Problems with delete in hash perl

Question

Another hash contains name as key (split by _) and has 4 parts(I am comparing the first three parts). The value is a string whose length i need to compare and If the name is same and already exists in the hash (the first three parts), I need to keep the string with longest length and if the lengths are equal, I don't care which one it keeps. I know with iteration it is creating issues as the key-value pairs are being deleted.

 my %temphash=();
 %temphash=%anotherhash;
 foreach my $tempkey(keys %temphash){
        my @tempkey_splitted = split /\_/, $tempkey;
        my $tempkey_newfamily = $tempkey_splitted[0];
        my $tempkey_newgenera = $tempkey_splitted[1];
        my $tempkey_newspecies = $tempkey_splitted[2];
        my $tempkey_catstring ="$tempkey_newfamily"."_$tempkey_newgenera"."_$tempkey_newspecies";
        my $sequence_realkey="";
        my $sequence_tempkey="";
        my $length_realkey="";
        my $length_tempkey="";
        if ($realkey_catstring eq $tempkey_catstring){
            $sequence_realkey = $anotherhash{$realkey};
            $length_realkey = length($sequence_realkey);
            #print "$anotherhash{$realkey}";
            #print "$length_realkey";
            $sequence_tempkey = $temphash{$tempkey};
            #print "$anotherhash{$tempkey}";
            $length_tempkey = length($sequence_tempkey);
            # print "$length_tempkey";

            if($length_realkey>$length_tempkey){
                delete($temphash{$tempkey});
               #print ">$realkey\n$anotherhash{$realkey}\n\n";
                }
            elsif($length_tempkey>$length_realkey){
               delete($temphash{$realkey});
               #print ">$tempkey\n$anotherhash{$tempkey}\n\n";
                }
            elsif($length_tempkey eq $length_realkey){
                delete($temphash{$realkey});
                }
            }
        else{
             print "do nothing"; 
            }
        }
    }

Answer 1

Your code is quite convoluted, but I think the example code below does what you need. It works by copying the contents of %anotherhash (please don't use such dreadful meaningless names for your variables) to %temphash one element at a time, and only if the new value from %anotherhash is longer than the %temphash

my %temphash;

while (my ($key, $val) = each %anotherhash) {

  my @key = split /_/, $key;
  my $tempkey = join '_', @key[0,1,2];
  my $tempval = $temphash{$tempkey};

  unless (defined $tempval and length $tempval >= length $val) {
    $temphash{$tempkey} = $val;
  }
}