CSequence_Mol2_1.h

/* fast_TIGER (version 1.0), a program for computing TIGER rates
 * (TIGER: Tree Independent Generation of Evolutionary Rates).
 *
 * Copyright (C) January 2015 by Paul Frandsen and Christoph Mayer
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.

 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License along
 * with this program; if not, see <http://www.gnu.org/licenses/>.
 *
 * For any enquiries send an email to
 * Paul Frandsen: paulbfrandsen@gmail.com
 * or
 * Christoph Mayer: c.mayer.zfmk@uni-bonn.de
 *
 * When publishing work that is based on the results of Fast_Tiger please cite:
 * 
 * Frandsen, P.B., Calcott, B., Mayer, C., Lanfear, R., 2015, Automatic selection of
 * partitioning schemes for phylogenetic analyses using iterative k-means clustering
 * of site rates, BMC Evolutionary Biology 15:13.
 *
 */
#ifndef CSEQUENCE_MOL_H
#define CSEQUENCE_MOL_H

// Changes:
// 22.08.2009: According to the fasta format specifications, the sequence name is only the string between the ">" and
//             the first space. The rest of the line contains a sequence description or annotation information.
//             This is now reflected in the CSequence_Mol class. The variables full_name, name, and description
//             contain this information.
//             The sequence name variables are now of type faststring (not string as before).
// 07.01.2011: Migrated to fastring2.h and CFile2_1.h
// 30.04.2011: Sliding window CG-AT analysis

// TODO: Data type detection does not fully implemented yet.

#include <vector>
#include <iostream>
#include "faststring2.h"
//#include "CFile/CFile3b.h"
#include "CFile/CFile2_1.h"
#include <cstdio>
#include "basic-DNA-RNA-AA-routines.h"
#include "fast-realloc-vector.h"
#include <cstring>
#include <cstdio>

class CSequence_Mol
{
  public:
  enum DataTypesEnum {
    standard = 1,	// characters with arbitrarily-assigned,
			// discrete states, such as discrete
			// morphological data

    dna,		/* DNA sequences (states A, C, G, T) */
    rna,		/* RNA sequences (states A, C, G, U) */
    //    nucleotide,		/* nucleotide sequences */
    protein,		/* amino acid sequences         */
    molecular,          /* AA, dna, or rna              */
    unknown,
    auto_detect_type,   // Not fully implemented
    mixed
  };

  enum processing_flag {
    convert_toupper                  = 1,
    convert_ambig2N                  = 2,
    autodetecttype                   = 4,
    removegaps                       = 8,
    removestars                      = 16,
    compute_numbers_of_residues_flag = 32,
    gaps_to_ambig                    = 64
  };

  private:
  faststring      full_name;   // with description after first space
  faststring      short_name;  // without desciption after first space
  faststring      description; // description only

  faststring      data;
  DataTypesEnum   type;
  char            general_ambig;

  bool            numbers_of_residues_computed;

  // General rule: If we convert symbols, we count the new symbols we have in memory,
  // not the symbols in the file. This is more consitent.
  //
  unsigned        number_of_AAs;           // Does not include Xs and ?
  unsigned        number_of_DNARNA_bases;  // Includes Us but not ambigs
  unsigned        number_of_DNARNA_amgibs; // Includes Ns and '?'
  unsigned        number_of_Us;
  unsigned        number_of_gaps;
  unsigned        number_of_stars;
  unsigned        number_of_questionmarks;

  unsigned        number_of_Xs;
  unsigned        number_of_Ns;
  unsigned        number_raw_original_length; //no spaces, but gaps and stars 
  unsigned        number_CG;                   // CG content in case of DNA/RNA


  fastvector<unsigned> gap_and_star_koords;
  fastvector<unsigned> gap_koords;
  fastvector<unsigned> star_koords;

  bool bitcode_recoded;

 private:
  void find_shortname_and_description()
  {
    full_name.removeSpacesBack();
    full_name.removeSpacesFront();

    unsigned pos_description = full_name.find(' ');
    unsigned full_name_len   = full_name.size();

    if (pos_description == -1u) // no space -> no sequence description
    {
      pos_description = full_name.size();
      short_name.assign(full_name.begin(), full_name.begin() + pos_description);
    }
    else
    {
      short_name.assign(full_name.begin(), full_name.begin() + pos_description);
      // Skip all spaces
      while (pos_description < full_name_len && full_name.get_unckecked(pos_description) == ' ')
      {
	//	std::cerr << "-------------DEBUG: ++pos_description" << std::endl;	
	++pos_description;
      }

      //      std::cerr << "-------------DEBUG: " << full_name.end() - (full_name.begin() + pos_description) << std::endl;
      //      if (full_name.end() - (full_name.begin() + pos_description) > 0) // Should not occur any more!!!!
      //      {
      description.assign(full_name.begin() + pos_description, full_name.end() );
      //      }
    }
  }

  void toupper_this_sequence()
  {
    data.toupper();
  }

  

  // TODO: Keep case of ambig.
  void change_ambigs()
  {
    if (type == dna || type == rna)
    {
      char * it     = data.begin();
      char * it_end = data.end();

      number_of_questionmarks  = 0;
      number_of_Ns             = 0;

      while (it != it_end)
      {
	char c       = *it;

	if ( c == 'N' || c == 'n' || c == '?' )
	{
	  // Convert all ambiguity codes to N's
	  *it = general_ambig;
	  if (general_ambig == '?')
	    ++number_of_questionmarks;
	  else
	    ++number_of_Ns;
	  //	  ++number_of_DNARNA_amgibs;
	}
	++it;
      }
    }
    else if (type == protein)
    {
      char * it     = data.begin();
      char * it_end = data.end();
      
      number_of_questionmarks  = 0;
      number_of_Xs             = 0;

      while (it != it_end)
      {
	char c       = *it;

	if ( c == 'X' || c == 'x' || c == '?' )
	{
	  // Convert all ambiguity codes to N's
	  ++number_of_DNARNA_amgibs;
	  *it = general_ambig;
	  if (general_ambig == '?' )
	    ++number_of_questionmarks;
	  else
	    ++number_of_Xs;
	}
	++it;
      }
    }

    // TODO: Adjust the numbers
  }


  void remove_gaps_stars_this_sequence(bool rg, bool rs)
  {
    //    char     *it     = data.begin();
    //    char     *it_end = data.end();
    
    
    // TODO

    if (rg)
    {}

    if (rs)
    {}
      

  }

  void gaps_to_ambig_this_sequence()
  {
    char     *it     = data.begin();
    char     *it_end = data.end();
    unsigned count=0;

    while (it != it_end)
    {
      if (*it =='-')
      {
	*it = general_ambig;
	++count;
      }
      ++it;
    }

    // If residues are computed, we have to change the numbers counts for gaps and ambigs:
    if (numbers_of_residues_computed)
    {
      number_of_gaps -= count; // Should indeed be 0 now.

      if (general_ambig == 'N' || general_ambig == 'n')
      {
	number_of_Ns += count;
      }
      else if (general_ambig == 'X' || general_ambig == 'x')
      {
	number_of_Xs += count;
      }
      else if (general_ambig == '?')
      {
	number_of_questionmarks += count;
      }
      if (type == dna || type == rna)
      {
	++number_of_DNARNA_amgibs += count;
      }
    }
  }


  void addCharToData(char c)
  {
    data.push_back(c);
  }


  // Reads line of data and removes only white spaces
  void getRawLine(CFile& infile)
  {
    register char            c;

    c = infile.getchar();
    while (!infile.fail() && c != '\n')
    {
      if ( isspace(c) )   // We remove white spaces.
      {
	c = infile.getchar();
	continue;
      }
      addCharToData(c);
      c = infile.getchar();
    }
  }

  // Reads line of data and removes only white spaces
  void getRawLine_toupper(CFile& infile)
  {
    register char            c;

    c = infile.getchar();
    while (!infile.fail() && c != '\n')
    {
      if ( isspace(c) )   // We remove white spaces.
      {
	c = infile.getchar();
	continue;
      }
      addCharToData(toupper(c));
      c = infile.getchar();
    }
  }


  // Called e.g. from Phobos
  void getLineOfDNA_toupper_ambig2N_removegaps (CFile& infile, unsigned &pos_in_seq)
  {
    register char       c;
    unsigned            pos = pos_in_seq;
    register char       local_general_ambig = toupper(general_ambig);

