Short of moving to an XS based parser, stop making redundant copies of your data ($Record, and @row into a fresh arrayref).

while( <SRC> ) {
  chomp;
  my( $key, @row ) = split( /\|/, $_ );
  $prod{ $key } = \@row;
}
[download]

you can take this two steps further by getting rid of $key and using a map:

%prod=map{chomp;@a=split/\|/;shift@a,[@a]}<SRC>;
[download]

or, expanded,

%prod = map {
  chomp;
  @a = split /\|/;
  shift(@a) => [@a]
} <SRC>
[download]

You have to do [@a] instead of \@a if you do not localize the array with "my".

I think that's one step too far. Specifically, getting rid of a descriptive scalar in favor of lumping all the values together and using shift. I don't see any advantage in performance or clarity.

---

Caution: Contents may have been coded under pressure.

Could someone please explain the significance of shift@a,[@a] in the above map? I tried removing this statement and it did not work. I really dont know what that statement is doing.
Also I did not understand why [@a] is to be used if my is not used.
thanks
narashima

Thanks Fletch. But I thought there would something more perlish......

use Text::xSV;

my $parser = Text::xSV->new(
    fh => \*SRC,
    sep => '|',
);

my %prod;
while ( my @row = $parser->get_row ) {
    $prod{ shift @row } = \@row;
}
[download]

--traveler

I checked with my data. It did not make a big difference as my data is mostly a huge matrix of decimals.
Thanks,
narashima

I don't know about 'optimal', really. The code you list implies that you can rely on having 4-element rows, so this might be some fun:

while (<SRC>) {
   $prod{$1} = [$2,$3,$4] if m{(.+?)\|(.+?)\|(.+?)\|(.+)};
}
[download]

Or, another more general approach with regex captures, which works with any data size:

while (<SRC>) {
   @{ $prod{$1} } = split('\|', $2) if m/^(.+?)\|(.+)/;
}
[download]

Of course, I doubt those are very fast, compared to, say:

while (<SRC>) {
   my @row = split(/\|/, $_);
   $prod{$row[0]} = \@row[1..$#row]; #using a slice;
}
[download]

Just to reinforce dragonchild’s response – use Text::xSV. Rolling your own is not recommended.

Makeshifts last the longest.