Your script:

open IN, "infile.dat";
open OUT, ">data.csv";

while (<IN>) {
    chomp $_;
    my @fields = split(/|/, $_);
    my $newLine = join(",", @fields);
    print OUT "$newLine\n"
}
close IN;
close OUT;
[download]

would corrupt the data with:

hi|I'm with GNU,Inc.|I need a comma in my title
this|will|produce|,|to|many|fields
[download]

Using pipes to delimit data is a better idea (usually) because it is less common in data. There are also several modules to parse CSV. Incuding Text::CSV.

UPDATE: tilly has pointed out that Text::CSV_XS is a better solution than Text::CSV. I handles imbedded line endings in the data.

grep

grep> cd pub 
grep> more beer