genome_prepare/script/get_longseq.pl

#!/usr/bin/perl -w
use strict;
use List::Util qw(min max);

die "Usage:\n\tperl $0 cds.fa/pep.fa > uniq_id.cdsfa\n" if(@ARGV!=1);
my %gene;

open IN ,$ARGV[0]||die $!;

local $/=">";
<IN>;
while(<IN>){
	chomp;
	my @a = split/\n/,$_;
	my $id = shift @a;
	my $seq = join("\n",@a);
	my $len = length(join("",@a));
	$gene{$id}{$len}{$seq}=1;
}

local $/="\n";
close IN;

foreach(sort keys %gene){
	my $max = max keys %{$gene{$_}};
	my @a = keys %{$gene{$_}{$max}};
	my $seq = shift @a;
	print ">$_\n$seq\n";
}