    //     ++pos;
    c = infile.getchar();
    while (!infile.fail() && c != '\n')
    {
      // if the char is a space or something we do not count, we subtract this further below
      ++number_raw_original_length;

      c = toupper(c);

      if ( c == 'A' || c == 'T' || c == 'G' || c == 'C')
      {
	addCharToData(c);
	++number_of_DNARNA_bases;
      }
      else if ( c == 'U')
      {
	addCharToData('U');
	++number_of_DNARNA_bases;
	++number_of_Us;
      }
      else if ( c == 'R' || c == 'Y' || c == 'K' || c == 'M' ||
		c == 'S' || c == 'W' || c == 'B' || c == 'D' ||
		c == 'H' || c == 'V' )
      {
	// Convert all ambiguity codes to N's
	addCharToData(local_general_ambig);
	++number_of_DNARNA_amgibs; // includes N and '?'
	if (local_general_ambig == 'N')
	  ++number_of_Ns;
	else
	  ++number_of_questionmarks;
      }
      else if (c == '?' || c == 'N')
      {
	addCharToData(local_general_ambig);
	if (local_general_ambig == 'N')
	  ++number_of_Ns;
	else
	  ++number_of_questionmarks;
	++number_of_DNARNA_amgibs;
      }
      else if ( isspace(c) )
      {
	--number_raw_original_length; // we do not count spaces
      }                              // Spaces are removed
      else if ( c=='-' || c=='*')     // What do we do with gaps and *s - Here they are removed,
	                              // but their positions are stored in order to able to reconstuct the origianl coordinates.
      {                               // I don't really know why we tolerate '*' ? They occur in protein sequences as stop codons.
	gap_and_star_koords.push_back(pos);    // If we want to reconstruct the true position in the sequence,
	                              // we need to keep track of the positions we cut out.
	if (c=='-')
	{
	  gap_koords.push_back(pos);
	  ++number_of_gaps;
	}
	else
	{
	  star_koords.push_back(pos);
	  ++number_of_stars;
	}
	//	c = infile.getchar();
	//	++pos;
      }
      else
      {
	// we do not count this -just to be consistent before we bail out 
	--number_raw_original_length;
	infile.clear(CFile::__fail_flag);
	break;
      }
      c = infile.getchar();
      ++pos;
    }
    pos_in_seq = pos;
  }


  void getLineOfDNA_generic (CFile& infile, unsigned &pos_in_seq,
			     processing_flag pflag)
  {
    register char            c;
    register char            c_upper;
    register char local_general_ambig = general_ambig;
    unsigned        pos = pos_in_seq;
    //    register signed char     delta;                  
    register bool            convert_toupper_bool  = (pflag & convert_toupper);   // handled
    register bool            convert_ambig2N_bool  = (pflag & convert_ambig2N);   // handled
    register bool            removegaps_bool       = (pflag & removegaps);        // handled
    register bool            removestars_bool      = (pflag & removestars);       // handled
    register bool            gaps_to_ambig_bool    = (pflag & gaps_to_ambig);     // handled

    // In the routine calling this function it should be checked, that the general_ambig
    // character is upper or lower case, as specified.


    if (convert_toupper_bool)
      local_general_ambig = toupper(general_ambig);

    //     ++pos;
    c = infile.getchar();
    c_upper = toupper(c);
    while (!infile.fail() && c != '\n')
    {
      // if the char is a space or something we do not count, we subtract this further below
      ++number_raw_original_length;

      if (c >= 'a' && c <= 'z' && convert_toupper_bool)
	c = c_upper;

      if ( c_upper == 'A' || c_upper == 'T' || c_upper == 'G' || c_upper == 'C')
      {
	addCharToData(c);
	++number_of_DNARNA_bases;
      }
      else if ( c_upper == 'U')
      {
	addCharToData(c);
	++number_of_DNARNA_bases;
	++number_of_Us;
      }
      else if (c_upper == 'R' || c_upper == 'Y' || c_upper == 'K' || c_upper == 'M' ||
	       c_upper == 'S' || c_upper == 'W' || c_upper == 'B' || c_upper == 'D' ||
	       c_upper == 'H' || c_upper == 'V' )
      {
	if (convert_ambig2N_bool)
	{
	  // Convert all ambiguity codes to N's, but we preserve the case. 'w' -> 'n' and 'W' -> 'N' etc.
	  if (c <='Z') // i.e. lower case.
	    addCharToData(  tolower(local_general_ambig)  );
	  else
	    addCharToData(  toupper(local_general_ambig)  );

	  ++number_of_DNARNA_amgibs; // includes N and '?'
	  if (local_general_ambig == '?' )
	    ++number_of_questionmarks;
	  else
	    ++number_of_Ns;
	}
	else
	{
	  addCharToData(c);
	}
      }
      else if (c == '?' || c_upper == 'N') // c is either: 'N', 'n', '?'
      {
	++number_of_DNARNA_amgibs;
	if (local_general_ambig == '?')
	{
	  addCharToData(local_general_ambig);
	  ++number_of_questionmarks;
	}
	else //	c is '?' or 'N' or 'n' and general_ambig is 'N' or 'n' --- add 'N' or 'n'
	{
	  ++number_of_Ns;
 	  if (c == '?')  // convert to general ambig with proper case:
	  {
	    addCharToData(local_general_ambig); // Append general_ambig with appropriate case.
	    if (local_general_ambig == '?')
	      ++number_of_questionmarks;
	    else
	      ++number_of_Ns;
	  }
	  else // c is 'N' or 'n', general_ambig is 'N' or 'n'
	  {
	    addCharToData(c); // Append N or n and preserve case !!! to be consistent with the flags.
	  }
	  // Note: In case of: convert to upper, we have already converted c to upper.
	}
      }
      else if ( isspace(c) )
      {
	--number_raw_original_length; // we do not count spaces
      }                              // Spaces are removed
      else if ( c=='-' || c=='*')     // What do we do with gaps and *s - Here they are removed,
	                              // but their positions are stored in order to able to reconstuct the origianl coordinates.
      {                               // I don't really know why we tolerate '*' ? They occur in protein sequences as stop codons.

	if (c=='-')
	{
	  if ( gaps_to_ambig_bool)
	  {
	    //	    std::cerr << "Gaps to ambig " << std::endl;
	    addCharToData(local_general_ambig);
	    if (general_ambig == '?' )
	      ++number_of_questionmarks;
	    else
	      ++number_of_Ns;
	    ++number_of_DNARNA_amgibs;
	  }
	  else if ( removegaps_bool )
	  {
	    //	    std::cerr << "Remove gaps " << std::endl;
	    gap_and_star_koords.push_back(pos);    // If we want to reconstruct the true position in the sequence,
	                                         // we need to keep track of the positions we cut out.
	    gap_koords.push_back(pos);
	    ++number_of_gaps;
	  }
	  else
	  {
	    addCharToData('-');
	    ++number_of_gaps;
	  }
	}
	else if (removestars_bool)
	{
	  gap_and_star_koords.push_back(pos);    // If we want to reconstruct the true position in the sequence,
	                                         // we need to keep track of the positions we cut out.
	  star_koords.push_back(pos);
	  ++number_of_stars;
	}
	//	c = infile.getchar();
	//	++pos;
      }
      else
      {
	// we do not count this -just to be consistent before we bail out 
	--number_raw_original_length;
	infile.clear(CFile::__fail_flag);
	break;
      }
      c = infile.getchar();
      c_upper = toupper(c);
      ++pos;
    }
    pos_in_seq = pos;
  }


  void getLineOfProtein_generic (CFile& infile, unsigned &pos_in_seq,
				 processing_flag pflag)
  {
    register char            c;
    register char            c_upper;
    register char            local_general_ambig = general_ambig;


    unsigned        pos = pos_in_seq;
    //    register signed char     delta;                  
    register bool            convert_toupper_bool  = (pflag & convert_toupper);   // handled
    //    register bool            convert_ambig2N_bool  = (pflag & convert_ambig2N);   // handled
    register bool            removegaps_bool       = (pflag & removegaps);        // handled
    //    register bool            removestars_bool      = (pflag & removestars);       // handled
    register bool            gaps_to_ambig_bool    = (pflag & gaps_to_ambig);// handled

    // In the routine calling this function it should be checked, that the general_ambig
    // character is upper or lower case, as specified.

    if (convert_toupper_bool)
      local_general_ambig = toupper(general_ambig);

