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.util.ArrayList; 58 import java.util.Arrays; 59 import java.util.Iterator; 60 import java.util.List; 61 62 /** 63 * The Gener object helps in the discarding of nodes which break the reduction effort and defend the 64 * structure against large reductions. 65 */ 66 public class Gener extends Reduce { 67 /** Constructor for the Gener object. */ Gener()68 public Gener() {} 69 70 /** 71 * Return a Trie with infrequent values occurring in the given Trie removed. 72 * 73 * @param orig the Trie to optimize 74 * @return a new optimized Trie 75 */ 76 @Override optimize(Trie orig)77 public Trie optimize(Trie orig) { 78 List<CharSequence> cmds = orig.cmds; 79 List<Row> rows = new ArrayList<>(); 80 List<Row> orows = orig.rows; 81 int[] remap = new int[orows.size()]; 82 83 Arrays.fill(remap, 1); 84 for (int j = orows.size() - 1; j >= 0; j--) { 85 if (eat(orows.get(j), remap)) { 86 remap[j] = 0; 87 } 88 } 89 90 Arrays.fill(remap, -1); 91 rows = removeGaps(orig.root, orows, new ArrayList<Row>(), remap); 92 93 return new Trie(orig.forward, remap[orig.root], cmds, rows); 94 } 95 96 /** 97 * Test whether the given Row of Cells in a Trie should be included in an optimized Trie. 98 * 99 * @param in the Row to test 100 * @param remap Description of the Parameter 101 * @return <code>true</code> if the Row should remain, <code>false 102 * </code> otherwise 103 */ eat(Row in, int[] remap)104 public boolean eat(Row in, int[] remap) { 105 int sum = 0; 106 for (Iterator<Cell> i = in.cells.values().iterator(); i.hasNext(); ) { 107 Cell c = i.next(); 108 sum += c.cnt; 109 if (c.ref >= 0) { 110 if (remap[c.ref] == 0) { 111 c.ref = -1; 112 } 113 } 114 } 115 int frame = sum / 10; 116 boolean live = false; 117 for (Iterator<Cell> i = in.cells.values().iterator(); i.hasNext(); ) { 118 Cell c = i.next(); 119 if (c.cnt < frame && c.cmd >= 0) { 120 c.cnt = 0; 121 c.cmd = -1; 122 } 123 if (c.cmd >= 0 || c.ref >= 0) { 124 live |= true; 125 } 126 } 127 return !live; 128 } 129 } 130