refactored assembly class to be parser agnostic

lib/bio-assembly.rb

@@ -1,169 +1,55 @@
-
-require 'bio/sequence'
+require 'bio/sequence' 
 require 'bio-assembly/contig'
 require 'bio-assembly/read'
 
 module Bio
 
-  class Assembly
-    attr_accessor :contigs
-
-    def initialize(path)
-      @file = File.new(path, 'r') 
-      @contigs = Array.new
-      parse_as
-    end
-
-    def contigs
-      # use each_contig to stream large files
-      parse_whole_file if @contigs.empty?
-      @contigs
-    end
-
-    def each_contig
-      # check if file is already parsed
-      if @total_num_contigs.to_i == @contigs.size
-        @contigs.each{ |contig| yield contig }
-      else
-        each_identifier do |identifier, attrs|
-          next unless identifier == 'CO'
-          contig = parse_contig(attrs)
-          @contigs.push contig
-          yield(contig)   
-        end
+   class Assembly
+      attr_accessor :contigs
+
+      @@formats = { }
+
+      def self.create(path, format)
+         streamer = @@formats[format]
+         if streamer
+            streamer.new(path)
+         else
+            raise "Format type '#{format}' is not supported"
+         end
       end
-    end
-
-    def to_ace
-      ace = ""
-      ace += "AS " + num_contigs.to_s + " " + num_reads.to_s + "\n\n"
-      each_contig { |contig| ace += contig.to_ace + "\n" }
-      ace
-    end
-
-    private
-
-    def parse_contig(attrs)
-      contig = Bio::Assembly::Contig.new
-      contig.name, base_num, @num_reads, base_segments_num, contig.orientation = attrs.split(" ")
-      # keep track of the number of RD identifiers parsed
-      @num_rds_parsed = 0
 
-      # get sequence
-      seq = @file.gets("\n\n").tr(" \r\n", "")
-      contig.seq = seq
-
-      # loop through identifiers (e.g AF, RD, etc)
-      each_identifier do |identifier, attrs|    
-        case identifier
-          when "BQ" then parse_bq(contig)
-          when "AF" then parse_af(contig, attrs)
-          when "BS" then parse_bs(contig, attrs)
-          when "RD" then parse_rd(contig, attrs); break if @num_rds_parsed == @num_reads.to_i
-          when "WR" then parse_wr(contig, attrs)
-          when "RT" then parse_rt(contig, attrs)
-          when "CT" then parse_ct(contig, attrs)
-          when "WA" then parse_wa(contig, attrs)
-        end
+      def self.register_parser name
+         @@formats[name] = self
       end
-
-     contig
-    end
-
-    # Finds the next_identifier
-    def each_identifier
-      @file.each do |line|
-        next if line !~ /^[ABCDQRW][ADFOQRST][\s\n].*/
-        yield(line[0..1], line[3..-1])
+
+      def contigs
+        # use each_contig to stream large files
+        parse_whole_file if @contigs.empty?
+        @contigs
       end
-    end
-
-    # parse assembly meta data
-    def parse_as
-      line = @file.gets
-      identifier, @total_num_contigs, total_num_reads = line.split(" ")
-    end
-
-    # parse contig sequence quality data
-    def parse_bq(contig)
-      contig.quality = @file.gets("\n\n").tr("\r\n", "").gsub(/^\s/, "").split(' ')
-    end
-
-    # parse read meta data
-    def parse_af(contig, attrs)
-      read = Bio::Assembly::Read.new
-      read.name , read.orientation, read.from = attrs.split(" ")
-      contig.add_read read
-    end
-
-    # parse base sequence data
-    def parse_bs(contig, attrs)
-      from, to, read_name = attrs.split(" ")
-      read = contig.find_read_by_name( read_name )
-      read.add_base_sequence(from, to, read_name)
-    end
-
-    # parse read sequence and position data
-    def parse_rd(contig, attrs)
-      # increment counter
-      @num_rds_parsed += 1
 
-      # parse read
-      read_name, num_padded_bases, num_read_infos, num_read_tags = attrs.split(" ") 
-      seq = @file.gets("\n\n").tr( " \r\n", "")
-
-      # get read with matching name
-      read = contig.find_read_by_name( read_name )
-      read.seq = seq
-      read.to = read.from.to_i + read.seq.length
-      # set read.to to contig length if read runs off contig
-      read.to = contig.seq.length if read.to > contig.seq.length
-
-      # if present parse QA and DS associated with this read
-      each_identifier do |identifier, attrs|
-        case identifier
-          when "QA" then parse_qa(read, attrs)
-          when "DS" then parse_ds(read, attrs); break
-        end
+      def each_contig
+         # implemented by each format subclass
+      end
+
+      private
+
+      def num_contigs
+        contigs.size
       end
-
-    end
-
-    # parse a read's clear ranges (the part of the read that contributes to the contig)
-    def parse_qa(read, attrs)
-      start, stop, clear_range_from, clear_range_to = attrs.split(" ")
-      read.clear_range_from = clear_range_from
-      read.clear_range_to = clear_range_to
-    end
-
-    # parse file data - ignored
-    def parse_ds(read, attrs)
-    end
-
-    # parse run meta data - ignored
-    def parse_wa(contig, attrs)
-    end
-
-    # parse run meta data - ignored
-    def parse_ct(contig, attrs)
-    end
-
-    def num_contigs
-      contigs.size
-    end
-
-    def num_reads
-      read_num = 0
-      each_contig { |contig| read_num += contig.num_reads }
-      read_num
-    end
-
-    def parse_whole_file
-      each_contig { |x| 1 }
-    end
-
-  end
 
