Beefy Boxes and Bandwidth Generously Provided by pair Networks
laziness, impatience, and hubris
 
PerlMonks  

Re: A demanding parser

by Juerd (Abbot)
on Jan 25, 2002 at 16:43 UTC ( [id://141484]=note: print w/replies, xml ) Need Help??


in reply to A demanding parser

Is there any way of avoiding the external module and catch an arbitrary number of nested parentheses with a "normal" Regex? I know that the Owl book says it can't be done, but I would like to put my mind at rest on this issue.

It depends on what you call a "normal" regex. If normal means without any perl specific things, it can not be done. But if you don't mind a perl specific regex, perlre has the solution:

The following pattern matches a parenthesized group: 

$re = qr{
          \(
          (?:
            (?> [^()]+ )    # Non-parens without backtracking
           |
            (??{ $re })     # Group with matching parens
          )*
          \)
        }x;

[download]

2;0 juerd@ouranos:~$ perl -e'undef christmas'
Segmentation fault
2;139 juerd@ouranos:~$

[download]

Replies are listed 'Best First'.
Re: Re: A demanding parser
by TheDamian (Vicar) on Jan 25, 2002 at 17:12 UTC
    Is there any way of avoiding the external module and catch an arbitrary number of nested parentheses with a "normal" Regex? 
    use Regexp::Common;

$str =~ /$RE{balanced}{-parens=>'()'}/

    [download]
[reply]
[d/l]
      That's what he does already. He asked if it were possible without the module. (Not a very good idea, but I don't know the motivation (education, perhaps)) 

      2;0 juerd@ouranos:~$ perl -e'undef christmas'
Segmentation fault
2;139 juerd@ouranos:~$

      [download]

[reply]
[d/l]
        Ah, sorry. Missed that bit.

        How about: 

            use Regexp::Common;

    print $RE{balanced}{-parens=>'()'}, "\n";
    print $RE{balanced}{-parens=>'{}'}, "\n";
    print $RE{balanced}{-parens=>'{}()'}, "\n";
    # etc.

        [download]
        followed by reading the Regexp::Common source?

        Damian

[reply]
[d/l]
Re: Re: A demanding parser
by gmax (Abbot) on Jan 26, 2002 at 18:56 UTC
    Thanks for the tip. I am not sure I understand how to use it, though.
    My purpose, as you have pointed out, is to replace Regexp::Common with some normal Perl RegEx. By normal I mean a non-module dependant expression.
    As for the motivation, you guessed right that it's related to education. Personally, I wouldn't bother. I need to distribute this module as part of a more extensive educational material aiming at the build-up of a huge database. I would like to avoid pointing to a CPAN module, since many people in the audience are not experienced Perl users. They should just copy this module to their computers and execute the import/export script.
    Of course I can provide them with a copy of the module, or instruct them to connect to the CPAN, download the module and install it, or use "perl -MCPAN -e shell" but it would steal valuable time from my lectures.

    That aside, here is a test script for your RegEx, which does not seem to give me what I want.
    Was it my misunderstanding, or were you trying to show me how to catch the inner parenthesized text only?
    #!/usr/bin/perl -w
use strict;
use Regexp::Common;

my $re = qr{ \( (?: (?> [^()]+ ) | (??{ $re }))* \) }x;

my $input = "aa bb cc (dd ee (ff gg (hh) jj) kk)";

print "With module\n";
while ($input =~ m/(\w+|$RE{balanced}{-parens=>'()'})\s*/g) { 
    print "$1\n";
}

print "With recursive RegExp\n";
while ($input =~ m/(\w+|$re)\s*/g) { 
    print "$1\n";
}

__END__
# output: 
With module
aa
bb
cc
(dd ee (ff gg (hh) jj) kk)

With recursive RegExp
aa
bb
cc
dd
ee
ff
gg
(hh)
jj
kk

    [download]
    update
    Found the problem. Recursive RegExes don't work properly with use strict
    Changing
    my $re = qr{ \( (?: (?> [^()]+ ) | (??{ $re }))* \) }x;

    [download]
    into
    no strict 'vars';
$rec_re = qr{ \( (?: (?> [^()]+ ) | (??{ $rec_re }))* \) }x;
my $re = $rec_re;
use strict;

    [download]
    makes the same output from both regexes.
     _  _ _  _  
(_|| | |(_|><
 _|
[reply]
[d/l]
[select]
Re^2: A demanding parser
by Aristotle (Chancellor) on Jan 25, 2002 at 17:07 UTC
    Wait a second - is that really a recursive regex there? :-o

    Makeshifts last the longest.

[reply]
      Affirmative 

      2;0 juerd@ouranos:~$ perl -e'undef christmas'
Segmentation fault
2;139 juerd@ouranos:~$

      [download]

[reply]
[d/l]

In Section Meditations

Log In?
Username:
Password:

What's my password?
Create A New User
Domain Nodelet?

www.com | www.net | www.org
Node Status?
node history
Node Type: note [id://141484]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? • Last hour • Other CB clients
Other Users?
Others making s'mores by the fire in the courtyard of the Monastery: (5)
As of 2024-04-18 02:31 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    No recent polls found