// ReadAlignment.java
//
// (c) 1999-2001 PAL Development Core Team
//
// This package may be distributed under the
// terms of the Lesser GNU General Public License (LGPL)


package pal.alignment;

import java.io.*;
import java.util.*;

import pal.io.*;
import pal.misc.*;


/**
 * reads aligned sequence data from plain text files.<p>
 *
 * recognizes PHYLIP 3.4 INTERLEAVED,
 *              PHYLIP SEQUENTIAL,
 *              CLUSTAL and derived formats.<p>
 *
 * Other features:
 * - the dot as "copy character" is recognized,
 * - all base characters are capitalized,
 * - automatic data type estimation
 * - determination of corresponding base frequencies.
 *
 * @version $Id: ReadAlignment.java,v 1.7 2003/09/04 03:22:34 matt Exp $
 *
 * @author Korbinian Strimmer
 * @author Alexei Drummond
 * @author Alex Moore a.d.moore@ex.ac.uk
 */
public class ReadAlignment extends AbstractAlignment
{
	//
	// Public stuff
	//

	/** read from stream */
	public ReadAlignment(PushbackReader input)
		throws AlignmentParseException, IOException
	{
		readFile(input);
	}

	/** read from file */
	public ReadAlignment(String file)
		throws AlignmentParseException, IOException
	{
		PushbackReader input = InputSource.openFile(file);
		readFile(input);
		input.close();
	}


	// Implementation of abstract Alignment method

	/** sequence alignment at (sequence, site) */
	public char getData(int seq, int site)
	{
		return data[seq][site];
	}


	//
	// Private stuff
	//

	private int lineLength;
	private Vector names, seqs, sites;
	private FormattedInput fi = FormattedInput.getInstance();

	// Raw sequence alignment [sequence][site]
	private char[][] data = null;
	private final boolean isType(PushbackReader in, String id) throws IOException {

		for (int i = 0; i < id.length(); i++)	{
			int c = fi.readNextChar(in);
			if (c != id.charAt(i) )	{
				in.unread(c);
				return false;
			}
		}
		return true;

	}

	private void readFile(PushbackReader in) throws AlignmentParseException, IOException	{
		if (isType(in,"CLUSTAL")) {
			fi.nextLine(in);
			readCLUSTALW(in);
		} else if(isType(in,"#NEXUS")||isType(in,"#nexus")) {
			readNEXUS(in);
		}	else {
			readPHYLIP(in);
		}
		// Capitalize
		for (int i = 0; i < numSeqs; i++){
			for (int j = 0; j < numSites; j++){
				data[i][j] = Character.toUpperCase(data[i][j]);
			}
		}
		// Estimate data type
		guessDataType();
	}

	private void readNEXUS(PushbackReader in) throws AlignmentParseException {

		NexusTokenizer ntp = null;
		try {
			// construct and set up parser
			ntp = new NexusTokenizer(in);
			ntp.setReadWhiteSpace(false);
			ntp.setWordModification(NexusTokenizer.WORD_UPPERCASE);
			ntp.setIgnoreComments(true);

			// vars
			boolean seqsOK = false;
			boolean sitesOK = false;
			boolean isInterleaved = false;

			String nToken = ntp.readToken();

			while(!(seqsOK && sitesOK)) {
				if(nToken.equals("NCHAR")) {
					ntp.readToken();	// =
					numSites = Integer.parseInt(ntp.readToken());
					sitesOK = true;
				}
				else if(nToken.equals("NTAX")) {
					ntp.readToken();	// =
					numSeqs = Integer.parseInt(ntp.readToken());
					seqsOK = true;
				}

				nToken = ntp.readToken();
			}

			// move on down to MATRIX, looking for INTERLEAVE as we go
			while(!nToken.equals("MATRIX")) {
				if(nToken.equals("INTERLEAVE")) {
					ntp.readToken();
					nToken = ntp.readToken();
					if(nToken.equals("YES")) {
						isInterleaved = true;
					}
				}
				nToken = ntp.readToken();
			}

