lib/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    my $seqin = Bio::SeqIO::MultiFile->new( -format => 'Fasta',
  21                                            -files  => ['file1','file2'] );
  22    while (my $seq = $seqin->next_seq) {
  23        # do something with $seq
  24    }
  25
  26 =head1 DESCRIPTION
  27
  28 Bio::SeqIO::MultiFile provides a simple way of bundling a whole
  29 set of identically formatted sequence input files as a single stream.
  30 File format is automatically determined by C<Bio::SeqIO>.
  31
  32 =head1 FEEDBACK
  33
  34 =head2 Mailing Lists
  35
  36 User feedback is an integral part of the evolution of this and other
  37 Bioperl modules. Send your comments and suggestions preferably to one
  38 of the Bioperl mailing lists.  Your participation is much appreciated.
  39
  40   bioperl-l@bioperl.org                  - General discussion
  41   http://bioperl.org/wiki/Mailing_lists  - About the mailing lists
  42
  43 =head2 Support
  44
  45 Please direct usage questions or support issues to the mailing list:
  46
  47 I<bioperl-l@bioperl.org>
  48
  49 rather than to the module maintainer directly. Many experienced and
  50 reponsive experts will be able look at the problem and quickly
  51 address it. Please include a thorough description of the problem
  52 with code and data examples if at all possible.
  53
  54 =head2 Reporting Bugs
  55
  56 Report bugs to the Bioperl bug tracking system to help us keep track
  57 the bugs and their resolution.
  58 Bug reports can be submitted via the web:
  59
  60   https://github.com/bioperl/bioperl-live/issues
  61
  62 =head1 AUTHOR - Ewan Birney
  63
  64 Email birney@ebi.ac.uk
  65
  66 =head1 APPENDIX
  67
  68 The rest of the documentation details each of the object methods. Internal methods are usually preceded with a _
  69
  70 =cut
  71
  72
  73 # Let the code begin...
  74
  75
  76 package Bio::SeqIO::MultiFile;
  77
  78 use strict;
  79
  80 use base qw(Bio::SeqIO);
  81
  82
  83 # _initialize is where the heavy stuff will happen when new is called
  84
  85 sub _initialize {
  86     my($self, @args) = @_;
  87
  88     $self->SUPER::_initialize(@args);
  89
  90     my ($file_array, $format) = $self->_rearrange([qw(FILES FORMAT)], @args);
  91     if( !defined $file_array || ! ref $file_array ) {
  92         $self->throw("Must have an array files for MultiFile");
  93     }
  94
  95     $self->{'_file_array'} = [];
  96     $self->_set_file(@$file_array);
  97
  98     $self->format($format) if defined $format;
  99
 100     if( $self->_load_file() == 0 ) {
 101         $self->throw("Unable to initialise the first file");
 102     }
 103 }
 104
 105
 106 =head2 next_seq
 107
 108  Title   : next_seq
 109  Usage   :
 110  Function:
 111  Example :
 112  Returns :
 113  Args    :
 114
 115 =cut
 116
 117 sub next_seq{
 118     my ($self, @args) = @_;
 119     my $seq = $self->_current_seqio->next_seq();
 120     if( !defined $seq ) {
 121         if( $self->_load_file() == 0) {
 122             return;
 123         } else {
 124             return $self->next_seq();
 125         }
 126     } else {
 127         return $seq;
 128     }
 129 }
 130
 131
 132 =head2 next_primary_seq
 133
 134  Title   : next_primary_seq
 135  Usage   :
 136  Function:
 137  Example :
 138  Returns :
 139  Args    :
 140
 141 =cut
 142
 143 sub next_primary_seq{
 144     my ($self, @args) = @_;
 145     my $seq = $self->_current_seqio->next_primary_seq();
 146     if( !defined $seq ) {
 147         if( $self->_load_file() == 0) {
 148             return;
 149         } else {
 150             return $self->next_primary_seq();
 151         }
 152     } else {
 153         return $seq;
 154     }
 155 }
 156
 157
 158 =head2 _load_file
 159
 160  Title   : _load_file
 161  Usage   :
 162  Function:
 163  Example :
 164  Returns :
 165  Args    :
 166
 167 =cut
 168
 169 sub _load_file{
 170     my ($self, @args) = @_;
 171     my $file = shift @{$self->{'_file_array'}};
 172     if( !defined $file ) {
 173         return 0;
 174     }
 175     my $seqio;
 176     my $format = $self->format;
 177     if ($format) {
 178         $seqio = Bio::SeqIO->new( -file => $file, -format => $format );
 179     } else {
 180         $seqio = Bio::SeqIO->new( -file => $file );
 181         $self->format($seqio->format) if not $format;
 182     }
 183
 184     # should throw an exception - but if not...
 185     if( !defined $seqio) {
 186         $self->throw("Could not build SeqIO object for $file!");
 187     }
 188     $self->_current_seqio($seqio);
 189     return 1;
 190 }
 191
 192
 193 =head2 _set_file
 194
 195  Title   : _set_file
 196  Usage   :
 197  Function:
 198  Example :
 199  Returns :
 200  Args    :
 201
 202 =cut
 203
 204 sub _set_file{
 205     my ($self, @files) = @_;
 206     push @{$self->{'_file_array'}}, @files;
 207 }
 208
 209
 210 =head2 _current_seqio
 211
 212  Title   : _current_seqio
 213  Usage   : $obj->_current_seqio($newval)
 214  Function:
 215  Example :
 216  Returns : value of _current_seqio
 217  Args    : newvalue (optional)
 218
 219 =cut
 220
 221 sub _current_seqio{
 222     my ($obj, $value) = @_;
 223     if( defined $value) {
 224         $obj->{'_current_seqio'} = $value;
 225     }
 226     return $obj->{'_current_seqio'};
 227 }
 228
 229
 230 # We overload the format() method of Bio::Root::IO by a simple get/set
 231
 232 sub format{
 233     my ($obj, $value) = @_;
 234     if( defined $value) {
 235         $obj->{'_format'} = $value;
 236     }
 237     return $obj->{'_format'};
 238 }
 239
 240
 241 1;