package jalview.io;

import jalview.analysis.SequenceIdMatcher;
import jalview.datamodel.AlignmentAnnotation;
import jalview.datamodel.AlignmentI;
import jalview.datamodel.Annotation;
import jalview.datamodel.SequenceI;

import java.awt.Color;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileReader;
import java.io.IOException;
import java.io.Reader;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;

/**
 * A file parse for T-Coffee score ascii format. This file contains the alignment consensus 
 * for each resude in any sequence.
 * <p>
 * This file is procuded by <code>t_coffee</code> providing the option 
 * <code>-output=score_ascii </code> to the program command line
 * 
 * An example file is the following 
 * 
 * <pre>
 * T-COFFEE, Version_9.02.r1228 (2012-02-16 18:15:12 - Revision 1228 - Build 336)
 * Cedric Notredame 
 * CPU TIME:0 sec.
 * SCORE=90
 * *
 *  BAD AVG GOOD
 * *
 * 1PHT   :  89
 * 1BB9   :  90
 * 1UHC   :  94
 * 1YCS   :  94
 * 1OOT   :  93
 * 1ABO   :  94
 * 1FYN   :  94
 * 1QCF   :  94
 * cons   :  90
 * 
 * 1PHT   999999999999999999999999998762112222543211112134
 * 1BB9   99999999999999999999999999987-------4322----2234
 * 1UHC   99999999999999999999999999987-------5321----2246
 * 1YCS   99999999999999999999999999986-------4321----1-35
 * 1OOT   999999999999999999999999999861-------3------1135
 * 1ABO   99999999999999999999999999986-------422-------34
 * 1FYN   99999999999999999999999999985-------32--------35
 * 1QCF   99999999999999999999999999974-------2---------24
 * cons   999999999999999999999999999851000110321100001134
 * 
 * 
 * 1PHT   ----------5666642367889999999999889
 * 1BB9   1111111111676653-355679999999999889
 * 1UHC   ----------788774--66789999999999889
 * 1YCS   ----------78777--356789999999999889
 * 1OOT   ----------78877--356789999999997-67
 * 1ABO   ----------687774--56779999999999889
 * 1FYN   ----------6888842356789999999999889
 * 1QCF   ----------6878742356789999999999889
 * cons   00100000006877641356789999999999889
 * </pre>
 * 
 * 
 * @author Paolo Di Tommaso
 *
 */
public class TCoffeeScoreFile {
	
	/** The {@link Header} structure holder */
	Header header;
	
	/** 
	 * Holds the consensues values for each sequences. It uses a LinkedHashMap to maintaint the 
	 * insertion order. 
	 */
	LinkedHashMap<String,StringBuilder> scores = new LinkedHashMap<String,StringBuilder>();

	Integer fWidth;

	/**
	 * Parse the specified file.
	 * 
	 * @param file The file to be parsed 
	 */
	public static TCoffeeScoreFile load(File file) {
		try {
			return load(new FileReader(file));
		} 
		catch (FileNotFoundException e) {
			throw new RuntimeException(e);
		}
	}
	
	/**
	 * Parse the provided reader for the T-Coffee scores file format
	 * 
	 * @param reader 
	 */
	public static TCoffeeScoreFile load(Reader reader) {

		try {
			BufferedReader in = (BufferedReader) (reader instanceof BufferedReader ? reader : new BufferedReader(reader));
			TCoffeeScoreFile result = new TCoffeeScoreFile();
			result.doParsing(in);
			return result.header != null && result.scores != null ? result : null;
		}
		catch( Exception e) {
			throw new RuntimeException(e);
		}
	}
		
	/**
	 * @return The 'height' of the score matrix i.e. the numbers of score rows that should matches 
	 * the number of sequences in the alignment
	 */
	public int getHeight() {
		// the last entry will always be the 'global' alingment consensus scores, so it is removed 
		// from the 'height' count to make this value compatible with the number of sequences in the MSA
		return scores != null && scores.size() > 0 ? scores.size()-1 : 0;
	}
	
	/**
	 * @return The 'width' of the score matrix i.e. the number of columns. 
	 * Since teh score value are supposd to be calculated for an 'aligned' MSA, all the entries 
	 * have to have the same width.  
	 */
	public int getWidth() {
		return fWidth != null ? fWidth : 0;
	}
	
	/**
	 * The default constructor is marked as {@code protected} since this class is meant to created 
	 * through the {@link #load(File)} or {@link #load(Reader)} factory methods
	 */
	protected TCoffeeScoreFile() { } 
	
