move strand into seqspan

[mussa.git] / alg / seq_span.hpp

diff --git a/alg/seq_span.hpp b/alg/seq_span.hpp
index 8d1991360df0bdcebee6e83cab7b926686630a9a..5a2da1268d3a62cf807f323d7e39bb450fea5841 100644 (file)
--- a/alg/seq_span.hpp
+++ b/alg/seq_span.hpp
@@ -35,14 +35,34 @@ public:
   typedef SeqString::size_type size_type;
   typedef SeqString::value_type value_type;
   static const size_type npos = SeqString::npos;
+  //! Define strand types
+  /**!
+   *  Unknown strand is treated as "either" strand
+   *  Plus refers to the initially created strand
+   *  Minus is the opposite strand
+   *  Both is for any feature that applies to "both" strands
+   *    (which may not actually be useful) 
+   *  Same strand is only used when creating a subsequence
+   *    and implies the subsequence has the same orientation as the parent
+   *  Opposite is only used for creating a subsequence
+   *    and implies the subsequence has the opposite orientation as the parent
+   *  Single indicates that this is single stranded and there can't be
+   *    an opposite strand.
+   */ 
+  enum strand_type { UnknownStrand, MinusStrand, PlusStrand, 
+                     BothStrand, SameStrand, OppositeStrand, SingleStrand };
 
 public:
   SeqSpan(const SeqSpan &);
   SeqSpan(const SeqSpan *);
   explicit SeqSpan(const std::string &, 
-               AlphabetRef = reduced_nucleic_alphabet
+               AlphabetRef a = dna_alphabet,
+               strand_type strand=PlusStrand
            );
-  SeqSpan(const SeqSpanRef, size_type start=0, size_type count=npos);
+  SeqSpan(const SeqSpanRef, 
+          size_type start=0, 
+          size_type count=npos, 
+          strand_type strand=SameStrand);
 
   //! assignment
   SeqSpan& operator=(const SeqSpan&);
@@ -87,6 +107,7 @@ public:
   size_type stop() const { return seq_start + seq_count; }
   //! set one past the last position relative to the root sequence.
   void setStop(size_type);
+  strand_type strand() const { return seq_strand; }
   
   //! get start position relative to the parent sequence
   size_type parentStart() const;
@@ -100,12 +121,17 @@ public:
   
 
   //! return a subsequence, copying over any appropriate annotation
-  SeqSpanRef subseq(size_type start=0, size_type count = std::string::npos);
+  SeqSpanRef subseq(size_type start=0, 
+                    size_type count = std::string::npos, 
+                    strand_type = PlusStrand);
   //! get sequence
   std::string sequence() const; 
   //! are both sequences derived from the same sequence tree?
   static bool isFamily(const SeqSpan& a, const SeqSpan& b);
   
+  //! fill in our rc_seq variable 
+  void initialize_rc_seq() const;
+  
   friend class Sequence;
 private:
   //! do not statically initialize, only create with new
@@ -117,11 +143,15 @@ protected:
   size_type seq_start;
   //! how big we ware
   size_type seq_count;  
-  // Do I need to track the strand here?
+  //! strand orientation
+  strand_type seq_strand;
   
   //! keep a reference to who our parent span is 
   SeqSpanRef parent;
 
+  //! hold a reverse complement version of our sequence if needed
+  SeqStringRef rc_seq;
+  
   // boost::serialization support
   friend class boost::serialization::access;
   template<class Archive>
@@ -129,6 +159,7 @@ protected:
     ar & BOOST_SERIALIZATION_NVP(seq);
     ar & BOOST_SERIALIZATION_NVP(seq_start);
     ar & BOOST_SERIALIZATION_NVP(seq_count);
+    ar & BOOST_SERIALIZATION_NVP(seq_strand);
     ar & BOOST_SERIALIZATION_NVP(parent);
   }
 };