#ifndef ALIGNMENT_H
#define ALIGNMENT_H

#include "Alphabet/Alphabet.h"
#include "AlphabetTuple/AlphabetTuple.h"
#include <iostream>


class alignment {
 private:
    
    void alloc(void);	// allocate arrays based on num_seqs and width

    void alloc_for_a2m(istream &in);
	// read the a2m file from in, retaining only width and num_seqs
	// then call alloc to allocate space for second pass read.

    inline void clear(void)	// set all pointers to 0
    {	alignmentID = NULL;
	Filename = NULL;
        names = NULL;
	data=NULL;
	weights=NULL;
	insert_before=NULL;
	
	num_seqs=0;	width=0;
    };
    
    void delete_all(void);	// remove all arrays, and clear pointers
    
 public:
    const Alphabet *alphabet;
    
    char *Filename;
    char *alignmentID;	// just up to first ; of ID field
    int width;	// number of characters per sequence
    int num_seqs;	// number of sequences
        
    char **names;	// array of num_seqs names
    Base **data;	// array of sequences	data[num_seqs][width]
    char **insert_before; // non-zero if there is an insertion before 
			// data[i][j] (array has width+1 positions)
    float *weights;	// array of num_seqs sequence weights
    
    alignment(void) {clear();}
    ~alignment(void) {delete_all();}
    
    void read_a2m(const char *full_filename, const AlphabetTuple *alpha);	
        // open an read SAM-style alignment
	// (a FASTA file with . and lowercase in insert positions)
	// only the match positions are used, insertions are just counted.


    inline double total_weight(void) const
    {   if (!weights) return 0;
	double total=0.;
	for (int i=num_seqs-1; i>=0; i--)
		total += weights[i];
	return total;
    }
    
    // report the weights in a format suitable for SAM 
    void report_weights(ostream &out) const;
    
    inline int alphabet_size(void) const
    {    return alphabet->norm_length();
    }

    // Returns a newly allocated copy of the first sequence of data.
    // This sequence is taken to be the alignemnts guide sequence.
    Base *first_sequence(void) const;
};

// CHANGE LOG
// 20 July 1997 Kevin Karplus
//	borrowed and rewritten from estimate-dist, keeping only a2m format.
// 21 April 2004 Sol Katzman
//      Changed first_sequence to Base* from char*
// 24 May 2004 Kevin Karplus
//	inlined simple functions

#endif