks ago I sent a message here, "Onigurama - Problem with
> Subexpression Call", sent on Montag, 27. Juni 2005 23:22 -
> unfortunately there was no comment from someone. So I try to ask the
> major part again.

It's easier to track if you point to the number of the message instead.
It can be found at http://www.ruby-talk.org/.

> Onigurama does not recognize left recursion correctly.

Why should it?  (See below.)

> As I wrote in my earlier message
> 
> -----------------------------------------------------------
> orgstring= "5-((3+4)*5)+6+xx"
> pattern = /^(?<bal>[^()]*(\(\g<bal>\)[^()]*)*)$/

You make my eyes bleed...  Please don't use regular expressions for
matching parentheses, it simply can't be done using regular expressions
(well, .NET-"regular expressions" can, and perhaps there's some support
for it in Oniguruma as well, but definitely not in the way you're doing
it.).

Use a context free grammar for that instead.  You can use a tool like
racc to do that.

> puts "\nMuster: #{pattern.inspect}\nString: '#{orgstring}'"
> if (res = orgstring.match(pattern)) then
>   puts "Match:  '#{res}'"
> else
>   puts "+++++ No Match"
> end
> -----------------------------------------------------------
> 
> produces the Error Message
> 
> -----------------------------------------------------------
> never ending recursion: /^(?<bal>[^()]*(\(\g<bal>\)[^()]*)*)$/
> -----------------------------------------------------------
> 
> which is definitely wrong.

No, it's definitely right.  Your trying to match a <bal> inside a <bal>,
right?  That sounds like infinite recursion to me,
        nikolai

-- 
Nikolai Weibull: now available free of charge at http://bitwi.se/!
Born in Chicago, IL USA; currently residing in Gothenburg, Sweden.
main(){printf(&linux["\021%six\012\0"],(linux)["have"]+"fun"-97);}