    //     ++pos;
    c = infile.getchar();
    c_upper = toupper(c);
    while (!infile.fail() && c != '\n')
    {
      // if the char is a space or something we do not count, we subtract this further below
      ++number_raw_original_length;

      if (c >= 'a' && c <= 'z' && convert_toupper_bool)
	c = c_upper;

      if (c == '?' || c_upper == 'X' )
      {
	if (local_general_ambig == '?')
	{
	  addCharToData(local_general_ambig);
	  ++number_of_questionmarks;
	}
	else // c is '?' or 'X' or 'x' and general_ambig is 'X' or 'x'
	{
	  ++number_of_Xs;
	  if (c == '?')
	  {
	    addCharToData(local_general_ambig);
	  }
	  else // c is 'X' 'x', general_ambig is 'X' 'x'
	    addCharToData(c); // Append X or x and preserve case !!! to be consistent with the flags.
	  // Note: In case of convert to upper, we have already converted c to upper.
	}
      }
      else // The following code assumes we do not have X or ?. Otherwise Xs are duplicated in the data.
	if ( is_aa_or_aa_ambig_extended(c) ) // includes other ambigs (!'X', !'x' ,!'?')
      {
	addCharToData(c);
	++number_of_AAs; // Counts ambigs but not X or ?
      }
      else if ( isspace(c) )
      {
	--number_raw_original_length; // we do not count spaces
      }                              // Spaces are removed
      else if ( c=='-' || c=='*')     // What do we do with gaps and *s - Here they are removed,
	                              // but their positions are stored in order to able to reconstuct the origianl coordinates.
      {                               // I don't really know why we tolerate '*' ? They occur in protein sequences as stop codons.

	if (c=='-')
	{
	  if ( gaps_to_ambig_bool)
	  {
	    addCharToData(local_general_ambig);
	    if (local_general_ambig == '?' )
	      ++number_of_questionmarks;
	    else
	      ++number_of_Xs;
	  }
	  else if ( removegaps_bool )
	  {
	    gap_and_star_koords.push_back(pos);    // If we want to reconstruct the true position in the sequence,
	                                         // we need to keep track of the positions we cut out.
	    gap_koords.push_back(pos);
	    ++number_of_gaps;
	  }
	  else
	  {
	    addCharToData('-');
	    ++number_of_gaps;
	  }
	}

/* 	else if (removestars_bool) */
/* 	{ */
/* 	  gap_and_star_koords.push_back(pos);    // If we want to reconstruct the true position in the sequence, */
/* 	                                         // we need to keep track of the positions we cut out. */
/* 	  star_koords.push_back(pos); */
/* 	  ++number_of_stars; */
/* 	} */
	//	c = infile.getchar();
	//	++pos;
      }
      else // Unknown symbol
      {
	// we do not count this -just to be consistent before we bail out 
	--number_raw_original_length;
	infile.clear(CFile::__fail_flag);
	break;
      }
      c = infile.getchar();
      c_upper = toupper(c);
      ++pos;
    }
    pos_in_seq = pos;
  }


  void getLineOfDNA_toupper_removegaps (CFile& infile, unsigned &pos_in_seq)
  {
    register char       c;
    unsigned            pos = pos_in_seq;
    register char       local_general_ambig = toupper(general_ambig);

    //     ++pos;
    c = infile.getchar();
    while (!infile.fail() && c != '\n')
    {
      // if the char is a space or something we do not count, we subtract this further below
      ++number_raw_original_length; 
      c = toupper(c);

      if ( c == 'A' || c == 'T' || c == 'G' || c == 'C')
      {
	addCharToData(c);
	++number_of_DNARNA_bases;
      }
      else if ( c == 'N' || c == '?')
      {
	addCharToData(local_general_ambig);
	if (general_ambig == 'N')
	  ++number_of_Ns;
	else
	  ++number_of_questionmarks;
	++number_of_DNARNA_amgibs; // includes N and '?'
      }
      else if ( c == 'U')
      {
	addCharToData('U');
	++number_of_DNARNA_bases;
	++number_of_Us;
      }
      else if ( c == 'R' || c == 'Y' || c == 'K' || c == 'M' ||
		c == 'S' || c == 'W' || c == 'B' || c == 'D' ||
		c == 'H' || c == 'V' )
      {
	addCharToData(c);
	++number_of_DNARNA_amgibs;
      }
      else if ( isspace(c) )          // Spaces are removed
      {
	--number_raw_original_length; // we do not count spaces
      }                              
      else if ( c=='-' || c=='*')     // What do we do with gaps and *s - Here they are removed, but their positions are stored in order to reconstuct the origianl coordinates.
      {                               // I don't really know why we tolerate '*' ?
	gap_and_star_koords.push_back(pos);    // If we want to reconstruct the true position in the sequence,
	                              // we need to keep track of the positions we cut out.
	if (c=='-')
	{
	  gap_koords.push_back(pos);
	  ++number_of_gaps;
	}
	else
	{
	  star_koords.push_back(pos);
	  ++number_of_stars;
	}
	//	c = infile.getchar();
	//	++pos;
      }
      else
      {
	// Something went wrong -> the last reading event from the file failed
	// -> we set the fail flag, this informs the user that something went wrong
	// We do not count this - just to be consistent before we bail out 
	--number_raw_original_length;
	infile.clear(CFile::__fail_flag);
	break;
      }
      c = infile.getchar();
      ++pos;
    }
    pos_in_seq = pos;
  }



  void getLineOfProtein_toupper_removegaps (CFile& infile, unsigned &pos_in_seq)
  {
    register char            c;
    unsigned        pos = pos_in_seq;

    //     ++pos;
    c = infile.getchar();
    char local_general_ambig = toupper(general_ambig);

    while (!infile.fail() && c != '\n')
    {
      // if the char is a space or something we do not count, we subtract this further below
      ++number_raw_original_length;
      c = toupper(c);

      if (c == 'X' || c=='?')
      {
	addCharToData(local_general_ambig);
	if (local_general_ambig == 'X' )
	  ++number_of_Xs;
	else
	  ++number_of_questionmarks;
      }
      else if (is_aa_or_aa_ambig_extended(c) )
      {
	addCharToData(c);
	++number_of_AAs;
      }
      else if ( isspace(c) )// Spaces are removed
      {
	--number_raw_original_length; // we do not count spaces
      }
      else if ( c=='-' || c=='*')     // What do we do with gaps and *s - Here they are removed, but their positions are stored in order to reconstuct the origianl coordinates.
      {                                        // I don't really know why we tolerate '*' ?
	gap_and_star_koords.push_back(pos);    // If we want to reconstruct the true position in the sequence,
	                                       // we need to keep track of the positions we cut out.
	if (c=='-')
	{
	  gap_koords.push_back(pos);
	  ++number_of_gaps;
	}
	else
	{
	  star_koords.push_back(pos);
	  ++number_of_stars;
	}
      }
      else
      {
	// we do not count this -just to be consistent before we bail out 
	--number_raw_original_length;

	infile.clear(CFile::__fail_flag);
	break;
      }
      c = infile.getchar();
      ++pos;
    }
    pos_in_seq = pos;
  }

  void getLineOfDNA_toupper_ambig2N_gaps2ambig (CFile& infile)
  {
    register char            c;

    c = infile.getchar();
    while (!infile.fail() && c != '\n')
    {
      // if the char is a space or something we do not count, we subtract this further below
      ++number_raw_original_length;

      c = toupper(c);

      if ( c == 'A' || c == 'T' || c == 'G' || c == 'C')
      {
	addCharToData(c);
	++number_of_DNARNA_bases;
      }
      else if ( c == 'N' || c == '?')
      {
	addCharToData(general_ambig);
	if (general_ambig == 'N')
	  ++number_of_Ns;
	else
	  ++number_of_questionmarks;
	++number_of_DNARNA_amgibs;
      }
      else if ( c == 'U')
      {
	addCharToData('U');
	++number_of_DNARNA_bases;
	++number_of_Us;
      }
      else if ( c == 'R' || c == 'Y' || c == 'K' || c == 'M' ||
		c == 'S' || c == 'W' || c == 'B' || c == 'D' ||
		c == 'H' || c == 'V' )
      {
	// Convert all ambiguity codes to N's
	addCharToData(general_ambig);
	if (general_ambig == 'N')
	  ++number_of_Ns;
	else
	  ++number_of_questionmarks;
	++number_of_DNARNA_amgibs;
      }
      else if (c=='-')
      {
	addCharToData(general_ambig);
	if (general_ambig == 'N')
	  ++number_of_Ns;
	else
	  ++number_of_questionmarks;
	++number_of_DNARNA_amgibs;
      }
      else if ( isspace(c) )   // We remove white spaces.
      {                        // I don't really know why we tolerate '*' ?
	--number_raw_original_length; // we do not count spaces
      }
      else
      {
	// we do not count this -just to be consistent before we bail out 
	--number_raw_original_length;

	infile.clear(CFile::__fail_flag);
	return;
      }
      c = infile.getchar();
    }
  }





 public:
 CSequence_Mol(DataTypesEnum dt, char p_general_ambig = '\0'):
  type(dt), general_ambig(p_general_ambig),
    numbers_of_residues_computed(false), bitcode_recoded(false)
  {
    data.reserve(10000);
    if (p_general_ambig == '\0') // Use the typical ambig character
    {
      if      (dt == dna)     general_ambig = 'N';
      else if (dt == protein) general_ambig = 'X';
      else                    general_ambig = '?';
    }
    reset_numbers_of_residues();
  }


