use strict;
use warnings;
use Data::Dumper;

my $in_fasta=$ARGV[0];
open(IN,$in_fasta) or die "cannot open $in_fasta";
my $out_fasta="$in_fasta.gi_header";
open(OUT,">",$out_fasta);
my $err_file="$in_fasta.headers_wNOgi";
open(ERR,">",$err_file);

my $header_count=0;
my $gi_count=0; my $protid_count=0;
my %giTogeneid;
while (my $line=<IN>){
                if ($line =~ /^>/ && $line =~ /^>.+GI="(\w+)"/){
                        my $gi=$1;
                        $gi_count ++;
                        my @header_columns=split(/\s+/,$line);
                        my $SM_geneid=$header_columns[0];
                        $SM_geneid=~s/^>//g;
                        $giTogeneid{$gi}=$SM_geneid;
                        print OUT ">gi$gi\n";
                }
                elsif ($line =~ /^>/ &&  $line !~ /^>.+GI="(\w+)"/ &&  $line !~ /^>.+protein_id="(\w+)"/){
                        print ERR "$line\n";
                }
                elsif ($line =~ /^>/ && $line =~ /^>.+protein_id="(\w+)"/){
                        my $protid=$1;
                        $protid_count++;
                        print OUT ">$protid\n";
                        my @header_columns=split(/\s+/,$line);
                        my $SM_geneid=$header_columns[0];
                        $SM_geneid=~s/^>//g;
                        $giTogeneid{$protid}=$SM_geneid;
                }

                elsif ($line !~ /^>/){
                        print OUT $line;
                }
}
close(IN);
print "number of headers seen\t$header_count\n";
print "number of gis seen\t$gi_count\n";
print "number of protids seen\t$protid_count\n";
close(OUT);