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

I have searched cpan and here and haven't found a good answer to this basic question. I have a string which has comma separated values, but there is nesting within parens. 
$str = 'sue,fred,x(mary,jane)';

[download]
I want to break this string into a three elements... 
['sue','fred','x(mary,jane)']

[download]

where, 'mary' and 'jane' are two args of function x which I am going to resolve by processing the third item later on

Not four elements... 
['sue','fred','x(mary','jane)']

[download]
I have tried several ideas like splitting and then recombining adjacent elements that should not have been split, but was hoping for a cleaner solution.

Replies are listed 'Best First'.
Re: csv with parens that include commas
by Anonymous Monk on Dec 05, 2010 at 00:48 UTC
    FWIW, that is bogus csv

    I would preprocess with 

    my $csv = Text::CSV->new({ qw!
  allow_loose_escapes 1
  allow_loose_quotes 1 ! });

...

s/x\(([^\(\)]+)\)/"$1"/g;
$csv->parse($_);

    [download]
    Seems to work 
    
$ perl -MText::CSV -e" $t = Text::CSV->new({qw! auto_diag 1 allow_loos
+e_escapes 1 allow_loose_quotes 1 ! }); $_ = q!sue,fred,x(mary,jane)!;
+ s/x\(([^\(\)]+)\)/\x22$1\x22/g; warn $_; warn $t->parse( $_ ); warn$
+_ for $t->fields"
sue,fred,"mary,jane" at -e line 1.
1 at -e line 1.
sue at -e line 1.
fred at -e line 1.
mary,jane at -e line 1.

    [download]
[reply]
[d/l]
[select]

      I want the third item to be the string

      x(mary,jane)

      i.e. mary and jane are two args of function x which I am going to resolve by processing the third item later on

      P.S. Although this may not meet the technical definition of "csv" it was useful in the title.

[reply]
[d/l]
        Yes, and? You want to treat something that is not CSV as CSV. Thats a no-go situation.
[reply]
        A reply falls below the community's threshold of quality. You may see it by logging in.
Re: csv'ish string with parens that include commas
by BrowserUk (Patriarch) on Dec 05, 2010 at 05:02 UTC 

    for(  
    'sue,fred,x(mary,jane)', 
    'x(a,b),c(d,e,f),g,h,i(j,k)', 
    'a,b,cdef(g),h,i(j)' 
) {
    print join' | ', m/( [^,(]+ \( [^)]+ \) | [^,]+ )(?:$|,)/gx;
};;

sue | fred | x(mary,jane)
x(a,b) | c(d,e,f) | g | h | i(j,k)
a | b | cdef(g) | h | i(j)

    [download]
    Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
    "Science is about questioning the status quo. Questioning authority".
    In the absence of evidence, opinion is indistinguishable from prejudice.
[reply]
[d/l]
      m/\G( [^,()]* \( [^()]* \) | [^,()]+ )(?:,|$)/g­x;

      [download]
[reply]
[d/l]

        Is that intended as an improvement? What purpose does the \G serve?

[reply]
Re: csv with parens that include commas
by Anonymous Monk on Dec 05, 2010 at 00:38 UTC
[reply]
Re: csv'ish string with parens that include commas
by james2vegas (Chaplain) on Dec 05, 2010 at 03:57 UTC
    Use Regexp::Common::list, something like this (presuming your strings and all word characters): 
    use Regexp::Common 'list';
my $re = qr{
        \w+  # word characters (word or function name)
        (?:\($RE{list}{-pat=>'\w+'}\))? # (optional parameters)
        }x;
my $x = 'joe,sam,x(monkey,lemur)';
my @x = ($x =~ m{$re}g);
print join("\n", @x);

    [download]
    returns: 
    joe
sam
x(monkey,lemur)

    [download]
[reply]
[d/l]
[select]

      Thanks. Actually, some of the args will be numbers, but I should be able to take it from here.

      I also looked at using Text::Balanced for a bit, but it seemed like using a hammer to drive a screw.

[reply]
[d/l]
        \w matches letters, digts and underscore, so that should be sufficient.
[reply]

Back to Seekers of Perl Wisdom