There is a standard solution to this problem, mostly from Mastering Regular Expressions:

# You need to match a double quoted string with the following regex
# [^"\\]*(\\.[^"\\]*)*",?
#
# But to get the text between double quotes use some ( )
# ([^"\\]*(\\.[^"\\]*)*)",?
# gets text inside quotes as $1
#
# but you also have non quoted fields, thus
# ([^,]+),?
# which should match things optionally followed by a comma
#
# and then a match for separation commas
# ,
#
# this must be repeated with m/.../g
[download]

Before attempting this yourself, take at look at Text::ParseWords and the quotewords routine. This should solve your problem. If the module is not available to you then the following untested code from Mastering Regular Expressions should work:

@fields = ();
while ($text =~ m/"([^"\\]*(\\.[^"\\]*)*)",?|([^,]+),?|,/g {
push (@fields, defined ($1) ? $1 : $3) ;
}
push (@fields, undef) if $text =~ m/,$/; # Account for the special cas
+e of an empty last field.

# all data is now in @fields
[download]

Note: untested.

It seems to me that you're trying to parse a comma-delimited entries. Each entry may or may not be quoted. If quoted, the entry may contain commas. Sounds like a good job for Text::xSV or similar.

Keep in mind that most languages use escape routines to indicate embedded delimiters. Perl, for example, would require that the string in question be something like "a like <A HREF=\"foo\"> b</A>" in order to avoid bareword errors on foo.

2,5,"foobar","tape 2" white", 0122435992020<br>
                    ^
[download]

Of course if you don't have data like that, you'll be fine.

Best of Luck

Cheers

Sweetblood

Here's an example straight from the Text::CSV POD:

use Text::CSV;

$csv = Text::CSV->new();              # create a new object

$status = $csv->parse($line);         # parse a CSV string into fields
@columns = $csv->fields();            # get the parsed fields
[download]

Don't use a regex to do what can be done more reliably with a tried and proven module. Text::CSV is the right tool for the job. You'll find it even properly handles commas embeded in quoted strings.

Better is Text::xSV. In addition to allowing any character as the separator (instead of hard-coding comma), it actually takes into account embedded newlines, which Text::CSV doesn't do. Plus, it's a pure-perl solution that's almost as fast as Text::CSV_XS. And, tilly wrote it. :-)

------
We are the carpenters and bricklayers of the Information Age.

Then there are Damian modules.... *sigh* ... that's not about being less-lazy -- that's about being on some really good drugs -- you know, there is no spoon. - flyingmoose

I shouldn't have to say this, but any code, unless otherwise stated, is untested

Great advice. ...a new favorite module. :)



Dave

my $data = q("one"," one,two",3,"a like <A HREF="foo"> b</A>");

my @pieces = split(q(,), $data);
[download]

#!/usr/bin/perl -w
use strict;
use Dumpvalue;

my $dumper = new Dumpvalue;
my @split_elms = map{chomp;
    my $line = $_;
    my @data = map{my $substr = $_;
        $substr =~ s/,/:innerc:/g;
        $line =~ s/\Q$_\E/$substr/;} $line =~ /("[^"]*")/g;
    @data = map{$_ =~ s/:innerc:/,/g;
        $_} split(/,/,$line);
    \@data;} <DATA>;
$dumper->dumpValue(\@split_elms);
exit;

__DATA__
"one"," one,two",3,"a like <A HREF="foo"> b</A>"
0,0,0,0
[download]

0  ARRAY(0x81a89dc)
   0  '"one"'
   1  '" one,two"'
   2  3
   3  '"a like <A HREF="foo"> b</A>"'
1  ARRAY(0x8151afc)
   0  0
   1  0
   2  0
   3  0
[download]