			// create a data array to hold taxa:sequence
			String[][] d = new String[numSeqs][2];	// 0 = id, 1 = sequence
			for(int i = 0; i < d.length; i++) {
				d[i][1] = new String();
			}
			nToken = ntp.seek(NexusTokenizer.WORD_TOKEN);
			int taxaNumber = 0;


			while(!nToken.equals(";")) {
				d[taxaNumber][0] = nToken;	// insert name
				System.out.println("Read id: " + nToken);
				nToken = ntp.readToken();

				while(ntp.getLastTokenType() != NexusTokenizer.NEWLINE_TOKEN) {
					d[taxaNumber][1] += nToken;
					nToken = ntp.readToken();
				}

				taxaNumber++;
				if(taxaNumber == numSeqs & isInterleaved) {
					taxaNumber = 0;
					nToken = ntp.readToken();
					if(ntp.getLastTokenType() == NexusTokenizer.NEWLINE_TOKEN) {
						nToken = ntp.readToken();
					}
				}
				else {
					nToken = ntp.readToken();
				}
			}

			idGroup = new SimpleIdGroup(numSeqs);
			data = new char[numSeqs][numSites];

			for(int i = 0; i < d.length; i++) {
				idGroup.setIdentifier(i, new Identifier(d[i][0]));
				for(int j = 0; j < d[i][1].length(); j++) {
					if(d[i][1].length() != numSites) {
						throw new AlignmentParseException("Taxa " + d[i][0] + " sequence length not equal to " + numSites);
					}
					data[i][j] = d[i][1].charAt(j);
					if(i == 0 & data[i][j] == '.') {
						throw new AlignmentParseException(
								"Copy character (.) in first sequence not allowed @ pos " + j);
					}
				}
			}
		}
		catch(IOException e) {
			throw new AlignmentParseException("I/O error after row:" + ntp.getRow() + ", col:" + ntp.getCol());
		}
		catch(NumberFormatException e) {
			throw new AlignmentParseException("Number parse error - row:" + ntp.getRow() + ", col:" + ntp.getCol());
		}
		catch(NexusParseException e) {
			throw new AlignmentParseException("Parse error row:" + ntp.getRow() + ", col:" + ntp.getCol());
		}
	}


	// Read alignment (in CLUSTAL W format)
	private void readCLUSTALW(PushbackReader in)
		throws AlignmentParseException
	{

		int EOF = -1;
		int c, seq = 0, pos = 0;
		try	{
			// Find start block
			c = fi.readNextChar(in);
			in.unread(c);
			names = new Vector();		seqs = new Vector();

			// Reading first data block
			c = in.read();
			while (!Character.isWhitespace((char)c))	{
				in.unread(c);
				names.addElement(fi.readLabel(in, 10));
				readSeqLineC(in, seq, pos);
				seq++;
				c = in.read();
			}
			in.unread(c);
			// Skip CLUSTAL W status line
			fi.nextLine(in);

			pos += lineLength;
			numSeqs = seq;

			// Go to next block
			c = fi.readNextChar(in);
			in.unread(c);

			// Reading remaining blocks
			while (c != EOF)
			{
				for (seq = 0; seq < numSeqs; seq++)
				{
					// goto next blank
					do
					{
						c = in.read();
						if(c<0) { throw new AlignmentParseException("Unexpected end of file exception!"); }
					}
					while (Character.isWhitespace((char)c));
					readSeqLineC(in, seq, pos);

				}

				// Skip CLUSTAL W status line
				fi.nextLine(in);

				pos += lineLength;

				// Go to next block
				c = fi.readNextChar(in);
				in.unread(c);
			}

			numSites = pos;

			// Copy to array
			idGroup = new SimpleIdGroup(numSeqs);
			data = new char[numSeqs][numSites];
			for (int i = 0; i < numSeqs; i++)
			{
				idGroup.setIdentifier(i,
									new Identifier((String)names.elementAt(i)));
			}
			for (int i = 0; i < numSeqs; i++)
			{
				for (int j = 0; j < numSites; j++)
				{
					data[i][j] =
						((Character)
							((Vector) seqs.elementAt(i)
						).elementAt(j)).charValue();
				}
			}

