1 /* 2 Egothor Software License version 1.00 3 Copyright (C) 1997-2004 Leo Galambos. 4 Copyright (C) 2002-2004 "Egothor developers" 5 on behalf of the Egothor Project. 6 All rights reserved. 7 8 This software is copyrighted by the "Egothor developers". If this 9 license applies to a single file or document, the "Egothor developers" 10 are the people or entities mentioned as copyright holders in that file 11 or document. If this license applies to the Egothor project as a 12 whole, the copyright holders are the people or entities mentioned in 13 the file CREDITS. This file can be found in the same location as this 14 license in the distribution. 15 16 Redistribution and use in source and binary forms, with or without 17 modification, are permitted provided that the following conditions are 18 met: 19 1. Redistributions of source code must retain the above copyright 20 notice, the list of contributors, this list of conditions, and the 21 following disclaimer. 22 2. Redistributions in binary form must reproduce the above copyright 23 notice, the list of contributors, this list of conditions, and the 24 disclaimer that follows these conditions in the documentation 25 and/or other materials provided with the distribution. 26 3. The name "Egothor" must not be used to endorse or promote products 27 derived from this software without prior written permission. For 28 written permission, please contact Leo.G@seznam.cz 29 4. Products derived from this software may not be called "Egothor", 30 nor may "Egothor" appear in their name, without prior written 31 permission from Leo.G@seznam.cz. 32 33 In addition, we request that you include in the end-user documentation 34 provided with the redistribution and/or in the software itself an 35 acknowledgement equivalent to the following: 36 "This product includes software developed by the Egothor Project. 37 http://egothor.sf.net/" 38 39 THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED 40 WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 41 MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 42 IN NO EVENT SHALL THE EGOTHOR PROJECT OR ITS CONTRIBUTORS BE LIABLE 43 FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 44 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 45 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR 46 BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, 47 WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE 48 OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN 49 IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 50 51 This software consists of voluntary contributions made by many 52 individuals on behalf of the Egothor Project and was originally 53 created by Leo Galambos (Leo.G@seznam.cz). 54 */ 55 package org.egothor.stemmer; 56 57 import java.io.LineNumberReader; 58 import java.nio.charset.Charset; 59 import java.nio.file.Files; 60 import java.nio.file.Paths; 61 import java.util.Locale; 62 import java.util.StringTokenizer; 63 import org.apache.lucene.util.SuppressForbidden; 64 65 /** The DiffIt class is a means generate patch commands from an already prepared stemmer table. */ 66 public class DiffIt { 67 68 /** no instantiation */ DiffIt()69 private DiffIt() {} 70 get(int i, String s)71 static int get(int i, String s) { 72 try { 73 return Integer.parseInt(s.substring(i, i + 1)); 74 } catch ( 75 @SuppressWarnings("unused") 76 Exception x) { 77 return 1; 78 } 79 } 80 81 /** 82 * Entry point to the DiffIt application. 83 * 84 * <p>This application takes one argument, the path to a file containing a stemmer table. The 85 * program reads the file and generates the patch commands for the stems. 86 * 87 * @param args the path to a file containing a stemmer table 88 */ 89 @SuppressForbidden(reason = "System.out required: command line tool") main(java.lang.String[] args)90 public static void main(java.lang.String[] args) throws Exception { 91 92 int ins = get(0, args[0]); 93 int del = get(1, args[0]); 94 int rep = get(2, args[0]); 95 int nop = get(3, args[0]); 96 97 for (int i = 1; i < args.length; i++) { 98 // System.out.println("[" + args[i] + "]"); 99 Diff diff = new Diff(ins, del, rep, nop); 100 String charset = System.getProperty("egothor.stemmer.charset", "UTF-8"); 101 try (LineNumberReader in = 102 new LineNumberReader( 103 Files.newBufferedReader(Paths.get(args[i]), Charset.forName(charset)))) { 104 for (String line = in.readLine(); line != null; line = in.readLine()) { 105 try { 106 line = line.toLowerCase(Locale.ROOT); 107 StringTokenizer st = new StringTokenizer(line); 108 String stem = st.nextToken(); 109 System.out.println(stem + " -a"); 110 while (st.hasMoreTokens()) { 111 String token = st.nextToken(); 112 if (token.equals(stem) == false) { 113 System.out.println(stem + " " + diff.exec(token, stem)); 114 } 115 } 116 } catch ( 117 @SuppressWarnings("unused") 118 java.util.NoSuchElementException x) { 119 // no base token (stem) on a line 120 } 121 } 122 } 123 } 124 } 125 } 126