 CSequence_Mol(DataTypesEnum dt, faststring name, unsigned len, char p_general_ambig = '\0'):
  full_name(name),type(dt), general_ambig(p_general_ambig),
    numbers_of_residues_computed(false), bitcode_recoded(false)
  {
    if (len < -1u)
      ++len;
    data.reserve(len);

    find_shortname_and_description();

    if (p_general_ambig == '\0') // Use the typical ambig character
    {
      if      (dt == dna)     general_ambig = 'N';
      else if (dt == protein) general_ambig = 'X';
      else                    general_ambig = '?';
    }
    reset_numbers_of_residues();
  }


  //**************************************************************************
  // pos1 must be the first column starting with 0.
  // pos2 must be the index after the last column.
  // pos2-pos1 must be the number of bases that are copied to this sequence.
  //**************************************************************************
 CSequence_Mol(const CSequence_Mol& s, unsigned pos1=0, unsigned pos2 = -1u):
  full_name(s.full_name),
    short_name(s.short_name),
    description(s.description),
    type(s.type),
    general_ambig(s.general_ambig),
    numbers_of_residues_computed(false),
    bitcode_recoded(s.bitcode_recoded)
  {
    if (pos1 == 0 && pos2 >= s.length() )
      data = s.data;
    else
    {
      if (pos2 > s.length() )
	pos2 = s.length();
      if (pos1 >= pos2)
      {
	return;
      }
      data.assign(s.data, pos1, pos2-pos1);
    }
  }
  


  // Should only be applied to dna sequences
  void convert_ambig2N_this_sequence()
  {
    if ( !(type == dna || type == rna) )
      return;

    // This is a DNA sequence:

    char * it     = data.begin();
    char * it_end = data.end();

    while (it != it_end)
    {
      char c       = *it;
      char c_upper = toupper(c);

      if ( c_upper == 'R' || c_upper == 'Y' || c_upper == 'K' || c_upper == 'M' ||
	   c_upper == 'S' || c_upper == 'W' || c_upper == 'B' || c_upper == 'D' ||
	   c_upper == 'H' || c_upper == 'V' )
      {
	// Convert all ambiguity codes to N's
	*it = general_ambig;
	if (general_ambig == 'N' || general_ambig == 'n')
	  ++number_of_Ns;
	else
	  ++number_of_questionmarks;
	++number_of_DNARNA_amgibs;
      }
      ++it;
    }
  }


  unsigned length() const
  {
    return data.size();
  }

  DataTypesEnum get_datatype()
  {
    return type;
  }

  faststring type_as_string()
  {
    if (type == dna)
      return faststring("DNA");
    else if (type == rna)
      return faststring("RNA");
    else if (type == protein)
      return faststring("Protein");
    return "Unknown";
  }
  

  void set_taxon_and_sequence(const faststring& taxonName,
                              const faststring& seqData, DataTypesEnum datatype=auto_detect_type)
  {
/*     unsigned pos; */

    if (numbers_of_residues_computed)
    {
      reset_numbers_of_residues();
    }

    full_name = taxonName;

    find_shortname_and_description();
/*     if ( (pos = taxonName.find(' '))!=faststring::npos) */
/*     { */
/*       short_name  = taxonName.substr(0,pos+1); */
/*       description = taxonName.substr(pos+1); */
/*     } */
/*     else */
/*     { */
/*       short_name  = taxonName; */
/*       description.clear(); */
/*     } */
    data = seqData;

    //
    //    toupper_this_sequence();

    compute_numbers_of_residues();
    if (datatype == auto_detect_type || datatype == unknown)
      auto_detect_datatype();

    //    std::cerr << "Data type: " << type_as_string()  << std::endl;
    //    std::cerr << "General ambig: " << general_ambig  << std::endl;

    if (auto_detect_ambig_detect_conflict() )
    {
      std::cerr << "Conflict in amgiguity code used in imported sequence." << std::endl;
    }
    
    //    std::cerr << "General ambig: " << general_ambig  << std::endl;

    change_ambigs();
    //    remove_gaps_stars_this_sequence(true, true);

    //    general_ambig = ambig;
    //    type = t;
  }


  void set_taxon_and_random_DNA_sequence(const faststring& taxonName,
					 unsigned len,
					 double piA, double piC, double piG)
  {
    unsigned pos;

    type = dna;
    general_ambig = '?';

    data.clear();
    data.reserve(len+1);

    reset_numbers_of_residues();

    full_name = taxonName;
    if ( (pos = taxonName.find(' '))!=faststring::npos)
    {
      short_name  = taxonName.substr(0,pos+1);
      description = taxonName.substr(pos+1);
    }
    else
    {
      short_name  = taxonName;
      description.clear();
    }

    // Compute random sequence:
    unsigned i;
    double   z;

    double bA = piA;
    double bC = bA+piC;
    double bG = bC+piG;

    for (i=0; i<len; ++i)
    {
      z = rand()/(RAND_MAX+1.0);
      if ( z < bA )
	data.push_back('A');
      else if ( z < bC )
	data.push_back('C');
      else if ( z < bG )
	data.push_back('G');
      else
	data.push_back('T');
    }

    compute_numbers_of_residues();

    //    std::cerr << "Data type: " << type_as_string()  << std::endl;
    //    std::cerr << "General ambig: " << general_ambig  << std::endl;
    //    std::cerr << "General ambig: " << general_ambig  << std::endl;
  }


  char& operator[](unsigned pos)
  {
    return data[pos];
  }

  char get_pos(unsigned pos)
  {
    return data[pos];
  }

/*   void set_pos(unsigned pos, char c) */
/*   { */
/*     data[pos] = c; */
/*   } */

  bool is_bitcoderecoded()
  {
    return bitcode_recoded;
  }

  unsigned full_original_length() const
  {
    // Remember that some numbers include others:
    // number_of_DNARNA_amgibs includes 'Ns' and '?', so they are not added again:
    // number_of_DNARNA_bases  includes 'Us'
    if (type == dna || type == rna )
    {
      return number_of_DNARNA_bases + number_of_DNARNA_amgibs + number_of_gaps + number_of_stars;
    }
    else if (type == protein)
    {
      return number_of_AAs + number_of_gaps + number_of_stars
	                   + number_of_Xs + number_of_questionmarks;
    }
    else
    {
      return number_raw_original_length;
    }
  }

  unsigned length_DNARNA_bases() const
  {
    return number_of_DNARNA_bases;
  }

  const char* getName() // const
  {
    return short_name.c_str();
  }

  const faststring& getName_faststring() const
  {
    return short_name;
  }

  const faststring& getSeq_faststring() const
  {
    return data;
  }

  faststring getPhylipName()
  {
    faststring str(short_name);

    str.resize(10, ' ');
    return str;
  }

  const char* getFullName() // const
  {
    return full_name.c_str();
  }

  const faststring getFullName_faststring() const
  {
    return full_name;
  }

  const char* getDescription() // const
  {
    return description.c_str();
  }

  const char* getSeqStr()
  {
    return data.c_str();
  }

  faststring getPartialSeq(size_t pos, size_t n)
  {
    return data.substr(pos, n);
  }

  const char* getSeqBegin()
  {
    return data.begin();
  }

  const char* getSeqEnd()
  {
    return data.end();
  }

  const char* getSeqRBegin()
  {
    return data.rbegin();
  }

  const char*getSeqREnd()
  {
    return data.rend();
  }

  void reset_numbers_of_residues()
  {
    numbers_of_residues_computed = false;

