rwfile.java

用TFIDF和特征增益两种方式实现了特征向量空间的建立

package yus.baseline;

import java.io.BufferedReader;
import java.io.File;
import java.io.FileOutputStream;
import java.io.FileReader;
import java.io.IOException;
import java.io.OutputStreamWriter;

public class RWFile {

	/**
	 * @param args
	 * @throws IOException
	 */
	public static void main(String[] args) throws IOException {
		// TODO Auto-generated method stub
		long startTime = System.currentTimeMillis();
		mergeFile();
		separateFile();
		long endTime = System.currentTimeMillis();
		System.out.println("The spending time is: "
				+ String.valueOf(endTime - startTime) + " ms");
	}

	private static void separateFile() throws IOException {
		// TODO Auto-generated method stub
		String spath = new File("").getCanonicalPath() + "/text/labeled/";
		String dpath = new File("").getCanonicalPath() + "/text/desFile/";
		String[] fileArray = { "H1_cla.txt", "H2_cla.txt", "H3_cla.txt",
				"H4_cla.txt", "H5_cla.txt", "H6_cla.txt", "H7_cla.txt",
				"H8_cla.txt", "H9_cla.txt", "H10_cla.txt" };
		int count = 0;

		for (int i = 0; i < fileArray.length; i++) {
			BufferedReader br = new BufferedReader(new FileReader(spath + "/"
					+ fileArray[i]));

			StringBuffer sb = new StringBuffer();
			String line = br.readLine();
			while (line != null) {
				if (!"TheEndOfThisTextOhYeah/nx".equals(line.trim())) {
					sb.append(line);
				} else {
					if (!"".equals(sb.toString().trim())) {
						count++;
						OutputStreamWriter osw = new OutputStreamWriter(
								new FileOutputStream(dpath
										+ String.valueOf(count) + ".txt"));
						osw.write(sb.toString());
						osw.flush();
						osw.close();
						sb = new StringBuffer();
					}
				}
				line = br.readLine();
			}
			br.close();

		}
	}

	private static void mergeFile() throws IOException {
		// TODO Auto-generated method stub
		String path = new File("").getCanonicalPath();

		String[] pathArray = { path + "/text/C000007", path + "/text/C000008",
				path + "/text/C000010", path + "/text/C000013",
				path + "/text/C000014", path + "/text/C000016",
				path + "/text/C000020", path + "/text/C000022",
				path + "/text/C000023", path + "/text/C000024" };
		String[] fileArray = { "H1.txt", "H2.txt", "H3.txt", "H4.txt",
				"H5.txt", "H6.txt", "H7.txt", "H8.txt", "H9.txt", "H10.txt" };

		for (int i = 0; i < pathArray.length; i++) {
			File file = new File(pathArray[i]);
			StringBuffer sb = new StringBuffer();
			if (file.isDirectory()) {
				String[] fileList = file.list();
				for (String fileName : fileList) {
					BufferedReader br2 = new BufferedReader(new FileReader(
							pathArray[i] + "/" + fileName));
					String line = br2.readLine();
					while (line != null) {
						sb.append(line);
						line = br2.readLine();
					}

					sb.append("\n\rTheEndOfThisTextOhYeah\n\r");
					br2.close();
				}
			}

			String filePath = path + "/text/waitLabel/" + fileArray[i];
			File file2 = new File(filePath);
			if (file2.exists()) {
				file2.delete();
			}
			OutputStreamWriter osw = new OutputStreamWriter(
					new FileOutputStream(filePath));
			osw.write(sb.toString());
			osw.flush();
			osw.close();

		}

	}

}