Bio/SeqIO/MultiFile.pm

   1 #
   2 # BioPerl module for Bio::SeqIO::MultiFile
   3 #
   4 # Please direct questions and support issues to <bioperl-l@bioperl.org>
   5 #
   6 # Cared for by Ewan Birney <birney@ebi.ac.uk>
   7 #
   8 # Copyright Ewan Birney
   9 #
  10 # You may distribute this module under the same terms as perl itself
  11
  12 # POD documentation - main docs before the code
  13
  14 =head1 NAME
  15
  16 Bio::SeqIO::MultiFile - Treating a set of files as a single input stream
  17
  18 =head1 SYNOPSIS
  19
  20    $seqin = Bio::SeqIO::MultiFile( '-format' => 'Fasta',
  21                                    '-files'  => ['file1','file2'] );
  22    while((my $seq = $seqin->next_seq)) {
  23        # do something with $seq
  24    }
  25
  26 =head1 DESCRIPTION
  27
  28 Bio::SeqIO::MultiFile provides a simple way of bundling a whole
  29 set of identically formatted sequence input files as a single stream.
  30
  31 =head1 FEEDBACK
  32
  33 =head2 Mailing Lists
  34
  35 User feedback is an integral part of the evolution of this and other
  36 Bioperl modules. Send your comments and suggestions preferably to one
  37 of the Bioperl mailing lists.  Your participation is much appreciated.
  38
  39   bioperl-l@bioperl.org                  - General discussion
  40   http://bioperl.org/wiki/Mailing_lists  - About the mailing lists
  41
  42 =head2 Support
  43
  44 Please direct usage questions or support issues to the mailing list:
  45
  46 I<bioperl-l@bioperl.org>
  47
  48 rather than to the module maintainer directly. Many experienced and
  49 reponsive experts will be able look at the problem and quickly
  50 address it. Please include a thorough description of the problem
  51 with code and data examples if at all possible.
  52
  53 =head2 Reporting Bugs
  54
  55 Report bugs to the Bioperl bug tracking system to help us keep track
  56 the bugs and their resolution.
  57 Bug reports can be submitted via the web:
  58
  59   https://redmine.open-bio.org/projects/bioperl/
  60
  61 =head1 AUTHOR - Ewan Birney
  62
  63 Email birney@ebi.ac.uk
  64
  65 =head1 APPENDIX
  66
  67 The rest of the documentation details each of the object methods. Internal methods are usually preceded with a _
  68
  69 =cut
  70
  71
  72 # Let the code begin...
  73
  74
  75 package Bio::SeqIO::MultiFile;
  76 use strict;
  77
  78 use base qw(Bio::SeqIO);
  79
  80
  81 # _initialize is where the heavy stuff will happen when new is called
  82
  83 sub _initialize {
  84   my($self,@args) = @_;
  85
  86   $self->SUPER::_initialize(@args);
  87
  88   my ($file_array,$format) = $self->_rearrange([qw(
  89                                          FILES
  90                                          FORMAT
  91                                         )],
  92                                      @args,
  93                                      );
  94   if( !defined $file_array || ! ref $file_array ) {
  95       $self->throw("Must have an array files for MultiFile");
  96   }
  97
  98   if( !defined $format ) {
  99       $self->throw("Must have a format for MultiFile");
 100   }
 101
 102   $self->{'_file_array'} = [];
 103
 104   $self->_set_file(@$file_array);
 105   $self->_format($format);
 106   if( $self->_load_file() == 0 ) {
 107      $self->throw("Unable even to initialise the first file");
 108   }
 109 }
 110
 111 =head2 next_seq
 112
 113  Title   : next_seq
 114  Usage   :
 115  Function:
 116  Example :
 117  Returns :
 118  Args    :
 119
 120
 121 =cut
 122
 123 sub next_seq{
 124    my ($self,@args) = @_;
 125
 126    my $seq = $self->_current_seqio->next_seq();
 127    if( !defined $seq ) {
 128        if( $self->_load_file() == 0) {
 129            return;
 130        } else {
 131            return $self->next_seq();
 132        }
 133    } else {
 134        return $seq;
 135    }
 136
 137 }
 138
 139 =head2 next_primary_seq
 140
 141  Title   : next_primary_seq
 142  Usage   :
 143  Function:
 144  Example :
 145  Returns :
 146  Args    :
 147
 148
 149 =cut
 150
 151 sub next_primary_seq{
 152    my ($self,@args) = @_;
 153
 154    my $seq = $self->_current_seqio->next_primary_seq();
 155    if( !defined $seq ) {
 156        if( $self->_load_file() == 0) {
 157            return;
 158        } else {
 159            return $self->next_primary_seq();
 160        }
 161    } else {
 162        return $seq;
 163    }
 164
 165 }
 166
 167 =head2 _load_file
 168
 169  Title   : _load_file
 170  Usage   :
 171  Function:
 172  Example :
 173  Returns :
 174  Args    :
 175
 176
 177 =cut
 178
 179 sub _load_file{
 180    my ($self,@args) = @_;
 181
 182    my $file = shift(@{$self->{'_file_array'}});
 183    if( !defined $file ) {
 184        return 0;
 185    }
 186    my $seqio = Bio::SeqIO->new( '-format' => $self->_format(), -file => $file);
 187    # should throw an exception - but if not...
 188    if( !defined $seqio) {
 189        $self->throw("no seqio built for $file!");
 190    }
 191
 192    $self->_current_seqio($seqio);
 193    return 1;
 194 }
 195
 196 =head2 _set_file
 197
 198  Title   : _set_file
 199  Usage   :
 200  Function:
 201  Example :
 202  Returns :
 203  Args    :
 204
 205
 206 =cut
 207
 208 sub _set_file{
 209    my ($self,@files) = @_;
 210
 211    push(@{$self->{'_file_array'}},@files);
 212
 213 }
 214
 215 =head2 _current_seqio
 216
 217  Title   : _current_seqio
 218  Usage   : $obj->_current_seqio($newval)
 219  Function:
 220  Example :
 221  Returns : value of _current_seqio
 222  Args    : newvalue (optional)
 223
 224
 225 =cut
 226
 227 sub _current_seqio{
 228    my ($obj,$value) = @_;
 229    if( defined $value) {
 230       $obj->{'_current_seqio'} = $value;
 231     }
 232     return $obj->{'_current_seqio'};
 233
 234 }
 235
 236 =head2 _format
 237
 238  Title   : _format
 239  Usage   : $obj->_format($newval)
 240  Function:
 241  Example :
 242  Returns : value of _format
 243  Args    : newvalue (optional)
 244
 245
 246 =cut
 247
 248 sub _format{
 249    my ($obj,$value) = @_;
 250    if( defined $value) {
 251       $obj->{'_format'} = $value;
 252     }
 253     return $obj->{'_format'};
 254
 255 }
 256
 257 1;