I would suggest something like
Text::CSV for this but that isn't always an option...
You're fields all seem to follow a rather basic pattern, so I would suggest that to make your life easier, you simplify how you are dealing with this... at a possible minor cost to efficiency you can just parse through the fields one by one and use a much simpler and MUCH easier to maintain regexp, like so
#!/usr/local/bin/perl
use warnings;
use strict;
my $line = "2006-01-01,Kims,common,406,560(centrifuge,refrig.),569b,60
+7(dark room),210-211,101(ultracentrifuge),104-105(crystal growth room
+s),660(centrifuge,refrig.)";
my @fields;
push @fields, $1 while $line =~ /([^,(]+(?:\([^)]*\))?)/g;
my $i = 1;
print join(', ', map { $i++.": $_" } @fields),"\n";
Which outputs
1: 2006-01-01, 2: Kims, 3: common, 4: 406, 5: 560(centrifuge,refrig.),
+ 6: 569b, 7: 607(dark room), 8: 210-211, 9: 101(ultracentrifuge), 10:
+ 104-105(crystal growth rooms), 11: 660(centrifuge,refrig.)
Your original regexp is working just fine, but you only have 4 sets of capturing parens so you only get 4 fields...
I would strongly suggest using the
x modifier in big regexps like that to improve readability and also creating variables holding regexp pieces which match any fields that you can re-use the regexp for, so you only have to define a segment once.. should also add readability.
- Ant
- Some of my
best work - (1 2 3)