Package edu.umd.cloud9.collection.wikipedia

Source Code of edu.umd.cloud9.collection.wikipedia.WikipediaPagesBz2InputStream

/*
* Cloud9: A MapReduce Library for Hadoop
*
* Licensed under the Apache License, Version 2.0 (the "License"); you
* may not use this file except in compliance with the License. You may
* obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
* implied. See the License for the specific language governing
* permissions and limitations under the License.
*/

package edu.umd.cloud9.collection.wikipedia;

import java.io.BufferedReader;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStreamReader;

import org.apache.commons.cli.CommandLine;
import org.apache.commons.cli.CommandLineParser;
import org.apache.commons.cli.GnuParser;
import org.apache.commons.cli.HelpFormatter;
import org.apache.commons.cli.OptionBuilder;
import org.apache.commons.cli.Options;
import org.apache.commons.cli.ParseException;
import org.apache.hadoop.util.ToolRunner;
import org.apache.tools.bzip2.CBZip2InputStream;

import edu.umd.cloud9.collection.wikipedia.language.WikipediaPageFactory;

/**
* Class for working with bz2-compressed Wikipedia article dump files on local disk.
*
* @author Jimmy Lin
* @author Peter Exner
*/
public class WikipediaPagesBz2InputStream {
  private static int DEFAULT_STRINGBUFFER_CAPACITY = 1024;

  private BufferedReader br;
  private FileInputStream fis;

  /**
   * Creates an input stream for reading Wikipedia articles from a bz2-compressed dump file.
   *
   * @param file path to dump file
   * @throws IOException
   */
  public WikipediaPagesBz2InputStream(String file) throws IOException {
    br = null;
    fis = new FileInputStream(file);
    byte[] ignoreBytes = new byte[2];
    fis.read(ignoreBytes); // "B", "Z" bytes from commandline tools
    br = new BufferedReader(new InputStreamReader(new CBZip2InputStream(fis)));

  }

  /**
   * Reads the next Wikipedia page.
   *
   * @param page WikipediaPage object to read into
   * @return <code>true</code> if page is successfully read
   * @throws IOException
   */
  public boolean readNext(WikipediaPage page) throws IOException {
    String s = null;
    StringBuffer sb = new StringBuffer(DEFAULT_STRINGBUFFER_CAPACITY);

    while ((s = br.readLine()) != null) {
      if (s.endsWith("<page>"))
        break;
    }

    if (s == null) {
      fis.close();
      br.close();
      return false;
    }

    sb.append(s + "\n");

    while ((s = br.readLine()) != null) {
      sb.append(s + "\n");

      if (s.endsWith("</page>"))
        break;
    }

    WikipediaPage.readPage(page, sb.toString());

    return true;
  }

  private static final String INPUT_OPTION = "input";
  private static final String LANGUAGE_OPTION = "output";

  @SuppressWarnings("static-access")
  public static void main(String[] args) throws Exception {
    Options options = new Options();
    options.addOption(OptionBuilder.withArgName("path").hasArg()
        .withDescription("gzipped XML dump file").create(INPUT_OPTION));
    options.addOption(OptionBuilder.withArgName("lang").hasArg()
        .withDescription("output location").create(LANGUAGE_OPTION));

    CommandLine cmdline = null;
    CommandLineParser parser = new GnuParser();
    try {
      cmdline = parser.parse(options, args);
    } catch (ParseException exp) {
      System.err.println("Error parsing command line: " + exp.getMessage());
      System.exit(-1);
    }

    if (!cmdline.hasOption(INPUT_OPTION)) {
      HelpFormatter formatter = new HelpFormatter();
      formatter.printHelp(WikipediaPagesBz2InputStream.class.getCanonicalName(), options);
      ToolRunner.printGenericCommandUsage(System.out);
      System.exit(-1);
    }

    String path = cmdline.getOptionValue(INPUT_OPTION);
    String lang = cmdline.hasOption(LANGUAGE_OPTION) ? cmdline.getOptionValue(LANGUAGE_OPTION)
        : "en";
    WikipediaPage p = WikipediaPageFactory.createWikipediaPage(lang);

    WikipediaPagesBz2InputStream stream = new WikipediaPagesBz2InputStream(path);
    while (stream.readNext(p)) {
      System.out.println(p.getTitle() + "\t" + p.getDocid());
    }
  }
}
TOP

Related Classes of edu.umd.cloud9.collection.wikipedia.WikipediaPagesBz2InputStream

TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.