001/** 002 * Copyright (c) 2010, Regents of the University of Colorado 003 * All rights reserved. 004 * 005 * Redistribution and use in source and binary forms, with or without 006 * modification, are permitted provided that the following conditions are met: 007 * 008 * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 009 * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. 010 * Neither the name of the University of Colorado at Boulder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. 011 * 012 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" 013 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 014 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 015 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE 016 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 017 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 018 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 019 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 020 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 021 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 022 * POSSIBILITY OF SUCH DAMAGE. 023 */ 024package org.cleartk.feature; 025 026import java.util.Collections; 027import java.util.List; 028 029import org.apache.uima.jcas.JCas; 030import org.apache.uima.jcas.tcas.Annotation; 031import org.cleartk.ml.Feature; 032import org.cleartk.ml.feature.extractor.NamedFeatureExtractor1; 033 034/** 035 * <br> 036 * Copyright (c) 2010, Regents of the University of Colorado <br> 037 * All rights reserved. 038 * 039 * Extract a slice of the text covered by the annotation. Handles negative slice indices to make it 040 * easy to slice from the end of the string. 041 * 042 * @author Steven Bethard 043 */ 044public class TextSliceExtractor<T extends Annotation> implements NamedFeatureExtractor1<T> { 045 046 private int start; 047 048 private int stop; 049 050 private String featureName; 051 052 /** 053 * Create an extractor for a given slice of the text. E.g. 054 * <code>new TextSliceExtractor(1, -1)</code> would extract all of the text but its first and last 055 * characters. 056 * 057 * @param start 058 * The first character offset of the slice. If negative, it is assumed to count backwards 059 * from the end of the string. If the offset falls before the start of the string, the 060 * start of the string will be used instead. 061 * @param stop 062 * The last character offset of the slice. If negative, it is assumed to count backwards 063 * from the end of the string. If the offset falls after the end of the string, the end 064 * of the string will be used instead. 065 */ 066 public TextSliceExtractor(int start, int stop) { 067 this.start = start; 068 this.stop = stop; 069 this.featureName = "Suffix"; 070 } 071 072 /** 073 * Create an extractor for a slice of text from a single offset to the end of the string. E.g. 074 * <code>new TextSliceExtractor(-2)</code> would extract a suffix of length 2 from the text. 075 * 076 * @param start 077 * The first character offset of the slice. If negative, it is assumed to count backwards 078 * from the end of the string. If the offset falls before the start of the string, the 079 * start of the string will be used instead. 080 */ 081 public TextSliceExtractor(int start) { 082 this(start, Integer.MAX_VALUE); 083 } 084 085 @Override 086 public String getFeatureName() { 087 return this.featureName; 088 } 089 090 public List<Feature> extract(JCas view, T focusAnnotation) { 091 String text = focusAnnotation.getCoveredText(); 092 int startOffset = this.start; 093 if (startOffset < 0) { 094 startOffset += text.length(); 095 } 096 if (startOffset < 0) { 097 startOffset = 0; 098 } 099 int stopOffset = this.stop; 100 if (stopOffset < 0) { 101 stopOffset += text.length(); 102 } 103 if (stopOffset > text.length()) { 104 stopOffset = text.length(); 105 } 106 text = text.substring(startOffset, stopOffset); 107 return Collections.singletonList(new Feature(this.featureName, text)); 108 } 109 110}