			// Help garbage collector
			names = null;
			for (int i = 0; i < numSeqs; i++)
			{
				((Vector) seqs.elementAt(i)).removeAllElements();
			}
			seqs = null;
		}
			catch (IOException e)
			{
				throw new AlignmentParseException("IO error after pos. " + (pos + 1) + ", seq. " + (seq + 1));
				}
	}


	private void readSeqLineC(PushbackReader in, int s, int pos)
		throws IOException, AlignmentParseException
	{

		int c;

		if (pos == 0)
		{
			sites = new Vector();
			seqs.addElement(sites);
		}
		else
		{
			sites = (Vector) seqs.elementAt(s);
		}

		if (s == 0)
		{
			String thisLine = fi.readLine(in, false);
			lineLength = thisLine.length();

			for (int i = 0; i < lineLength; i++)
			{
				c = thisLine.charAt(i);
				if (c == '.')
				{
					throw new AlignmentParseException("Copy character (.) in first sequence not allowed (pos. "
					+ (i + pos + 1) + ")");
				}
				sites.addElement(new Character((char) c));
			}
		}
		else
		{
			for (int i = 0; i < lineLength; i++)
			{
				c = fi.readNextChar(in);
				if (c == '.')
				{
					c = ((Character)
							((Vector) seqs.elementAt(0)
						).elementAt(pos + i)).charValue();
				}
				sites.addElement(new Character((char) c));
			}
			fi.nextLine(in);
		}
	}

	// Read alignment (in PHYLIP 3.4 INTERLEAVED or PHYLIP SEQUENTIAL format)
	private void readPHYLIP(PushbackReader in)
		throws AlignmentParseException
	{

		int c, pos = 0, seq = 0;

		try
		{
			// Parse PHYLIP header line
			numSeqs = fi.readInt(in);
			numSites = fi.readInt(in);

			// Reserve memory
			idGroup = new SimpleIdGroup(numSeqs);
			data = new char[numSeqs][numSites];


			// Determine whether sequences are in INTERLEAVED
			// or in sequential format
			String header = fi.readLine(in, false);

			boolean interleaved = true;

			if (header.length() > 0)
			{
				if (header.charAt(0) == 'S')
				{
					interleaved = false;
				}
			}

			if (interleaved) // PHYLIP INTERLEAVED
			{
				//System.out.println("PHYLIP INTERLEAVED");


				// Reading data
				while (pos < numSites)
				{
					// Go to next block
					c = fi.readNextChar(in);
					in.unread(c);

					for (seq = 0; seq < numSeqs; seq++)
					{
						readSeqLineP(in, seq, pos, numSites);
					}
					pos += lineLength;
				}
			}
			else // PHYLIP SEQUENTIAL
			{
				//System.out.println("PHYLIP SEQUENTIAL");

				for (seq = 0; seq < numSeqs; seq++)
				{
					// Go to next block
					c = fi.readNextChar(in);
					in.unread(c);

					// Read label
					idGroup.setIdentifier(seq, new Identifier(fi.readLabel(in, 10)));

					// Read sequences
					for (pos = 0; pos < numSites; pos++)
					{
						data[seq][pos] = (char) fi.readNextChar(in);

						if (data[0][pos] == '.')
						{
							if (seq == 0)
							{
								throw new AlignmentParseException(
								"Copy character (.) in first sequence not allowed (pos. "
								+ (pos + 1) + ")");
							}
							else
							{
								data[seq][pos] = data[0][pos];
							}
						}
					}
				}
			}
		}
			catch (IOException e)
			{
				throw new AlignmentParseException("IO error after pos. " + (pos + 1) + ", seq. " + (seq + 1));
			}
	}

	private void readSeqLineP(PushbackReader in, int s, int pos, int maxPos)
		throws IOException, AlignmentParseException
	{

		if (pos == 0)
		{
			idGroup.setIdentifier(s, new Identifier(fi.readLabel(in, 10)));
		}

		if (s == 0)
		{
			String thisLine = fi.readLine(in, false);

			if (thisLine.length() > maxPos - pos)
			{
				lineLength = maxPos - pos;
			}
			else
			{
				lineLength = thisLine.length();
			}

			for (int i = 0; i < lineLength; i++)
			{
				data[0][pos + i] = thisLine.charAt(i);
				if (data[0][pos + i] == '.')
				{
					throw new AlignmentParseException("Copy character (.) in first sequence not allowed (pos. "
					+ (i + pos + 1) + ")");
				}
			}
		}
		else
		{
			for (int i = 0; i < lineLength; i++)
			{
				data[s][pos + i] = (char) fi.readNextChar(in);
				if (data[s][pos + i] == '.')
				{
					data[s][pos + i] = data[0][pos + i];
				}
			}
			fi.nextLine(in);
		}
	}
}