Package opennlp.tools.formats

Source Code of opennlp.tools.formats.EvalitaNameSampleStream

/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements.  See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 opennlp.tools.formats;

import java.io.IOException;
import java.io.InputStream;
import java.io.PrintStream;
import java.io.UnsupportedEncodingException;
import java.util.ArrayList;
import java.util.List;

import opennlp.tools.namefind.NameSample;
import opennlp.tools.util.InputStreamFactory;
import opennlp.tools.util.InvalidFormatException;
import opennlp.tools.util.ObjectStream;
import opennlp.tools.util.PlainTextByLineStream;
import opennlp.tools.util.Span;
import opennlp.tools.util.StringUtil;

/**
* Parser for the Italian NER training files of the Evalita 2007 and 2009 NER  shared tasks.
* <p>
* The data does not contain article boundaries,
* adaptive data will be cleared for every sentence.
* <p>
* Named Entities are annotated in the IOB2 format (as used in CoNLL 2002 shared task)
* <p>
* The Named Entity tag consists of two parts:
* 1. The  IOB2 tag: 'B'  (for 'begin')  denotes the  first token  of a
*    Named Entity,  I (for 'inside')  is used for  all other tokens  in a
*    Named Entity, and 'O' (for 'outside') is used for all other words;
* 2. The Entity  type tag: PER  (for Person), ORG  (for Organization),
*    GPE (for Geo-Political Entity), or LOC (for Location).
* <p>
* Each file  consists of four  columns separated by a  blank, containing
* respectively the  token, the Elsnet  PoS-tag, the Adige news  story to
* which the token belongs, and the Named Entity tag.
* <p>
* Data can be found on this web site:<br>
* http://www.evalita.it
* <p>
* <b>Note:</b> Do not use this class, internal use only!
*/
public class EvalitaNameSampleStream implements ObjectStream<NameSample>{

  public enum LANGUAGE {
      IT
    }

    public static final int GENERATE_PERSON_ENTITIES = 0x01;
    public static final int GENERATE_ORGANIZATION_ENTITIES = 0x01 << 1;
    public static final int GENERATE_LOCATION_ENTITIES = 0x01 << 2;
    public static final int GENERATE_GPE_ENTITIES = 0x01 << 3;

    public static final String DOCSTART = "-DOCSTART-";

    private final LANGUAGE lang;
    private final ObjectStream<String> lineStream;

    private final int types;

    public EvalitaNameSampleStream(LANGUAGE lang, ObjectStream<String> lineStream, int types) {
      this.lang = lang;
      this.lineStream = lineStream;
      this.types = types;
    }

    public EvalitaNameSampleStream(LANGUAGE lang, InputStreamFactory in, int types) throws IOException {
      this.lang = lang;
      try {
        this.lineStream = new PlainTextByLineStream(in, "UTF-8");
        System.setOut(new PrintStream(System.out, true, "UTF-8"));
      } catch (UnsupportedEncodingException e) {
        // UTF-8 is available on all JVMs, will never happen
        throw new IllegalStateException(e);
      }
      this.types = types;
    }

  /**
   * @param lang the language of the Evalita data file
   * @param in an Input Stream to read data.
   * @param types the types of the entities which are included in the Name Sample stream
   */
  @Deprecated
  public EvalitaNameSampleStream(LANGUAGE lang, InputStream in, int types) {

    this.lang = lang;
    try {
      this.lineStream = new PlainTextByLineStream(in, "UTF-8");
      System.setOut(new PrintStream(System.out, true, "UTF-8"));
    } catch (UnsupportedEncodingException e) {
      // UTF-8 is available on all JVMs, will never happen
      throw new IllegalStateException(e);
    }
    this.types = types;
  }

  static final Span extract(int begin, int end, String beginTag) throws InvalidFormatException {

    String type = beginTag.substring(2);

    if ("PER".equals(type)) {
      type = "person";
    }
    else if ("LOC".equals(type)) {
      type = "location";
    }
    else if ("GPE".equals(type)) {
      type = "gpe";
    }
    else if ("ORG".equals(type)) {
      type = "organization";
    }
    else {
      throw new InvalidFormatException("Unknown type: " + type);
    }

    return new Span(begin, end, type);
  }


  public NameSample read() throws IOException {

    List<String> sentence = new ArrayList<String>();
    List<String> tags = new ArrayList<String>();

    boolean isClearAdaptiveData = false;

    // Empty line indicates end of sentence

    String line;
    while ((line = lineStream.read()) != null && !StringUtil.isEmpty(line)) {

      if (line.startsWith(DOCSTART)) {
        isClearAdaptiveData = true;
        String emptyLine = lineStream.read();

        if (!StringUtil.isEmpty(emptyLine))
            throw new IOException("Empty line after -DOCSTART- not empty: '" + emptyLine +"'!");

        continue;
      }

      String fields[] = line.split(" ");

     // For Italian: WORD  POS-TAG SC-TAG NE-TAG
      if (LANGUAGE.IT.equals(lang) && (fields.length == 4)) {
        sentence.add(fields[0]);
        tags.add(fields[3]); // 3 is NE-TAG
      }
      else {
          throw new IOException("Incorrect number of fields per line for language: '" + line + "'!");
        }
    }

    // Always clear adaptive data for Italian
    if (LANGUAGE.IT.equals(lang))
      isClearAdaptiveData = true;

    if (sentence.size() > 0) {

      // convert name tags into spans
      List<Span> names = new ArrayList<Span>();

      int beginIndex = -1;
      int endIndex = -1;
      for (int i = 0; i < tags.size(); i++) {

        String tag = tags.get(i);

        if (tag.endsWith("PER") && (types & GENERATE_PERSON_ENTITIES) == 0)
          tag = "O";

        if (tag.endsWith("ORG") && (types & GENERATE_ORGANIZATION_ENTITIES) == 0)
          tag = "O";

        if (tag.endsWith("LOC") && (types & GENERATE_LOCATION_ENTITIES) == 0)
          tag = "O";

        if (tag.endsWith("GPE") && (types & GENERATE_GPE_ENTITIES) == 0)
          tag = "O";

        if (tag.startsWith("B-")) {

            if (beginIndex != -1) {
              names.add(extract(beginIndex, endIndex, tags.get(beginIndex)));
              beginIndex = -1;
              endIndex = -1;
            }

            beginIndex = i;
            endIndex = i +1;
          }
          else if (tag.startsWith("I-")) {
            endIndex++;
          }
          else if (tag.equals("O")) {
            if (beginIndex != -1) {
              names.add(extract(beginIndex, endIndex, tags.get(beginIndex)));
              beginIndex = -1;
              endIndex = -1;
            }
          }
          else {
            throw new IOException("Invalid tag: " + tag);
          }
        }

        // if one span remains, create it here
        if (beginIndex != -1)
          names.add(extract(beginIndex, endIndex, tags.get(beginIndex)));

        return new NameSample(sentence.toArray(new String[sentence.size()]), names.toArray(new Span[names.size()]), isClearAdaptiveData);
      }
      else if (line != null) {
        // Just filter out empty events, if two lines in a row are empty
        return read();
      }
      else {
        // source stream is not returning anymore lines
        return null;
      }
    }

  public void reset() throws IOException, UnsupportedOperationException {
    lineStream.reset();
  }

  public void close() throws IOException {
    lineStream.close();
  }
}
TOP

Related Classes of opennlp.tools.formats.EvalitaNameSampleStream

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.