Bio/SeqIO/raw.pm

   1 #-----------------------------------------------------------------------------
   2 # PACKAGE : Bio::SeqIO::raw
   3 # AUTHOR  : Ewan Birney <birney@ebi.ac.uk>
   4 # CREATED : Feb 16 1999
   5 #
   6 # Copyright (c) 1997-9 bioperl, Ewan Birney. All Rights Reserved.
   7 #           This module is free software; you can redistribute it and/or
   8 #           modify it under the same terms as Perl itself.
   9 #
  10 # _History_
  11 #
  12 # Ewan Birney <birney@ebi.ac.uk> developed the SeqIO
  13 # schema and the first prototype modules.
  14 #
  15 # This code is based on his Bio::SeqIO::Fasta module with
  16 # the necessary minor tweaks necessary to get it to read
  17 # and write raw formatted sequences made by
  18 # chris dagdigian <dag@sonsorol.org>
  19 #
  20 # October 18, 1999  Largely rewritten by Lincoln Stein
  21 #
  22 # Copyright Ewan Birney
  23 #
  24 # You may distribute this module under the same terms as perl itself
  25
  26 # POD documentation - main docs before the code
  27
  28 =head1 NAME
  29
  30 Bio::SeqIO::raw - raw sequence file input/output stream
  31
  32 =head1 SYNOPSIS
  33
  34 Do not use this module directly.  Use it via the L<Bio::SeqIO> class.
  35
  36 =head1 DESCRIPTION
  37
  38 This object can transform Bio::Seq objects to and from raw flat
  39 file databases.
  40
  41 =head1 FEEDBACK
  42
  43 =head2 Mailing Lists
  44
  45 User feedback is an integral part of the evolution of this and other
  46 Bioperl modules. Send your comments and suggestions preferably to one
  47 of the Bioperl mailing lists.  Your participation is much appreciated.
  48
  49   bioperl-l@bioperl.org                  - General discussion
  50   http://bioperl.org/wiki/Mailing_lists  - About the mailing lists
  51
  52 =head2 Support
  53
  54 Please direct usage questions or support issues to the mailing list:
  55
  56 I<bioperl-l@bioperl.org>
  57
  58 rather than to the module maintainer directly. Many experienced and
  59 reponsive experts will be able look at the problem and quickly
  60 address it. Please include a thorough description of the problem
  61 with code and data examples if at all possible.
  62
  63 =head2 Reporting Bugs
  64
  65 Report bugs to the Bioperl bug tracking system to help us keep track
  66 the bugs and their resolution.
  67 Bug reports can be submitted via the web:
  68
  69   https://github.com/bioperl/bioperl-live/issues
  70
  71 =head1 AUTHORS
  72
  73   Ewan Birney   E<lt>birney@ebi.ac.ukE<gt>
  74   Lincoln Stein E<lt>lstein@cshl.orgE<gt>
  75
  76 =head1 CONTRIBUTORS
  77
  78   Jason Stajich E<lt>jason@bioperl.org<gt>
  79
  80 =head1 APPENDIX
  81
  82 The rest of the documentation details each of the object methods.
  83 Internal methods are usually preceded with a _
  84
  85 =cut
  86
  87
  88 # Let the code begin...
  89
  90 package Bio::SeqIO::raw;
  91 use strict;
  92
  93 use Bio::Seq::SeqFactory;
  94
  95 use base qw(Bio::SeqIO);
  96
  97 our %variant = ( 'multiple' => undef, # default
  98                  'single'   => undef );
  99
 100 sub _initialize {
 101     my($self,@args) = @_;
 102     $self->SUPER::_initialize(@args);
 103     my ($variant) = $self->_rearrange([qw(VARIANT)], @args);
 104     $variant ||= 'multiple';
 105     $self->variant($variant);
 106     $self->{record_separator} = $variant eq 'single' ? undef : $/;
 107     if( ! defined $self->sequence_factory ) {
 108         $self->sequence_factory(Bio::Seq::SeqFactory->new
 109                     (-verbose => $self->verbose(),
 110                      -type => 'Bio::Seq'));
 111     }
 112     $self->variant;
 113 }
 114
 115 =head2 next_seq
 116
 117  Title   : next_seq
 118  Usage   : $seq = $stream->next_seq()
 119  Function: returns the next sequence in the stream
 120  Returns : Bio::Seq object
 121  Args    :
 122
 123
 124 =cut
 125
 126 sub next_seq{
 127     my ($self,@args) = @_;
 128     ## When its 1 sequence per line with no formatting at all,
 129     ## grabbing it should be easy :)
 130
 131     ## adding an option to assume the file is one sequence
 132     local $/ = $self->{record_separator};
 133     my $nextline = $self->_readline();
 134     return unless defined $nextline;
 135
 136     my $sequence = uc($nextline);
 137     $sequence =~ s/\W//g;
 138     return unless $sequence;
 139     return $self->sequence_factory->create(-seq => $sequence) if $sequence;
 140 }
 141
 142 =head2 write_seq
 143
 144  Title   : write_seq
 145  Usage   : $stream->write_seq($seq)
 146  Function: writes the $seq object into the stream
 147  Returns : 1 for success and 0 for error
 148  Args    : Array of Bio::PrimarySeqI objects
 149
 150
 151 =cut
 152
 153 sub write_seq {
 154     my ($self,@seq) = @_;
 155     foreach my $seq (@seq) {
 156         $self->throw("Must provide a valid Bio::PrimarySeqI object")
 157             unless defined $seq && ref($seq) && $seq->isa('Bio::PrimarySeqI');
 158         $self->_print($seq->seq, "\n") or return;
 159     }
 160     $self->flush if $self->_flush_on_write && defined $self->_fh;
 161     return 1;
 162 }
 163
 164 =head2 write_qual
 165
 166  Title   : write_qual
 167  Usage   : $stream->write_qual($seq)
 168  Function: writes the $seq object into the stream
 169  Returns : 1 for success and 0 for error
 170  Args    : Bio::Seq::Quality object
 171
 172
 173 =cut
 174
 175 sub write_qual {
 176     my ($self,@seq) = @_;
 177     my @qual = ();
 178     foreach (@seq) {
 179         unless ($_->isa("Bio::Seq::Quality")){
 180            warn("You cannot write raw qualities without supplying a Bio::Seq::".
 181                 "Quality object! You passed a ".ref($_)."\n");
 182            next;
 183         }
 184         @qual = @{$_->qual};
 185         if(scalar(@qual) == 0) {
 186            $qual[0] = "\n";
 187         }
 188         $self->_print (join " ", @qual,"\n") or return;
 189     }
 190     return 1;
 191 }
 192
 193 =head2 variant
 194
 195  Title   : variant
 196  Usage   : $format  = $obj->variant();
 197  Function: Get and set method for the sequence variant.  For raw sequence, this
 198            indicates whether to treat the input as multiple sequences (the
 199            default) or as a single sequence.
 200
 201            Current values accepted are:
 202
 203             'single'    single sequence
 204             'multiple'  multiple sequences (default)
 205  Returns : string
 206  Args    : new value, string
 207
 208 =cut
 209
 210 # variant() method inherited from Bio::Root::IO
 211
 212 # private method for testing record separator
 213
 214 sub _separator {
 215     shift->{record_separator};
 216 }
 217
 218 1;