-end
+      def num_reads
+        read_num = 0
+        each_contig { |contig| read_num += contig.num_reads }
+        read_num
+      end
 
+      def parse_whole_file
+        each_contig { |x| 1 }
+      end
+
+   end
+
+end
 
+require 'bio-assembly/ace'

lib/bio-assembly/ace.rb

@@ -0,0 +1,146 @@
+
+module Bio
+class Assembly
+
+class Ace < Bio::Assembly
+
+   # register parser with superclass
+   register_parser :ace
+
+   def initialize(path)
+      @file = File.new(path, 'r') 
+      @contigs = Array.new
+      parse_as
+    end
+
+   def each_contig
+     # check if file is already parsed
+     if @total_num_contigs.to_i == @contigs.size
+       @contigs.each{ |contig| yield contig }
+     else
+       each_identifier do |identifier, attrs|
+         next unless identifier == 'CO'
+         contig = parse_contig(attrs)
+         @contigs.push contig
+         yield(contig)   
+       end
+     end
+   end
+
+   def to_ace
+     ace = ""
+     ace += "AS " + num_contigs.to_s + " " + num_reads.to_s + "\n\n"
+     each_contig { |contig| ace += contig.to_ace + "\n" }
+     ace
+   end
+
+   private
+   def parse_contig(attrs)
+     contig = Bio::Assembly::Contig.new
+     contig.name, base_num, @num_reads, base_segments_num, contig.orientation = attrs.split(" ")
+     # keep track of the number of RD identifiers parsed
+     @num_rds_parsed = 0
+
+     # get sequence
+     seq = @file.gets("\n\n").tr(" \r\n", "")
+     contig.seq = seq
+
+     # loop through identifiers (e.g AF, RD, etc)
+     each_identifier do |identifier, attrs|    
+       case identifier
+         when "BQ" then parse_bq(contig)
+         when "AF" then parse_af(contig, attrs)
+         when "BS" then parse_bs(contig, attrs)
+         when "RD" then parse_rd(contig, attrs); break if @num_rds_parsed == @num_reads.to_i
+         when "WR" then parse_wr(contig, attrs)
+         when "RT" then parse_rt(contig, attrs)
+         when "CT" then parse_ct(contig, attrs)
+         when "WA" then parse_wa(contig, attrs)
+       end
+     end
+
+    contig
+   end
+
+   # Finds the next_identifier
+   def each_identifier
+     @file.each do |line|
+       next if line !~ /^[ABCDQRW][ADFOQRST][\s\n].*/
+       yield(line[0..1], line[3..-1])
+     end
+   end
+
+   # parse assembly meta data
+   def parse_as
+     line = @file.gets
+     identifier, @total_num_contigs, total_num_reads = line.split(" ")
+   end
+
+   # parse contig sequence quality data
+   def parse_bq(contig)
+     contig.quality = @file.gets("\n\n").tr("\r\n", "").gsub(/^\s/, "").split(' ')
+   end
+
+   # parse read meta data
+   def parse_af(contig, attrs)
+     read = Bio::Assembly::Read.new
+     read.name , read.orientation, read.from = attrs.split(" ")
+     contig.add_read read
+   end
+
+   # parse base sequence data
+   def parse_bs(contig, attrs)
+     from, to, read_name = attrs.split(" ")
+     read = contig.find_read_by_name( read_name )
+     read.add_base_sequence(from, to, read_name)
+   end
+
+   # parse read sequence and position data
+   def parse_rd(contig, attrs)
+     # increment counter
+     @num_rds_parsed += 1
+
+     # parse read
+     read_name, num_padded_bases, num_read_infos, num_read_tags = attrs.split(" ") 
+     seq = @file.gets("\n\n").tr( " \r\n", "")
+
+     # get read with matching name
+     read = contig.find_read_by_name( read_name )
+     read.seq = seq
+     read.to = read.from.to_i + read.seq.length
+     # set read.to to contig length if read runs off contig
+     read.to = contig.seq.length if read.to > contig.seq.length
+
+     # if present parse QA and DS associated with this read
+     each_identifier do |identifier, attrs|
+       case identifier
+         when "QA" then parse_qa(read, attrs)
+         when "DS" then parse_ds(read, attrs); break
+       end
+     end
+
+   end
+
+   # parse a read's clear ranges (the part of the read that contributes to the contig)
+   def parse_qa(read, attrs)
+     start, stop, clear_range_from, clear_range_to = attrs.split(" ")
+     read.clear_range_from = clear_range_from
+     read.clear_range_to = clear_range_to
+   end
+
+   # parse file data - ignored
+   def parse_ds(read, attrs)
+   end
+
+   # parse run meta data - ignored
+   def parse_wa(contig, attrs)
+   end
+
+   # parse run meta data - ignored
+   def parse_ct(contig, attrs)
+   end
+
+end # => end class Ace
+
+end # => end module Assembly
+end # => end module Bio