    number_of_AAs = 0;
    number_of_DNARNA_bases = 0;
    number_of_DNARNA_amgibs = 0;
    number_of_Us = 0;
    number_of_gaps = 0;
    number_of_stars = 0;
    number_of_questionmarks = 0;
    number_of_Xs = 0;
    number_of_Ns = 0;

    number_CG = 0;

    number_of_questionmarks = 0;
    number_raw_original_length = 0;
  }

  void compute_numbers_of_residues()
  {
    const char *it=data.begin(), *it_end=data.end();

    if (numbers_of_residues_computed)
    {
      std::cerr << "Request for computing number of residues - even though they have already been determined" << std::endl; 
      return;
    }

    reset_numbers_of_residues();

    for (;it<it_end; ++it)
    {
      char c = *it;

      if (is_DNA_or_RNA_base(c))
      {
	if (c == 'c' || c == 'C' || c == 'G' || c == 'g' )
	  ++number_CG;
	++number_of_DNARNA_bases;
      }

      if (is_DNA_iupac_ambig(c))
	++number_of_DNARNA_amgibs;

      if (is_aa_or_aa_ambig_extended(c) && c!= 'X' && c != '?')
	++number_of_AAs;

      if (c == 'N')
      {
	++number_of_Ns;
      }

      if (c == '-')
      {
	++number_of_gaps;
      }

      if (c == '*')
      {
	++number_of_stars;
      }

      if (c == 'X')
      {
	++number_of_Xs;
      }

      if (c == '?')
      {
	++number_of_questionmarks;
      }

      if (c == 'U' || c == 'u')
      {
	++number_of_Us;
      }

      numbers_of_residues_computed = true;
    }
  }

  void auto_detect_datatype()
  {
    if (!numbers_of_residues_computed)
      compute_numbers_of_residues();

    unsigned aa_stuff     = number_of_AAs + number_of_Xs + number_of_stars;
    unsigned DNARNA_stuff = number_of_DNARNA_bases + number_of_DNARNA_amgibs;

    // reine DNA: 100 > 0.25 * 100 - false - DNA

    if (aa_stuff < 0.1*data.length() && DNARNA_stuff < 0.1*data.length() )
      type = unknown;
    else if (aa_stuff > 0.25*DNARNA_stuff)
    {
      if (number_of_Us > 0)
	type = rna;
      else
	type = dna;
    }
    else
    {
      type = protein;
    }
  }

  bool auto_detect_ambig_detect_conflict()  // Return true in case of ambig conflict, else false
  {
    if (!numbers_of_residues_computed)
      compute_numbers_of_residues();

    if (type == protein)
      general_ambig = 'X';
    else general_ambig = 'N';

    if (number_of_questionmarks > 0)
      general_ambig = '?';

    // Determine conlict:
    return (   number_of_questionmarks > 0   && 
	       (number_of_Xs > 0 || ((type==dna || type==rna) && number_of_Ns > 0))); 
  }


  void assignSeq(faststring &newseq, processing_flag flag=(processing_flag)0)
  {
    data.assign(newseq);
    numbers_of_residues_computed = false;

    if (flag & compute_numbers_of_residues_flag)
      compute_numbers_of_residues();
    else
    {
      reset_numbers_of_residues();
    }
  }

  void maskSeq(const char *start, const char *end, char c)
  {
    char *pos = const_cast<char*>(start);
    while (pos != end)
    {
      *pos = c;
      ++pos;
    }
  }

  void maskSeq_tolower(const char *start, const char *end)
  {
    char *pos = const_cast<char*>(start);
    while (pos != end)
    {
      *pos = tolower(*pos);
      ++pos;
    }
  }

  unsigned find_pos_first_symbol_not_in_alphabet(const char * alphabet=NULL)
  {
    if (alphabet == NULL)
    {
      if (type == dna)
      {
	alphabet = DNARNA_symbols;
      }
      else if (type == protein)
      {
	alphabet = aa_symbols;
      }
      else
      {
	return -1u;
      }
    }

    return data.find_first_not_of(alphabet);
  }

  // Fills str with the sequnce data, that contains gaps and stars, that had been removed
  // temporarily.
  void getSequence_fill_in_gaps_and_stars(faststring &str)
  {
/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    unsigned upos = 0;
    unsigned pos_in_gap_and_star_koords=0;
    unsigned pos_in_gap_koords=0;
    unsigned pos_in_star_koords=0;

    const char *pos = data.begin();
    //    const char *end = data.end();

    unsigned N      = gap_and_star_koords.size();
    unsigned Ngaps  = gap_koords.size();
    unsigned Nstars = star_koords.size();

    //    unsigned left_this_line = char_per_line;

    str.clear();

    if (N == 0 && Ngaps == 0 && Nstars == 0)
    {
      str = data;
      return;
    }


    unsigned N_all_char = data.size()+N;
    str.reserve(N_all_char);

    while (upos < N_all_char)
    {
      if (pos_in_gap_and_star_koords == N)
	break;

      if ( gap_and_star_koords.get_unckecked(pos_in_gap_and_star_koords) == upos )
      {
	if ( pos_in_gap_koords < Ngaps && gap_koords.get_unckecked(pos_in_gap_koords) == upos ) // this is a gap
	{
	  str.push_back('-');
	  ++pos_in_gap_koords;
	  ++pos_in_gap_and_star_koords;
	}
	else
	  if ( pos_in_star_koords < Nstars && star_koords.get_unckecked(pos_in_star_koords) == upos ) // this is a gap
	  {
	    str.push_back('*');
	    ++pos_in_star_koords;
	    ++pos_in_gap_and_star_koords;
	  }
      }
      else
      {
	str.push_back(*pos);
	++pos;
      }
      ++upos;
      //      --left_this_line;

/*       if (left_this_line == 0) */
/*       { */
/* 	str.push_back('\n'); */
/* 	left_this_line = char_per_line; */
/*       } */
    }

    // copy rest without gaps and stars
    while (upos < N_all_char)
    {
      str.push_back(*pos);
      ++pos;
      ++upos;
      //      --left_this_line;

/*       if (left_this_line == 0) */
/*       { */
/* 	str.push_back('\n'); */

/* 	left_this_line = char_per_line; */
/*       } */
    }
/*     if (left_this_line != char_per_line)    // We have already written a partial line */
/*       str.push_back('\n'); */
  }

  // Replace part of the sequence. pos1 and pos2 are in the destination sequence.
  // - erases region from pos1 to pos2
  //   pos1 is the first column pos2-pos1 is the number of bases that are removed,
  //   i.e. pos2 is the column after the last column that is to be removed.
  // - insert s with full length.
  void replace_part_of_sequence(CSequence_Mol &s, unsigned pos1, unsigned pos2)
  {
    if (type != s.type)
    {
      std::cerr << "Critical warning: Replacing part of sequence with different data type." << std::endl;
    }
    data.replace(pos1, pos2-pos1, s.data);
  }

  void append_sequence(CSequence_Mol &s)
  {
    if (type != s.type)
    {
      std::cerr << "Critical warning: Appending sequence with different data type." << std::endl;
    }
    data.append(s.data);
  }

  void append_residue_unchecked(char c)
  {
    data.push_back(c);
  }

  void writeSequence_fill_in_gaps_and_stars(FILE *of, unsigned char_per_line=50)
  {
/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    unsigned upos = 0;
    unsigned pos_in_gap_and_star_koords=0;
    unsigned pos_in_gap_koords=0;
    unsigned pos_in_star_koords=0;

    const char *pos = data.begin();
    //    const char *end = data.end();

    unsigned N      = gap_and_star_koords.size();
    unsigned Ngaps  = gap_koords.size();
    unsigned Nstars = star_koords.size();

    unsigned left_this_line = char_per_line;

    unsigned N_all_char = data.size()+N;

    fprintf(of, ">%s\n", full_name.c_str());

    while (upos < N_all_char)
    {
      if (pos_in_gap_and_star_koords == N)
	break;

      if ( gap_and_star_koords.get_unckecked(pos_in_gap_and_star_koords) == upos )
      {
	if ( pos_in_gap_koords < Ngaps && gap_koords.get_unckecked(pos_in_gap_koords) == upos ) // this is a gap
	{
	  putc('-', of);
	  ++pos_in_gap_koords;
	  ++pos_in_gap_and_star_koords;
	}
	else
	  if ( pos_in_star_koords < Nstars && star_koords.get_unckecked(pos_in_star_koords) == upos ) // this is a gap
	  {
	    putc('*', of);
	    ++pos_in_star_koords;
	    ++pos_in_gap_and_star_koords;
	  }
      }
      else
      {
	putc(*pos, of);
	++pos;
      }
      ++upos;
      --left_this_line;

