1 /* 2 * Copyright (C) 2013 The Android Open Source Project 3 * 4 * Licensed under the Apache License, Version 2.0 (the "License"); 5 * you may not use this file except in compliance with the License. 6 * You may obtain a copy of the License at 7 * 8 * http://www.apache.org/licenses/LICENSE-2.0 9 * 10 * Unless required by applicable law or agreed to in writing, software 11 * distributed under the License is distributed on an "AS IS" BASIS, 12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 13 * See the License for the specific language governing permissions and 14 * limitations under the License. 15 */ 16 17 package com.android.inputmethod.latin.makedict; 18 19 import java.util.Random; 20 21 // Utility methods related with code points used for tests. 22 public class CodePointUtils { CodePointUtils()23 private CodePointUtils() { 24 // This utility class is not publicly instantiable. 25 } 26 27 public static final int[] LATIN_ALPHABETS_LOWER = { 28 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 29 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', 30 0x00E0 /* LATIN SMALL LETTER A WITH GRAVE */, 31 0x00E1 /* LATIN SMALL LETTER A WITH ACUTE */, 32 0x00E2 /* LATIN SMALL LETTER A WITH CIRCUMFLEX */, 33 0x00E3 /* LATIN SMALL LETTER A WITH TILDE */, 34 0x00E4 /* LATIN SMALL LETTER A WITH DIAERESIS */, 35 0x00E5 /* LATIN SMALL LETTER A WITH RING ABOVE */, 36 0x00E6 /* LATIN SMALL LETTER AE */, 37 0x00E7 /* LATIN SMALL LETTER C WITH CEDILLA */, 38 0x00E8 /* LATIN SMALL LETTER E WITH GRAVE */, 39 0x00E9 /* LATIN SMALL LETTER E WITH ACUTE */, 40 0x00EA /* LATIN SMALL LETTER E WITH CIRCUMFLEX */, 41 0x00EB /* LATIN SMALL LETTER E WITH DIAERESIS */, 42 0x00EC /* LATIN SMALL LETTER I WITH GRAVE */, 43 0x00ED /* LATIN SMALL LETTER I WITH ACUTE */, 44 0x00EE /* LATIN SMALL LETTER I WITH CIRCUMFLEX */, 45 0x00EF /* LATIN SMALL LETTER I WITH DIAERESIS */, 46 0x00F0 /* LATIN SMALL LETTER ETH */, 47 0x00F1 /* LATIN SMALL LETTER N WITH TILDE */, 48 0x00F2 /* LATIN SMALL LETTER O WITH GRAVE */, 49 0x00F3 /* LATIN SMALL LETTER O WITH ACUTE */, 50 0x00F4 /* LATIN SMALL LETTER O WITH CIRCUMFLEX */, 51 0x00F5 /* LATIN SMALL LETTER O WITH TILDE */, 52 0x00F6 /* LATIN SMALL LETTER O WITH DIAERESIS */, 53 0x00F7 /* LATIN SMALL LETTER O WITH STROKE */, 54 0x00F9 /* LATIN SMALL LETTER U WITH GRAVE */, 55 0x00FA /* LATIN SMALL LETTER U WITH ACUTE */, 56 0x00FB /* LATIN SMALL LETTER U WITH CIRCUMFLEX */, 57 0x00FC /* LATIN SMALL LETTER U WITH DIAERESIS */, 58 0x00FD /* LATIN SMALL LETTER Y WITH ACUTE */, 59 0x00FE /* LATIN SMALL LETTER THORN */, 60 0x00FF /* LATIN SMALL LETTER Y WITH DIAERESIS */ 61 }; 62 generateCodePointSet(final int codePointSetSize, final Random random)63 public static int[] generateCodePointSet(final int codePointSetSize, final Random random) { 64 final int[] codePointSet = new int[codePointSetSize]; 65 for (int i = codePointSet.length - 1; i >= 0; ) { 66 final int r = Math.abs(random.nextInt()); 67 if (r < 0) continue; 68 // Don't insert 0~0x20, but insert any other code point. 69 // Code points are in the range 0~0x10FFFF. 70 final int candidateCodePoint = 0x20 + r % (Character.MAX_CODE_POINT - 0x20); 71 // Code points between MIN_ and MAX_SURROGATE are not valid on their own. 72 if (candidateCodePoint >= Character.MIN_SURROGATE 73 && candidateCodePoint <= Character.MAX_SURROGATE) continue; 74 codePointSet[i] = candidateCodePoint; 75 --i; 76 } 77 return codePointSet; 78 } 79 80 /** 81 * Generates a random word. 82 */ generateWord(final Random random, final int[] codePointSet)83 public static String generateWord(final Random random, final int[] codePointSet) { 84 StringBuilder builder = new StringBuilder(); 85 // 8 * 4 = 32 chars max, but we do it the following way so as to bias the random toward 86 // longer words. This should be closer to natural language, and more importantly, it will 87 // exercise the algorithms in dicttool much more. 88 final int count = 1 + (Math.abs(random.nextInt()) % 5) 89 + (Math.abs(random.nextInt()) % 5) 90 + (Math.abs(random.nextInt()) % 5) 91 + (Math.abs(random.nextInt()) % 5) 92 + (Math.abs(random.nextInt()) % 5) 93 + (Math.abs(random.nextInt()) % 5) 94 + (Math.abs(random.nextInt()) % 5) 95 + (Math.abs(random.nextInt()) % 5); 96 while (builder.length() < count) { 97 builder.appendCodePoint(codePointSet[Math.abs(random.nextInt()) % codePointSet.length]); 98 } 99 return builder.toString(); 100 } 101 } 102