cormanaz has asked for the wisdom of the Perl Monks concerning the following question:

Howdy Bros. I am trying to use Text::Balanced to extract some balanced parens, brackts, etc. from a piece of text. In this example:
use Text::Balanced qw( extract_bracketed );
my $text = q/%%% rick ran errands . sem(1,    [     word(1001,rick),  
+   word(1002,ran),     word(1003,errands),     word(1004,'.')    ],  
+  [     pos(1001,'NN'),     pos(1002,'VBD'),     pos(1003,'NNS'),    
+ pos(1004,'.')    ] )./;
my ($extracted,$remainder) = extract_bracketed($text,'()');

[download]
$extracted comes out null, and $remainder contains the entire string. I thought I would get the contents of the outermost set of parens. What am I doing wrong?

TIA

Steve

Replies are listed 'Best First'.
Re: Text::Balanced not extracting
by kcott (Archbishop) on Oct 16, 2010 at 20:58 UTC

    You need to specify a prefix.

    Change

    extract_bracketed($text,'()')

    to

    extract_bracketed($text,'()', '%%% rick ran errands . sem')

    and you'll get your expected output.

    This is explained in the Text::Balanced documentation.

    -- Ken

[reply]
[d/l]
[select]
      Yeah I saw that but wasn't too sure what they meant by a prefix. I wonder why they built the module that way. Seems like there would be cases you'd want to use something like this were the "prefix" wouldn't be known in advance.
[reply]

        I haven't used this module before today so I can't offer any deep insight into the ways things have been set up.

        I did wonder (exactly as you have) why it's set up like this.

        I've just successfully tried the following:

        my ($extracted,$remainder) = extract_bracketed($text,'()', ($text =~  m{ \A ( [^(]+ ) }msx)[0]);

        You might choose to use something like that in all cases that might have an unknown prefix.

        I haven't checked out the solution below (by Khen1950fx) - that may be preferable.

        -- Ken

[reply]
[d/l]
Re: Text::Balanced not extracting
by Khen1950fx (Canon) on Oct 16, 2010 at 22:31 UTC
    Another way: 
    #!/usr/bin/perl

use strict;
use warnings;
use Text::Balanced qw(:ALL);
use Data::Dumper::Concise;

my $text = <DATA>;
my ($extracted, $remainder) = extract_multiple( 
           $text, 
           [ \&extract_quotelike ]);
print Dumper($extracted, $remainder);

__DATA__
q/%%% rick ran errands . sem(
[   word(1001, 'rick'),    word(1002, 'ran'),    word(1003,  'errands'
+),    word(1004,  '.')  ] ,
[   pos(1001, 'NN') , pos(1002, 'VBD') ,  pos(1003, 'NNS'),  pos(1004,
+ '.')  ] )/;

    [download]
[reply]
[d/l]

Back to Seekers of Perl Wisdom