/*
 * UniprotParser.java
 */
package org.ngbw.utils;


import java.io.IOException;
import java.io.InputStream;
import java.io.Reader;
import java.util.regex.Matcher;
import java.util.regex.Pattern;


/**
 * 
 * @author Paul Hoover
 *
 */
class UniprotParser extends Parser {

	private static final Pattern m_idPattern = Pattern.compile("^ID\\s+.*$");
	private static final Pattern m_accessionPattern = Pattern.compile("^AC\\s+.*$");
	private static final Pattern m_datePattern = Pattern.compile("^DT\\s+.*$");
	private static final Pattern m_versionPattern = Pattern.compile(".*entry version\\s+(\\d+).*");
	private static final Pattern m_descriptionPattern = Pattern.compile("^DE\\s+.*$");
	private static final Pattern m_organismPattern = Pattern.compile("^OS\\s+.*$");
	private static final Pattern m_recordEndPattern = Pattern.compile("^\\s*//\\s*$");
	private static final Pattern m_filterPattern = Pattern.compile("^CC\\s+-+$");


	/**
	 * 
	 * @param input
	 */
	public UniprotParser(String input)
	{
		super(input, m_filterPattern);
	}

	/**
	 * 
	 * @param input
	 */
	public UniprotParser(byte[] input)
	{
		super(input, m_filterPattern);
	}

	/**
	 * 
	 * @param input
	 */
	public UniprotParser(InputStream input)
	{
		super(input, m_filterPattern);
	}

	/**
	 * 
	 * @param input
	 */
	public UniprotParser(Reader input)
	{
		super(input, m_filterPattern);
	}


	/**
	 * 
	 * @return
	 * @throws IOException
	 */
	public SequenceRecord nextRecord() throws IOException
	{
		if (!findFirstLine(m_idPattern))
			return null;

		SequenceRecord record = new SequenceRecord();

		parseIdentity(record);

		while (true) {
			if (m_line == null)
				return null;

			if (m_accessionPattern.matcher(m_line).matches())
				parseAccession(record);
			else if (m_datePattern.matcher(m_line).matches())
				parseDate(record);
			else if (m_descriptionPattern.matcher(m_line).matches())
				parseDescription(record);
			else if (m_organismPattern.matcher(m_line).matches())
				parseOrganism(record);
			else if (m_recordEndPattern.matcher(m_line).matches()) {
				record.completeSource = getCompleteSource();
				record.filteredSource = getFilteredSource();

				break;
			}
			else
				readAndStoreLine();
		}

		return record;
	}


	/**
	 * 
	 * @param record
	 * @throws IOException
	 */
	private void parseIdentity(SequenceRecord record) throws IOException
	{
		String[] subStrings = m_line.substring(3).trim().split("\\s+");

		record.alternativeId = subStrings[0];

		readAndStoreLine();
	}

	/**
	 * 
	 * @param record
	 * @throws IOException
	 */
	private void parseAccession(SequenceRecord record) throws IOException
	{
		String[] subStrings = m_line.substring(3).split(";\\s*");

		record.primaryId = subStrings[0].trim();

		while (true) {
			readAndStoreLine();

			if (m_line == null || !m_accessionPattern.matcher(m_line).matches())
				break;
		}
	}

	/**
	 * 
	 * @param record
	 * @throws IOException
	 */
	private void parseDate(SequenceRecord record) throws IOException
	{
		while (true) {
			Matcher matcher = m_versionPattern.matcher(m_line);

			if (matcher.matches())
				record.version = matcher.group(1);

			readAndStoreLine();

			if (m_line == null || !m_datePattern.matcher(m_line).matches())
				break;
		}
	}

	/**
	 * 
	 * @param record
	 * @throws IOException
	 */
	private void parseDescription(SequenceRecord record) throws IOException
	{
		StringBuilder field = new StringBuilder(m_line.substring(3).trim());

		while (true) {
			readAndStoreLine();

			if (m_line == null || !m_descriptionPattern.matcher(m_line).matches())
				break;

			field.append(' ');
			field.append(m_line.substring(3).trim());
		}

		record.name = field.toString();
	}

	/**
	 * 
	 * @param record
	 * @throws IOException
	 */
	private void parseOrganism(SequenceRecord record) throws IOException
	{
		record.organism = m_line.substring(3).trim();

		readAndStoreLine();
	}
}