Corpus.java

import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;

public class Corpus 
{
	private String corpusname = "";
	private String directory = "";
	private String period = "";
	private String filter = "";
	private int startdate = 0;
	private int enddate = 0;
	private ArrayList<String> allWordsOfCorpus = new ArrayList<String>();
	
	public Corpus(String corpusname, String period, String directory)
	{
		this.corpusname = corpusname;
		this.period = period;
		this.directory = directory;
		setCorpusDates();
		readCorpus();
	}
	
	public String getCorpusName()
	{
		return corpusname;
	}
	
	public String getPeriod()
	{
		return period;
	}
	
	public ArrayList<String> getCorpus()
	{		
		return allWordsOfCorpus;
	}
	
	public int getStartDate()
	{
		return startdate;
	}
	
	public int getEndDate()
	{
		return enddate;
	}
	
	// location directory of corpus
	private void setCorpusDates()
	{
		switch (corpusname) 
		{
			case "ppcmbe": 
				startdate = 1700;
				enddate = 1914;
				//directory = "C:\\Users\\Peukert\\Corpora\\PENN Corpus\\PENN-CORPORA\\PPCMBE-RELEASE-1\\corpus\\pos";
				break;
			case "ppceme": 
				startdate = 1500;
				enddate = 1710;
				//directory = "C:\\Users\\Peukert\\Corpora\\PENN Corpus\\PENN-CORPORA\\PPCEME-RELEASE-2\\corpus\\pos\\penn2";
				break;
			case "ppcme2":
				startdate = 1150;
				enddate = 1500;
				//directory = "C:\\Users\\Peukert\\Corpora\\PENN Corpus\\PENN-CORPORA\\PPCME2-RELEASE-3\\corpus\\pos";
				break;
		}
		//System.out.println(directory);		
	}
	
	// set time information, method to read which files,
	private void readCorpus()
	{
		IO io = new IO();
		if (period.isEmpty())
		{
			filter = ".*\\.pos";
			allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
		}
		else
		{
			switch (period)
			{	
				case "m1": 
					startdate = 1150;
					enddate = 1250;
					filter = "[a-z0-9]+\\.[m][x]?[1]\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "m2":
					startdate = 1250;
					enddate = 1350;
					filter = "[a-z0-9]+\\.[m][2]\\d?\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "m3":
					startdate = 1350;
					enddate = 1420;
					filter = "[a-z0-9]+\\.[m][3]\\d?\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "m4":
					startdate = 1420;
					enddate = 1500;
					filter = "[a-z0-9]+\\.[m][x]?[4]\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "e1":
					startdate = 1500;
					enddate = 1569;
					filter = "[a-z0-9]+-e1-p2\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "e2":
					startdate = 1570;
					enddate = 1639;
					filter = "[a-z0-9]+-e2-p2\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "e3":
					startdate = 1640;
					enddate = 1710;
					filter = "[a-z0-9]+-e3-p2\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "Emod1":
					startdate = 1700;
					enddate = 1769;
					filter = "[a-z0-9]+-[1][7][^789](\\d|[x]?)\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "Emod2":
					startdate = 1770;
					enddate = 1839;
					filter = "[a-z0-9]+-[1]([7][789]|[8][0123])(\\d|[x]?)\\.pos";
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
				case "Emod3":
					startdate = 1840;
					enddate = 1914;
					filter = "[a-z0-9]+-[1]([8][456789]|[9][01])(\\d|[x]?)\\.pos";;
					allWordsOfCorpus = io.readFilesFromDirectory(directory, filter);
					break;
			}
			normalizeText();
		}
	}
	
	
	
	private void normalizeText()
	{
		//+a for æ; +t for þ; +d for ð, +g for ȝ
		for (int i = 0; i < allWordsOfCorpus.size(); i++) 
		{
			String word = allWordsOfCorpus.get(i);
			//delete $-sign of manually corrected words (see corpus docu)
			if (word.startsWith("$"))
			{
				word = word.replaceFirst("\\$", "");
				//System.out.println(word);
			}
			// normalize text data (only applicable for m1-m4)
			if (word.matches(".*\\+[agdtAGDT][a-zA-Z0-9]*[_|/][A-Z]+") && corpusname.equals("ppcme2"))
			{
				word = word.replaceAll("\\+[tT]", "th");//þ
				word = word.replaceAll("\\+[gG]", "z");//ȝ
				word = word.replaceAll("\\+[dD]", "th");//ð
				word = word.replaceAll("\\+[aA]", "ae");//æ
				allWordsOfCorpus.add(word);
				//System.out.println(word);
				allWordsOfCorpus.remove(i);
			}
		}
	}
		
	/*
	 * delete the main method once programm is finished
	 * exist only for test purposes
	 */
//	public static void main(String[] args) 
//	{  
//		Corpus cp = new Corpus("ppcme2", "m1");
//		cp.getCorpus();
//	}
	
	
	
}