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

Hi. I am trying to get up to speed on ParseRecDescent. The FAQ offers idiom (credited to Merlyn) to match CSV-like data 
 CSVLine: QuotedText(s Comma) { do something}

[download]
I'm trying to write a rule to match a term. A term is one or more words, separated by underscores. Neither terms nor words can contain whitespace. These are valid terms: apple, apple_pear, apple_plum_grape. These are not valid terms: _, _apple, apple_, apple__pear. Can anyone give me guidance on writing the rule for term and word? Do I need a <skip> directive to indicate a term cannot contain whitespace? Here's what I have, which does not work. Many thanks! 
my $g = q(

    Word: /^[A-Z]+$/i
    {return $item[1]}

    TermSep: '_'

    Term: Word(s TermSep) { use Data::Dumper; Dumper($item[1]) }

    Line: Term
    {$item[1]}

  );

[download]
rkg

Replies are listed 'Best First'.
Re: ParseRecDescent and csv-like data
by Elian (Parson) on Aug 21, 2003 at 13:45 UTC
    You don't have to get that fancy. Something like: 
    word: /[a-zA-Z]+(_[a-zA-Z]+)*/

    [download]
    should be sufficient.
[reply]
[d/l]
Re: ParseRecDescent and csv-like data
by dreadpiratepeter (Priest) on Aug 21, 2003 at 13:47 UTC
    Wouldn't: 
    
Term: /[A-Z]|[A-Z][A-Z_]*[A-Z]/i {return \split(/_/,$item[1])}

    [download]
    Give you what you want? I could be wrong, it's off the top of my head, but it should return a term as a list of it's parts (which seems to be what you want).

    -pete
    "Worry is like a rocking chair. It gives you something to do, but it doesn't get you anywhere."
[reply]
[d/l]
Re: ParseRecDescent and csv-like data
by Abigail-II (Bishop) on Aug 21, 2003 at 14:51 UTC
    Instead of using ParseRecDescent, may I suggest a Regexp::Common solution? Parsing lists is one of its options. 
    #!/usr/bin/perl

use strict;
use warnings;

use Regexp::Common;

my $re = $RE{list}{-sep => '_'}{-pat => '(?:(?!_)\w)+'};

while (<DATA>) {
    chomp;
    print "'$_' ", /^$re$/ ? "matched\n" : "did not match\n";
}


__DATA__
apple_pear
apple_plum_grape
_
apple
_apple
apple_
apple__pear




'apple_pear' matched
'apple_plum_grape' matched
'_' did not match
'apple' did not match
'_apple' did not match
'apple_' did not match
'apple__pear' did not match

    [download]

    Abigail

[reply]
[d/l]
Re: ParseRecDescent and csv-like data
by gjb (Vicar) on Aug 21, 2003 at 14:18 UTC

    Although I agree that it might be better to consider 'apple_pear' as one token rather than two (as suggested by Elian and dreadpiratepeter above, the code below should do what you want: 

    #!/usr/bin/perl

use strict;
use warnings;

use Parse::RecDescent;

my $text = <<EOI
apple_pear
cherry
munchy_nice_apricot
mint_
_banana
raspberry__pie
EOI

$Parse::RecDescent::skip = '[ \t]*';

my $grammar = q(

  { use Data::Dumper }

  data: line(s)
  line: term endofline
  term: word(s /_/) ...endofline
       { print Dumper(\%item); }
  word: /[a-z]+/
  endofline: /[\n\r]+/

);

my $parser = Parse::RecDescent->new($grammar);
if ($parser->data($text)) {
    print "ok\n";
}

    [download]

    Hope this helps, -gjb-

    Update: changed the code to suite rkg's requirement that '_apple', 'apple__juice' and 'apple_' should not be accepted. The grammar happens to be more elegant now and I learned about separator patterns.

    Update 2: removed an unused token from the grammar.

[reply]
[d/l]
      Thanks for your help. You code, I think, accepts apple__pear (two underscores), _apple (leading underscore), and pear_ (trailing) as valid... in my desired grammar, they shouldn't be. Any thoughts? rkg ps And yes, I like your approach of treating a term as multiple tokens, vs. one. This is a part of a larger grammar, and I need the flexibility of ParseRecDescent.
[reply]
Re: ParseRecDescent and csv-like data
by rkg (Hermit) on Aug 21, 2003 at 17:15 UTC
    Well, this ugly setup (below) does what I want.

    If anyone would like to help me clean it, I'd appreciate learning how! 

    Word1: /^[A-Z0-9\/:"'.-]+$/i {$item[1]}
Word2: /[A-Z0-9\/:"'.-]+/i   {$item[1]}
Term: Word1 | (<skip: ''> Word2 ('_' Word2)(s))

    [download]
    (Why am I going to all this hassle, you may ask? Words are nested into terms, terms into superterms, and so on, each with a unique delimiter. So the simple regexp 1 token doesn't quite work. rkg
[reply]
[d/l]

Back to Seekers of Perl Wisdom