added q_id and s_id options to findsim.rb

git-svn-id: http://biopieces.googlecode.com/svn/trunk@1795 74ccb610-7750-0410-82ae-013aeee3265d

diff --git a/code_ruby/lib/maasha/findsim.rb b/code_ruby/lib/maasha/findsim.rb
index 5a14065f5115e4a00660790a885d43bbc77b4085..cd07b8e4c3fa45ed5f8bd9233627df1e21221f2f 100644 (file)
--- a/code_ruby/lib/maasha/findsim.rb
+++ b/code_ruby/lib/maasha/findsim.rb
@@ -56,6 +56,8 @@ class FindSim
     @q_total_ary  = nil
     @result       = nil
     @result_count = 0
+    @q_ids        = []
+    @s_ids        = []
   end
 
   # Method to load sequences from a query file in FASTA format
@@ -69,6 +71,8 @@ class FindSim
 
     Fasta.open(file, 'r') do |ios|
       ios.each do |entry|
+        @q_ids << entry.seq_name if @opt_hash[:query_ids]
+
         oligos = str_to_oligo_rb_ary_c(entry.seq, @opt_hash[:kmer], 1).uniq.sort
 
         q_total << oligos.size
@@ -99,6 +103,8 @@ class FindSim
 
     Fasta.open(file, 'r') do |ios|
       ios.each_with_index do |entry, s_index|
+        @s_ids << entry.seq_name if @opt_hash[:subject_ids]
+
         zero_ary_c(oligo_ary,  (4 ** @opt_hash[:kmer]) * BYTES_IN_INT)
         zero_ary_c(shared_ary, @q_size                 * BYTES_IN_INT)
 
@@ -119,7 +125,7 @@ class FindSim
   end
 
   # Method that for each query index yields all hits, sorted according to
-  # decending score, as a list of Score objects.
+  # decending score, as Hit objects.
   def each
     sort_hits_c(@result, @result_count)
 
@@ -128,19 +134,23 @@ class FindSim
     hit_index = 0
 
     (0 ... @q_size).each do |q_index|
-      scores = []
       zero_ary_c(hit_ary, HIT_ARY_MAX * BYTES_IN_HIT)
       hit_ary_size = get_hits_c(@result, @result_count, hit_index, hit_ary, q_index)
 
-      max = (hit_ary_size > @opt_hash[:report_scores]) ? @opt_hash[:report_scores] : hit_ary_size
+      if @opt_hash[:report_scores]
+        max = (hit_ary_size > @opt_hash[:report_scores]) ? @opt_hash[:report_scores] : hit_ary_size
+      else
+        max = hit_ary_size
+      end
 
       (0 ... max).each do |i|
         q_index, s_index, score = hit_ary[BYTES_IN_HIT * i ... BYTES_IN_HIT * i + BYTES_IN_HIT].unpack("IIF")
 
-        scores << Score.new(score, s_index)
-      end
+        q_id = @opt_hash[:query_ids]   ? @q_ids[q_index] : q_index
+        s_id = @opt_hash[:subject_ids] ? @s_ids[s_index] : s_index
 
-      yield scores
+        yield Hit.new(q_id, s_id, score)
+      end
 
       hit_index += hit_ary_size
     end
@@ -366,6 +376,7 @@ class FindSim
     # Method that given a string (char array) encodes all kmers overlapping
     # with a given step size as integers that are pushed onto a Ruby array
     # which is returned.
+    # TODO should have an option for skipping oligos with ambiguity codes.
     builder.c %{
       VALUE str_to_oligo_rb_ary_c(
         VALUE _str,    // DNA or RNA string.
@@ -554,19 +565,20 @@ class FindSim
   # >>>>>>>>>>>>>>> Embedded classes <<<<<<<<<<<<<<<
 
   # Class for holding score information.
-  class Score
-    attr_reader :score, :s_index
-
-    # Method to initialize Score object with
-    # a subject sequence index and a score.
-    def initialize(score, s_index)
-      @s_index  = s_index
-      @score    = score
+  class Hit
+    attr_reader :q_id, :s_id, :score
+
+    # Method to initialize Hit object with
+    # query and subject id a score.
+    def initialize(q_id, s_id, score)
+      @q_id  = q_id
+      @s_id  = s_id
+      @score = score
     end
 
     # Method for outputting score objects.
     def to_s
-      "#{@s_index}:#{@score.round(2)}"
+      "#{@q_id}:#{@s_id}:#{@score.round(2)}"
     end
   end
 end