      if (left_this_line == 0)
      {
	putc('\n', of);

	left_this_line = char_per_line;
      }
    }

    while (upos < N_all_char)
    {
      putc(*pos, of);
      ++pos;
      ++upos;
      --left_this_line;

      if (left_this_line == 0)
      {
	putc('\n', of);

	left_this_line = char_per_line;
      }
    }
    if (left_this_line != char_per_line)    // We have already written a partial line
 	putc('\n', of);
  }


  void writeSequence(FILE *of, unsigned char_per_line=50) //const
  {
    writeSequence_fasta(of, 0, -1u, full_name.c_str(), char_per_line);
  }

  void writeSequence_fasta(FILE *of, unsigned char_per_line=50) //const
  {
    writeSequence_fasta(of, 0, -1u, full_name.c_str(), char_per_line);
  }

  void writeSequence(FILE *of,
		     unsigned    pos_beg,
		     unsigned    pos_end,
		     const char  *s_name,
		     unsigned    char_per_line=50) const
  {
    writeSequence_fasta(of, pos_beg, pos_end, s_name, char_per_line);
  }


  void writeSequence_fasta(FILE *of,
			   unsigned    pos_beg, // 0 based
			   unsigned    pos_end, // 0 based, stop index
			   const char  *s_name,
			   unsigned    char_per_line=50) const
  {
    const char *beg = data.begin() + pos_beg;
    const char *end = data.begin() + pos_end;
    const char *tmp_end;
 
/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    if (pos_end == -1u || end > data.end() )
      end = data.end();

    if (char_per_line == -1u)
      char_per_line = data.length();

    fprintf(of, ">%s\n", s_name);
    
    tmp_end = beg + char_per_line;
    if (tmp_end > end)
      tmp_end = end;

    while (beg < end)
    {
      fprintf(of, "%.*s\n", int(tmp_end - beg), beg);

      beg = tmp_end;
      tmp_end = beg + char_per_line;
      if (tmp_end > end)
	tmp_end = end;
    }
  }

  // Some programs do not accept too many Ns in a row.
  // This function export sequences so that Nregions are collapsed
  // to a maximum number of successive Ns.
  // USE WITH CAUTION: This function changes your sequence data!!!
  // The output is not identical to your original sequnece data.


  void writeSequence_fasta_collaps_NX_regions(FILE *of,
			   unsigned    char_per_line=50,
		           unsigned    maximum_number_of_successive_NXs=100)
  {
    writeSequence_fasta_collaps_NX_regions(of, 
					   0,
					   -1u,
					   full_name.c_str(),
					   char_per_line,
					   maximum_number_of_successive_NXs);
  }

  void writeSequence_fasta_collaps_NX_regions(FILE *of,
					      unsigned    pos_beg, // 0 based
					      unsigned    pos_end, // 0 based , stop index
					      const char  *s_name,
					      unsigned    char_per_line=50,
					      unsigned    maximum_number_of_successive_NXs=100) const
  {
    const char *beg = data.begin() + pos_beg;
    const char *end = data.begin() + pos_end;
    //    const char *tmp_end;
 
/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    if (pos_end == -1u || end > data.end() )
      end = data.end();

    fprintf(of, ">%s\n", s_name);

    unsigned chars_in_this_line=0;
    unsigned tmp_succNs=0;

    while (beg != end)
    {
      char c = *beg;
      char c_toupper = toupper(c);

      if (c_toupper == 'N' || c_toupper == 'X')
      {
	++tmp_succNs;
	if (tmp_succNs <= maximum_number_of_successive_NXs) // print the N or X
	{
	  putc(c, of);
	  ++chars_in_this_line;

	  if (chars_in_this_line == char_per_line)
	  {
	    putc('\n', of);
	    chars_in_this_line = 0;
	  }
	}
	// unwritten else: do not print the N or X
      }
      else
      {
	putc(c, of);
	tmp_succNs = 0;
	++chars_in_this_line;

	if (chars_in_this_line == char_per_line)
	{
	  putc('\n', of);
	  chars_in_this_line = 0;
	}
      }
      ++beg;
    }
    if (chars_in_this_line != 0)
	  putc('\n', of);
  }

  void writeSequence_partial_fasta(FILE *of,
				   unsigned    pos_beg, // 0 based
				   unsigned    pos_end, // 0 based , stop index
				   const char  *s_name,
				   unsigned    char_per_line=50,
				   bool        degap=true) const
  {
    const char *beg = data.begin() + pos_beg;
    const char *end = data.begin() + pos_end;

/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    if (pos_end == -1u || end > data.end() )
      end = data.end();

    fprintf(of, ">%s\n", s_name);

    unsigned chars_in_this_line=0;
    
    if (degap)
    {
      while (beg != end)
      {
	char c = *beg;

	if (c != '-')
	{
	  putc(c, of);
	  ++chars_in_this_line;

	  if (chars_in_this_line == char_per_line)
	  {
	    putc('\n', of);
	    chars_in_this_line = 0;
	  }
	}
	// unwritten else: do not print -
      ++beg;
      }
    }
    else
    {
      while (beg != end)
      {
	putc(*beg, of);
	++chars_in_this_line;
	
	if (chars_in_this_line == char_per_line)
	{
	  putc('\n', of);
	  chars_in_this_line = 0;
	}
	++beg;
      }
    }
  }

  void writeSequence_fasta_filter_sites(FILE *of,
					std::vector<unsigned> &site_filter,
					int         offset,   // 0 if site_filter is 0 based, 1 if site_filter is 1 based
					const char  *s_name,
					unsigned    char_per_line=50) const
  {
    const char *beg = data.begin();
    const char *end = data.end();

/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    fprintf(of, ">%s\n", s_name);

    unsigned chars_in_this_line=0;
    int i=offset;

    {
      while (beg != end)
      {
	if (site_filter[i] != 0)
	{
	  putc(*beg, of);
	  ++chars_in_this_line;
	
	  if (chars_in_this_line == char_per_line)
	  {
	    putc('\n', of);
	    chars_in_this_line = 0;
	  }
	}
	// Move on in sequence and site_filter vector
	++beg;
	++i;
      }
    }
  }

  void writeSequence_phylip(FILE *of,
			    unsigned    pos_beg, // 0 based
			    unsigned    pos_end, // 0 based, stop index.
			   const char  *s_name) const
  {
/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    const char *beg = data.begin() + pos_beg;
    // const char *end = data.begin() + pos_end;
 
/*     if (pos_end == -1u || end > data.end() ) */
/*       end = data.end(); */

    unsigned name_len = strlen(s_name);
    if (name_len < 11)
      fprintf(of, "%-10.10s", s_name);
/*     else */
/*     { */
/*       char str[11]; */
/*       std::strncpy(str, s_name, 10); */
/*       str[10] = '\0'; */
/*       fprintf(of, "%s", str); */
/*     } */
    fprintf(of, "%.*s\n", pos_end - pos_beg, beg);
  }

  // Expects 0 bases coordinates. The range starts at pos_beg and pos_end
  // points to the base after the last position in the range.
  // Clearly pos_end > pos_beg. In the empty range, pos_end = pos_beg+1.

  void writeSequence_revComp(FILE *of,
			     unsigned pos_beg,  // 0 based index
			     unsigned pos_end,  // 0 based index after the last pos.
			     const char *s_name,
			     unsigned    char_per_line=50) const
  {
/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    const char *beg = data.begin() + pos_beg;
    const char *end = data.begin() + pos_end;

    unsigned interleaved = char_per_line;

    fprintf(of, ">%s\n", s_name);

    
    // Walking backwards, we shift the range:
    --beg;   // 0 based indices. beg is the rend index. Move beg to pos before "first pos". This is the stop index. 
    --end;   // 0 based index. Move end to index at which we start to move.