	/**
	 * Get the string of score values for the specified seqeunce ID. 
	 * @param id The sequence ID 
	 * @return The scores as a string of values e.g. {@code 99999987-------432}. 
	 * 	It return an empty string when the specified ID is missing. 
	 */
	public String getScoresFor( String id ) {
		return scores.containsKey(id) ? scores.get(id).toString() : "";
	}
	
	/**
	 * @return The list of score string as a {@link List} object, in the same ordeer of the insertion i.e. in the MSA
	 */
	public List<String> getScoresList() {
		List<String> result = new ArrayList<String>( scores.size() );
		for( Map.Entry<String,StringBuilder> it : scores.entrySet() ) {
			result.add(it.getValue().toString());
		}
		
		return result;
	}
	
	/**
	 * @return The parsed score values a matrix of bytes
	 */
	public byte[][] getScoresArray() { 
		byte[][] result = new byte[ scores.size() ][];
		
		int rowCount = 0;
		for( Map.Entry<String,StringBuilder> it : scores.entrySet() ) {
			String line = it.getValue().toString();
			byte[] seqValues = new byte[ line.length() ];
			for( int j=0, c=line.length(); j<c; j++ ) {
				
				byte val = (byte)(line.charAt(j) - '0');

				seqValues[j] = ( val >= 0 && val <= 9 ) ? val : -1; 
			}

			result[rowCount++] = seqValues;
		}
		
		return result;
	}
	

	private void doParsing(BufferedReader in) throws IOException {

		/*
		 * read the header
		 */
		header = readHeader(in);

		if( header == null ) { return; }
		
	
		/*
		 * initilize the structure
		 */
		for( Map.Entry<String,Integer> entry : header.scores.entrySet() ) {
			scores.put( entry.getKey(), new StringBuilder());
		}
		
		/*
		 * go with the reading
		 */
		Block block;
		while( (block = readBlock(in, header.scores.size())) != null  ) {
			
			/*
			 * append sequences read in the block
			 */
			for( Map.Entry<String,String> entry : block.items.entrySet() ) {
				StringBuilder scoreStringBuilder = scores.get(entry.getKey());
				if( scoreStringBuilder == null ) {
					throw new RuntimeException(String.format("Invalid T-Coffee score file: Sequence ID '%s' is not declared in header section", entry.getKey()));
				}
				
				scoreStringBuilder.append( entry.getValue() );
			}
		}
		
		/*
		 * verify that all rows have the same width
		 */
		for( StringBuilder str : scores.values() ) {
			if( fWidth == null ) {
				fWidth = str.length();
			}
			else if( fWidth != str.length() ) {
				throw new RuntimeException("Invalid T-Coffee score file: All the score sequences must have the same length");
			}
		}
		
		
		
	}


	static int parseInt( String str ) {
		try {
			return Integer.parseInt(str);
		}
		catch( NumberFormatException e ) {
			// TODO report a warning ?
			return 0;
		}		
	}
	
	/**
	 * Reaad the header section in the T-Coffee score file format 
	 * 
	 * @param reader The scores reader 
	 * @return The parser {@link Header} instance 
	 * @throws RuntimeException when the header is not in the expected format
	 */
	static Header readHeader(BufferedReader reader) {
		
		Header result = null;
		try {
			result = new Header();
			result.head = reader.readLine();
			
			String line;

			while( (line = reader.readLine()) != null ) {
				if( line.startsWith("SCORE=")) {
					result.score = parseInt( line.substring(6).trim() );
					break;
				}
			}

			if( (line=reader.readLine())==null || !"*".equals(line.trim())) return null;
			if( (line=reader.readLine())==null || !"BAD AVG GOOD".equals(line.trim())) return null;
			if( (line=reader.readLine())==null || !"*".equals(line.trim())) return null;
			
			/*
			 * now are expected a list if sequences ID up to the first blank line
			 */
			while( (line=reader.readLine()) != null ) {
				if( "".equals(line) ) {
					break;
				}
				
				int p = line.indexOf(":");
				if( p == -1 ) {
					// TODO report a warning
					continue;
				}
				
				String id = line.substring(0,p).trim();
				int val = parseInt(line.substring(p+1).trim());
				if( "".equals(id) ) {
					// TODO report warning
					continue;
				}
				
				result.scores.put(id,val);
			}
			
		}
		catch( IOException e ) {
			throw new RuntimeException("Cannot parse T-Coffee score ascii file", e);
		}
		
		return result;
	} 
	
	/**
	 * Read a scores block ihe provided stream. 
	 * 
	 * @param reader The stream to parse
	 * @param size The expected number of the sequence to be read 
	 * @return The {@link Block} instance read or {link null} null if the end of file has reached.
	 * @throws IOException Something went wrong on the 'wire' 
	 */
	static Block readBlock( BufferedReader reader, int size ) throws IOException {
		Block result = new Block(size);
		String line;
		
