Split the line up on every comma.
Grab the last M fields from the back.
Grab the first N fields from the front.
Join whatever is left using commas into a single field
Join the three pieces together with tabs.

Here is a quick example to get you up to speed.

#!/usr/bin/perl
use strict;

my $line = 'john,doe,manager,Pepsi Cola, inc.,Detroit,Michigan';
my @fields = split(/,/,$line);      # split on commas

my @last2  = splice(@fields,-2);    # Grab two fields off the back
my @first3 = splice(@fields, 0,3);  # Grab three fields from the front
my @middle = @fields;               # Whatever is left is the difficul
+t "company" data

my $middlefield = join(',',@middle);  # join the middle portion back t
+ogether

print "First:  $_\n" for (@first3);
print "Middle: $_\n" for (@middle);
print "Last:   $_\n" for (@last2);
print "\n";

my $newline = join("\t",@first3,$middlefield,@last2);  # generate new 
+line

print "Old: $line\n";
print "New: $newline\n";
[download]

-Blake

It took a few minutes, but this looks like a good prototype for your needs. You will likely need to adjust the tests on the position variable $i to fit your need more precisely...

use strict;
my $str = 'a,b,c,d,e,f,g,h,i,j,k,l,m,n';
print "$str\n";

my $i = 0;
$str =~ s/(.)/$i++;($1 ne ',' or $i>15 && $i<25) ? $1 : '#'/ge;

print "$str\n";
[download]

a,b,c,d,e,f,g,h,i,j,k,l,m,n
a#b#c#d#e#f#g#h,i,j,k,l,m#n
[download]

Note: along the way I tried using pos instead of $i, but apparently it is undefined while still in the right side of the s///. Update: Hofmator is correct to say I am wrong to imply that pos would work outside the s///.   pos is not set at all for s///.

Note 2: I offer this as a solution to the problem as you described it but it seems very odd that you would have a list of companies with names that are exactly 10 characters long.

re: note 2

The actual is number is 23 characters. I gave 10 in my example 'cause it's a nice round number. True, my sample data does not show this, that's why I said...

"Now the one thing that is consistent (my example doesn't show this) is that the company names will always start at the 15th character from the left and end at the 25th character from the left..."

If a company name has more than 23 characters, the name would get cut off (I have no control over the data that is given to me).

Thank you for your example. I think that it is exactly what I needed. I will also look at pos() as I am unfamiliar with it.

Note: along the way I tried using pos() instead of $i, but apparently it is undefined while still in the right side of the s///

well, this is right, but it is also undefined outside the s/// construct!! That's because (from perldoc)

pos returns the offset of where the last "m//g" search left off

and that's it. Pos is not set with s///.

-- Hofmator

#!/usr/local/bin/perl -w

use strict;

my @data = ('client-up-to-,Doe,Smith , Inc, jwaters,pfloyd,jjoplin');

foreach my $line (@data) {
    chomp $line;
    my $beg_line = substr($line, 0, 14);
    my $name_field = substr($line, 14, 10);
    my $end_line = substr($line, 25);
    print "B: ~$name_field~\n";
    $name_field =~ tr/,/ /;
    $name_field =~ s/\s*$//;
    print "A: ~$name_field~\n";
    $line = "${beg_line}${name_field}${end_line}";
    $line =~ tr/,/\t/;
    print "~$line~\n";
}

__END__
B: ~Doe,Smith ~ #note space is still in at this point
A: ~Doe Smith~
~client-up-to-    Doe Smith Inc     jwaters    pfloyd    jjoplin~
[download]

use strict;
use warnings;

my $max = 4;  #Largest expected index.

while (<DATA>) {
    s/#.*$//;  #Clear out the comments
    chomp;
    my @foo = split /,/;
    my $company = $foo[1]; #prime a variable with a company name
    if ($#foo > $max) {        #If there are more elements than the ma
+x...
        my $range = (scalar @foo) - $max; #Get the top of the range
        $company = join '+', @foo[1..$range]; #The company name is in 
+a slice
    }
    print "$company\n";
    #do whatever with the data.
}

__END__
client,ABC Company, jwilkens,jdoe,mdeart #this one is fine
client,Cornell,Thayil,Cameron & Sheppard, LLC, klivingston,amarison,ps
+erton #thi
s one is not
client,Doe,Smith & Randall, Inc, jwaters,pfloyd,jjoplin #this one is n
+ot
[download]

Ira,

"So... What do all these little arrows mean?"
~unknown

    substr($foo, 14, 10) =~ tr/,/+/;
    $foo =~ tr/,/\t/;
    substr($foo, 14, 10) =~ tr/+/,/;
[download]

One problem with this is that you may eventually have a company name with plusses in it. Besides that, it's three steps. I'm guessing company names won't have tabs in them, so the following works just as well and is only two steps:

    $foo =~ tr/,/\t/;
    substr($foo, 14, 10) =~ tr/\t/,/;
[download]

For your if statement take a look at substr and index.

$foo =~ s/\s/\\g;

For your replacements you should look at the tr/// operator.

-- Hofmator

The last question about the spaces was not related to the main problem. I just saw someone post it and thought that you could not substitute something with nothing. I will check out tr///.