    // Example: Print first 3 bases.
    // Then: pos_beg=0, pos_end=3, 0 based, pos_end is index behind last pos in range. 
    // Thus: beg="-1+seq_beg" and end="2+seq_beg", so print 2, 1, 0, and stop at -1

    for (; beg < end; --end)
    {
      if (interleaved == 0)
      {
	fputc('\n', of);
	interleaved = char_per_line;
      }
      fputc( complementBase(*end), of ); // From CDnaString.h
      --interleaved;
    }
    fputc('\n', of);
  }

  unsigned baseComposition(std::vector<unsigned> &v) const
  {
    if ( v.size() < 256 )
    {
      std::cerr << "Internal error. -24" << std::endl;
      exit(0);
    }

    char *pos = data.begin();
    char *end = data.end();

    unsigned len = end - pos;

    while (pos != end)
    {
      ++v[*pos];
      ++pos;
    }
    return len;
  }

  void sliding_window_CG_AT_content(std::vector<double> &v, unsigned window_size)
  {
    char *end = data.end();

    const char *range_beg = data.begin();
    const char *range_end;

    double   CG_prop;
    unsigned CG_num;
    unsigned AT_num;

    while (1)
    {
      range_end = range_beg + window_size;

      if (range_end > end)
	range_end = end;

      if (range_end == range_beg)
	break;

      CG_AT_content_in_region(range_beg, range_end, AT_num, CG_num);

      //      std::cerr << AT_num << " " << CG_num << std::endl;

      if (CG_num == 0 && AT_num == 0) // e.g. if all chars are ambigs
	CG_prop = -1;
      else
	CG_prop = (double)CG_num/(CG_num+AT_num);
      
      v.push_back(CG_prop);
 
      range_beg = range_end;
    }
  }


  unsigned long long AsciiComposition_CpG_CpNpG(unsigned long long v[], unsigned long long &CpG, unsigned long long &CpNpG, unsigned long long &Nislands) const
  {
    char *pos = data.begin();
    char *end = data.end();
    char last = 'N';  // An island is not counted if located at the beginning or end of the sequnce.

    unsigned long long len = end - pos;
    

    end = end - 2;
    while (pos < end)
    {
      char c1 = *pos;

      if ( c1 == 'C' || c1 == 'c' )
      {
	char c2 = *(pos+1);
	char c3 = *(pos+2);
	if ( c2 == 'G' || c2 == 'g') ++CpG;
	if ( c3 == 'G' || c3 == 'g') ++CpNpG;
      }
      if (c1 == 'N' && last != 'N') 
	++Nislands;

      ++v[(int)c1];
      ++pos;
      last = c1;
    }

    // Now deal with the second last position:
    if (len > 1 )
    {
      char c1 = *pos;

      if ( c1 == 'C' || c1 == 'c' )
      {
	char c2 = *(pos+1);
	if ( c2 == 'G' || c2 == 'g' ) ++CpG;
      }
      if (c1 == 'N' && last != 'N') 
	++Nislands;
      ++v[(int)c1];
      ++pos;
      last = c1;
    }

    // Now deal with the last position:
    if (len > 0)
    {
      ++v[(int)*pos];
      if (*pos == 'N') // We don't count the island if at the end of the sequence 
      {
	if  (last == 'N') // it has already been counted so undo this 
	{
	  --Nislands;
	}
      // else
      //    we don't count it
      }
    }
    return len;
  }

  void debug_print_gap_coords(std::ostream &os)
  {
     unsigned *git = gap_and_star_koords.begin();

     os << "Gap positions:" << std::endl;
     for (; git < gap_and_star_koords.end(); ++git)
       os << *git << std::endl;
  }


  unsigned getSeqCoord(const char *seq_pos)
  {
    return seq_pos - data.begin() + 1;
  }

  char get_ambiguity_character()
  {
    return general_ambig;
  }

  unsigned getSeqCoord_gap_correted(const char *seq_pos, unsigned &seq_gap_offset)
  {
    unsigned corrected_pos = seq_pos - data.begin()+1+seq_gap_offset;

    unsigned i = seq_gap_offset;
    unsigned N = gap_and_star_koords.size();

    //   std::cerr << "Entering getSeqCoord_gap_correted i: " << i << " N: " << N << " corr coord: " << corrected_pos << std::endl; 

    while (i < N && gap_and_star_koords.get_unckecked(i) < corrected_pos)
    {
      ++i;
      ++corrected_pos;
    }

    seq_gap_offset = i;

    //    std::cerr << "Exiting getSeqCoord_gap_correted i: " << i << " : " << N << " corr coord: " << corrected_pos << std::endl; 

    return corrected_pos;
  }


  void writeSequence(std::ostream &os, unsigned char_per_line)
  {
    char *pos = data.begin();
    char *end = data.end();
    char *tmp_end;

/*     if (bitcode_recoded) */
/*       backrecode_bitcode_sequence(); */

    os << ">" << full_name.c_str() << std::endl;
    //    os << ">" << full_name.c_str() << std::endl;

    tmp_end = pos + char_per_line;
    if (tmp_end > end)
      tmp_end = end;

    while (pos != end)
    {
      while (pos != tmp_end)
      {
	os << *pos;
	++pos;
      }
      os << std::endl;
      tmp_end = pos + char_per_line;
      if (tmp_end > end)
	tmp_end = end;
    }
  }

/*   void set_to_using_phylip_sequential_sequence() */
/*   { */
/*   } */

  void readRawFastaSequence(CFile& infile)
  {
    char     c   = '\0';

    full_name = "";
    data.clear();

    reset_numbers_of_residues();

    c = infile.getchar();
    while (!infile.fail() && isspace(c)) // Ignore spaces
      c = infile.getchar();
    while (!infile.fail() && c == '#')   // Ignore lines starting with #
    {
      infile.ignore('\n');
      c = infile.getchar();
      while (!infile.fail() && isspace(c) )   // Ignore spaces
	c = infile.getchar();
    }

    if (c != '>')          // Should be new sequence
    {
      infile.clear(CFile::__fail_flag);
      return;
    }
    infile.getline(full_name);
    if (infile.fail())
    {
      return;
    }
    find_shortname_and_description();

    c = infile.peekchar();
    while ( !infile.fail() && c != '>' )
    {
      //      cerr << data.capacity() << endl;
      getRawLine(infile);
      c = infile.peekchar();
    }
  }

  void readSeqName_ignore_Sequence_data(CFile& infile)
  {
    char     c   = '\0';

    full_name = "";
    data.clear();

    reset_numbers_of_residues();

    c = infile.getchar();
    while (!infile.fail() && isspace(c)) // Ignore spaces
      c = infile.getchar();
    while (!infile.fail() && c == '#')   // Ignore lines starting with #
    {
      infile.ignore('\n');
      c = infile.getchar();
      while (!infile.fail() && isspace(c) )   // Ignore spaces
	c = infile.getchar();
    }

    if (c != '>')          // Should be new sequence
    {
      infile.clear(CFile::__fail_flag);
      return;
    }
    infile.getline(full_name);
    if (infile.fail())
    {
      return;
    }
    find_shortname_and_description();

    // Ignore the sequence data of this sequnce:
    c = infile.peekchar();
    while ( !infile.fail() && c != '>' )
    {
      //      std::cerr << data.capacity() << std::endl;
      infile.ignore('\n');
      c = infile.peekchar();
    }
  }


  void readRawFastaSequence_toupper(CFile& infile)
  {
    char     c   = '\0';

    full_name = "";
    data.clear();

    reset_numbers_of_residues();

    c = infile.getchar();
    while (!infile.fail() && isspace(c))
      c = infile.getchar();
    while (!infile.fail() && c == '#')
    {
      infile.ignore('\n');
      c = infile.getchar();
      while (!infile.fail() && isspace(c) )
	c = infile.getchar();
    }

    if (c != '>')
    {
      infile.clear(CFile::__fail_flag);
      return;
    }
    infile.getline(full_name);
    if (infile.fail())
    {
      return;
    }
    find_shortname_and_description();

    c = infile.peekchar();
    while ( !infile.fail() && c != '>' )
    {
      //      std::cerr << data.capacity() << std::endl;
      getRawLine_toupper(infile);
      c = infile.peekchar();
    }
  }

  // it1 is the first position to fill. This coorindate must be 0 based.
  // it2 is the position behind the last position to fill. It must be 0 based.