		/*
		 * read blank lines (eventually)
		 */
		while( (line=reader.readLine()) != null && "".equals(line.trim())) {
			// consume blank lines 
		}
		
		if( line == null ) return null;
		
		/*
		 * read the scores block
		 */
		do {
			if( "".equals(line.trim()) ) {
				// terminated
				break;
			}
			
			// split the line on the first blank 
			// the first part have to contain the sequence id
			// theramining part are the scores values
			int p = line.indexOf(" ");
			if( p == -1 ) {
				//TODO This is an unexpected condition, log a warning or throw an exception ? 
				continue;
			} 
			
			String id = line.substring(0,p).trim();
			String val = line.substring(p+1).trim();
			
			result.items.put(id, val);
			
		} while( (line = reader.readLine()) != null ); 
		

		return result;
	}

	/*
	 * The score file header 
	 */
	static class Header {
		String head;
		int score;

		LinkedHashMap<String,Integer> scores = new LinkedHashMap<String,Integer>();
		
		public int getScoreAvg() { return score; }
		
		public int getScoreFor( String ID ) { 

			return scores.containsKey(ID) ? scores.get(ID) : -1;
	
		}
	}
	
	/*
	 * Hold a single block values block in the score file
	 */
	static class Block {
		int size;
		Map<String,String> items;
		
		public Block( int size ) {
			this.size = size;
			this.items = new HashMap<String,String>(size);
		} 
	
		String getScoresFor( String id ) {
			return items.get(id);
		}
		
		String getConsensus() {
			return items.get("cons");
		}
	}
	/**
	 * TCOFFEE score colourscheme
	 */
	static final Color[] colors = {
	                new Color( 102, 102, 255 ),     // #6666FF
	                new Color( 0, 255, 0),          // #00FF00
	                new Color( 102, 255, 0),        // #66FF00
	                new Color( 204, 255, 0),        // #CCFF00
	                new Color( 255, 255, 0),        // #FFFF00
	                new Color( 255, 204, 0),        // #FFCC00
	                new Color( 255, 153, 0),        // #FF9900
	                new Color( 255, 102, 0),        // #FF6600
	                new Color( 255, 51, 0),         // #FF3300
	                new Color( 255, 34, 0)          // #FF2000
	        };
	public final static String TCOFFEE_SCORE="TCoffeeScore";
	/**
	 * generate annotation for this TCoffee score set on the given alignment
	 * @param al alignment to annotate
	 * @param matchids if true, annotate sequences based on matching sequence names
	 * @return true if alignment annotation was modified, false otherwise.
	 */
	public boolean annotateAlignment(AlignmentI al, boolean matchids)
	{
	  boolean added=false;
	  int i=0;
	  SequenceIdMatcher sidmatcher = new SequenceIdMatcher(al.getSequencesArray());
	  byte[][] scoreMatrix=getScoresArray();
	  // for 2.8 - we locate any existing TCoffee annotation and remove it first before adding this.
	  for (Map.Entry<String,StringBuilder> id:scores.entrySet())
	  {
	    byte[] srow=scoreMatrix[i];
	    SequenceI s;
	    if (matchids)
	    {
	      s=sidmatcher.findIdMatch(id.getKey());
	    } else {
	      s=al.getSequenceAt(i);
	    }
	    i++;
            if (s==null && i!=scores.size() && !id.getKey().equals("cons"))
            {
              System.err.println("No "+(matchids ? "match ":" sequences left ")+" for TCoffee score set : "+id.getKey());
              continue;
            }
            int jSize=al.getWidth()< srow.length ? al.getWidth() : srow.length;
	    Annotation[] annotations=new Annotation[al.getWidth()];
	    for (int j=0;j<jSize;j++) {
	      byte val = srow[j];
	      annotations[j]=new Annotation(s==null ? ""+val:null,s==null ? ""+val:null,(char) val,val*1f,val >= 0 && val < colors.length ? colors[val] : Color.white);
	    }
	    AlignmentAnnotation aa=null;
	    if (s!=null)
	    {
	      // TODO - set per sequence score
	      aa=new AlignmentAnnotation(TCOFFEE_SCORE, "Score for "+id.getKey(), annotations);
	      
	      aa.setSequenceRef(s);
	      aa.visible=false;
	      aa.belowAlignment=false;
	    } else {
	      aa=new AlignmentAnnotation("T-COFFEE", "TCoffee column reliability score", annotations);
              aa.belowAlignment=true;
	      aa.visible=true;
	      
	    }
	    al.addAnnotation(aa);
	    added=true;
	  }
	  return added;
	}
	  

}