views:

31

answers:

1

In Perl, we have IO::ScalarArray for treating the elements of an array like the lines of a file. In BioPerl, we have Bio::SeqIO, which can produce a filehandle that reads and writes Bio::Seq objects instead of strings representing lines of text. I would like to do a combination of the two: I would like to obtain a handle that reads successive Bio::Seq objects from an array of such objects. Is there any way to do this? Would it be trivial for me to implement a module that does this?

My reason for wanting this is that I would like to be able to write a subroutine that accepts either a Bio::SeqIO handle or an array of Bio::Seq objects, and I'd like to avoid writing separate loops based on what kind of input I get. Perhaps the following would be better than writing my own IO module?

sub process_sequences {
    my $input = $_[0];

    # read either from array of Bio::Seq or from Bio::SeqIO
    my $nextseq;
    if (ref $input eq 'ARRAY') {
        my $pos = 0
        $nextseq = sub { return $input->[$pos++] if $pos < @$input}; }
    }
    else {
        $nextseq = sub { $input->getline(); }
    }

    while (my $seq = $nextseq->()) {
        do_cool_stuff_with($seq)
    }
}
+1  A: 

Your solution looks like it should work. Unless you really want to spend a lot of time solving this problem, go with it until you don't like it anymore. I might have written that like so to avoid typing the variable name several times:

my $nextseq = do {
     if (ref $input eq ref [] ) {
         my $pos = 0;  #maybe a state variable if you have Perl 5.10
         sub { return $input->[$pos++] if $pos < @$input} }
         }
     else {
         sub { $input->getline() }
     }
 }

If you're interested in iterators, though, check out Mark Jason Dominus's Higher Order Perl, where he talks about all sorts of ways to do these sorts of things.

brian d foy
Thanks for leading me to HOP, which led me to [`Iterator::Simple`](http://search.cpan.org/~rintaro/Iterator-Simple-0.05/lib/Iterator/Simple.pm)
Ryan Thompson