  void fill_range_with(unsigned pos1, unsigned pos2, char c)
  {
    numbers_of_residues_computed  = false;

    char     *it1     = data.begin() + pos1;
    char     *it2     = data.begin() + pos2;
    char     *it_end  = data.end();

    if (it2 > it_end)
      it2 = it_end;

    while (it1  < it2)
    {
      *it1 = c;
      ++it1;
    }
  }


  // Called from Phobos:
  // TODO: I think we should not remove stars any more.
  // Stars should only be tolerated in protein sequences.
  void readFastaSequence_toupper_ambig2N_removegaps(CFile& infile)
  {
    char      c   = '\0';
    unsigned  pos = 0;

    full_name = "";
    data.clear();
    gap_and_star_koords.clear();
    gap_koords.clear();
    star_koords.clear();

    reset_numbers_of_residues();

    c = infile.getchar();
    while (!infile.fail() && isspace(c))
      c = infile.getchar();
    while (!infile.fail() && c == '#')
    {
      infile.ignore('\n');
      c = infile.getchar();
      while (!infile.fail() && isspace(c) )
	c = infile.getchar();
    }

    if (c != '>')
    {
      infile.clear(CFile::__fail_flag);
      return;
    }
    infile.getline(full_name);
    if (infile.fail())
    {
      return;
    }
    find_shortname_and_description();

    c = infile.peekchar();
    while ( !infile.fail() && c != '>' )
    {
      //      std::cerr << data.capacity() << std::endl;
      if (type == dna)
	getLineOfDNA_toupper_ambig2N_removegaps(infile, pos);
      else
	std::cout << "Not implemented. Use the readFastaSequence_toupper_removegaps(CFile& infile) function instead." << std::endl;
      c = infile.peekchar();
    }
  }


  void readFastaSequence_toupper_removegaps(CFile& infile)
  {
    char      c   = '\0';
    unsigned  pos = 0;

    full_name = "";
    data.clear();
    gap_and_star_koords.clear();
    gap_koords.clear();
    star_koords.clear();

    reset_numbers_of_residues();

    c = infile.getchar();
    while (!infile.fail() && isspace(c))
      c = infile.getchar();
    while (!infile.fail() && c == '#')
    {
      infile.ignore('\n');
      c = infile.getchar();
      while (!infile.fail() && isspace(c) )
	c = infile.getchar();
    }

    if (c != '>')
    {
      infile.clear(CFile::__fail_flag);
      return;
    }
    infile.getline(full_name);
    if (infile.fail())
    {
      return;
    }
    find_shortname_and_description();

    c = infile.peekchar();
    while ( !infile.fail() && c != '>' )
    {
      //      std::cerr << data.capacity() << std::endl;
      if (type == dna)
	getLineOfDNA_toupper_removegaps(infile, pos);
      else
	getLineOfProtein_toupper_removegaps(infile, pos);
      c = infile.peekchar();
    }
  }

  // Backward compatible function:
  // Called e.g. from Phobos
  void readFastaSequence_toupper_ambig2N_gaps2N(CFile& infile)
  {
    readFastaSequence_toupper_ambig2N_gaps2ambig(infile);
  }


  // Has resently been renamed: gaps2N -> gaps2ambig
  // since we want to prepare for reading proteins.
  void readFastaSequence_toupper_ambig2N_gaps2ambig(CFile& infile)
  {
    char     c   = '\0';

    full_name = "";
    data.clear();

    reset_numbers_of_residues();

    c = infile.getchar();
    while (!infile.fail() && isspace(c))
      c = infile.getchar();
    while (!infile.fail() && c == '#')
    {
      infile.ignore('\n');
      c = infile.getchar();
      while (!infile.fail() && isspace(c) )
	c = infile.getchar();
    }

    if (c != '>')
    {
      infile.clear(CFile::__fail_flag);
      return;
    }
    infile.getline(full_name);
    if (infile.fail())
    {
      return;
    }
    find_shortname_and_description();

    c = infile.peekchar();
    while ( !infile.fail() && c != '>' )
    {
      //      std::cerr << data.capacity() << std::endl;
      if (type == dna)
	getLineOfDNA_toupper_ambig2N_gaps2ambig(infile);
      else
	std::cout << "Not implemented." << std::endl;
      c = infile.peekchar();
    }
  }


  // chrashed in this function on linux - do not know why yet 64 bit problem??
  void check_allowed_symbols_in_sequence(const char *symbols)
  {
    unsigned pos=0;

    pos = data.find_first_not_of(symbols, pos);
    while (pos != faststring::npos)
    {
      std::cerr << "Found illegal symbol in sequnces: " << short_name 
		<< " Position " << pos << ". Found char " << data[pos] << std::endl;

      ++pos;
      pos = data.find_first_not_of(symbols, pos);
    }

  }


  void export_fasta()
  {
    

  }

  void export_phylip()
  {


  }

  void export_nexus()
  {


  }


  void readFastaSequence_generic(CFile& infile, processing_flag pflag)
  {
    char     c   = '\0';
    unsigned  pos = 0;
 
    // TODO: Prepare to remove gaps and/or stars:

    // Increase the readability:
    register bool   convert_toupper_bool  = (pflag & convert_toupper);   //
    register bool   convert_ambig2N_bool  = (pflag & convert_ambig2N);   //
    register bool   removegaps_bool       = (pflag & removegaps);        //
    register bool   removestars_bool      = (pflag & removestars);       //
    bool            gaps_to_ambig_bool    = (pflag & gaps_to_ambig);     //

    full_name = "";
    data.clear();

    reset_numbers_of_residues();

    c = infile.getchar();
    while (!infile.fail() && isspace(c))
      c = infile.getchar();
    while (!infile.fail() && c == '#')
    {
      infile.ignore('\n');
      c = infile.getchar();
      while (!infile.fail() && isspace(c) )
	c = infile.getchar();
    }

    if (c != '>')
    {
      infile.clear(CFile::__fail_flag);
      return;
    }
    infile.getline(full_name);
    if (infile.fail())
    {
      return;
    }
    find_shortname_and_description();

    c = infile.peekchar();
    while ( !infile.fail() && c != '>' )
    {
      //      std::cerr << data.capacity() << std::endl;
      if (type == dna)
	getLineOfDNA_generic(infile, pos, pflag);
      else if (type == protein)
	getLineOfProtein_generic(infile, pos, pflag);
      else if (type == auto_detect_type  || type == molecular)
      {
	getRawLine(infile);
	if (convert_toupper_bool)
	  toupper_this_sequence();
	auto_detect_datatype();

	if (convert_ambig2N_bool)
	  convert_ambig2N_this_sequence();
	if (gaps_to_ambig_bool)
	  gaps_to_ambig_this_sequence();
	if (removegaps_bool || removestars_bool)
	  remove_gaps_stars_this_sequence(removegaps_bool, removestars_bool);
      }
      c = infile.peekchar();
    }
  }

  unsigned get_number_of_Ns()
  {
    if (!numbers_of_residues_computed)
      compute_numbers_of_residues();
    return number_of_Ns;
  }

  unsigned get_number_of_gaps()
  {
    if (!numbers_of_residues_computed)
      compute_numbers_of_residues();
    return number_of_gaps;
  }

  unsigned get_number_raw_original_length()
  {
    if (!numbers_of_residues_computed)
      compute_numbers_of_residues();
    return number_raw_original_length;
  }

  unsigned get_number_of_DNARNA_bases()
  {
    if (!numbers_of_residues_computed)
      compute_numbers_of_residues();
    return number_of_DNARNA_bases;
  }

  unsigned get_number_of_AAs()
  {
    if (!numbers_of_residues_computed)
      compute_numbers_of_residues();
    return number_of_AAs;
  }

  unsigned get_number_of_DNARNA_ambigs()
  {
    return number_of_DNARNA_amgibs;
  }

  double get_CG_content()
  {
    //    std::cerr << "number_CG " << number_CG   << std::endl;
    //    std::cerr << "number_of_DNARNA_bases " << number_of_DNARNA_bases   << std::endl;
    //    //    std::cerr << "number_of_DNARNA_amgibs  " << number_of_DNARNA_amgibs << std::endl;

    return ((double) number_CG)/(number_of_DNARNA_bases);
  }


  bool range_contains_gaps_or_Ns(unsigned pos1, unsigned pos2)
  {
    unsigned len = length();

    if (pos1 >= pos2 || pos1 > len) // Empty range
      return false;
    if (pos2 > len)
      pos2 = len;

    const char *p1, *p2;
    p1 = getSeqBegin() + pos1;
    p2 = getSeqBegin() + pos2;

    char c='\0';

    while (p1 != p2)
    {
      c = *p1;
      if (c == 'N' || c == 'n' || c == '-')
	break;
      ++p1;
    }
    if (c == 'N' || c == 'n' || c == '-')      
      return true;
    else
      return false;
  }

};


#endif