| 1 | /* |
| 2 | * This file is part of lanterna (http://code.google.com/p/lanterna/). |
| 3 | * |
| 4 | * lanterna is free software: you can redistribute it and/or modify |
| 5 | * it under the terms of the GNU Lesser General Public License as published by |
| 6 | * the Free Software Foundation, either version 3 of the License, or |
| 7 | * (at your option) any later version. |
| 8 | * |
| 9 | * This program is distributed in the hope that it will be useful, |
| 10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 12 | * GNU Lesser General Public License for more details. |
| 13 | * |
| 14 | * You should have received a copy of the GNU Lesser General Public License |
| 15 | * along with this program. If not, see <http://www.gnu.org/licenses/>. |
| 16 | * |
| 17 | * Copyright (C) 2010-2015 Martin |
| 18 | */ |
| 19 | package com.googlecode.lanterna; |
| 20 | |
| 21 | import java.util.ArrayList; |
| 22 | import java.util.Arrays; |
| 23 | import java.util.LinkedList; |
| 24 | import java.util.List; |
| 25 | |
| 26 | /** |
| 27 | * This class contains a number of utility methods for analyzing characters and |
| 28 | * strings in a terminal context. The main purpose is to make it easier to work |
| 29 | * with text that may or may not contain double-width text characters, such as |
| 30 | * CJK (Chinese, Japanese, Korean) and other special symbols. This class assumes |
| 31 | * those are all double-width and in case the terminal (-emulator) chooses to |
| 32 | * draw them (somehow) as single-column then all the calculations in this class |
| 33 | * will be wrong. It seems safe to assume what this class considers double-width |
| 34 | * really is taking up two columns though. |
| 35 | * |
| 36 | * @author Martin |
| 37 | */ |
| 38 | public class TerminalTextUtils { |
| 39 | private TerminalTextUtils() { |
| 40 | } |
| 41 | |
| 42 | /** |
| 43 | * Given a character, is this character considered to be a CJK character? |
| 44 | * Shamelessly stolen from <a href="http://stackoverflow.com/questions/1499804/how-can-i-detect-japanese-text-in-a-java-string" |
| 45 | * >StackOverflow</a> where it was contributed by user Rakesh N |
| 46 | * |
| 47 | * @param c |
| 48 | * Character to test |
| 49 | * @return {@code true} if the character is a CJK character |
| 50 | * |
| 51 | */ |
| 52 | public static boolean isCharCJK(final char c) { |
| 53 | Character.UnicodeBlock unicodeBlock = Character.UnicodeBlock.of(c); |
| 54 | return (unicodeBlock == Character.UnicodeBlock.HIRAGANA) |
| 55 | || (unicodeBlock == Character.UnicodeBlock.KATAKANA) |
| 56 | || (unicodeBlock == Character.UnicodeBlock.KATAKANA_PHONETIC_EXTENSIONS) |
| 57 | || (unicodeBlock == Character.UnicodeBlock.HANGUL_COMPATIBILITY_JAMO) |
| 58 | || (unicodeBlock == Character.UnicodeBlock.HANGUL_JAMO) |
| 59 | || (unicodeBlock == Character.UnicodeBlock.HANGUL_SYLLABLES) |
| 60 | || (unicodeBlock == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS) |
| 61 | || (unicodeBlock == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS_EXTENSION_A) |
| 62 | || (unicodeBlock == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS_EXTENSION_B) |
| 63 | || (unicodeBlock == Character.UnicodeBlock.CJK_COMPATIBILITY_FORMS) |
| 64 | || (unicodeBlock == Character.UnicodeBlock.CJK_COMPATIBILITY_IDEOGRAPHS) |
| 65 | || (unicodeBlock == Character.UnicodeBlock.CJK_RADICALS_SUPPLEMENT) |
| 66 | || (unicodeBlock == Character.UnicodeBlock.CJK_SYMBOLS_AND_PUNCTUATION) |
| 67 | || (unicodeBlock == Character.UnicodeBlock.ENCLOSED_CJK_LETTERS_AND_MONTHS) |
| 68 | || (unicodeBlock == Character.UnicodeBlock.HALFWIDTH_AND_FULLWIDTH_FORMS && c < 0xFF61); // The |
| 69 | // magic |
| 70 | // number |
| 71 | // here |
| 72 | // is |
| 73 | // the |
| 74 | // separating |
| 75 | // index |
| 76 | // between |
| 77 | // full-width |
| 78 | // and |
| 79 | // half-width |
| 80 | } |
| 81 | |
| 82 | /** |
| 83 | * Checks if a character is expected to be taking up two columns if printed |
| 84 | * to a terminal. This will generally be {@code true} for CJK (Chinese, |
| 85 | * Japanese and Korean) characters. |
| 86 | * |
| 87 | * @param c |
| 88 | * Character to test if it's double-width when printed to a |
| 89 | * terminal |
| 90 | * @return {@code true} if this character is expected to be taking up two |
| 91 | * columns when printed to the terminal, otherwise {@code false} |
| 92 | */ |
| 93 | public static boolean isCharDoubleWidth(final char c) { |
| 94 | return isCharCJK(c); |
| 95 | } |
| 96 | |
| 97 | /** |
| 98 | * @deprecated Call {@code getColumnWidth(s)} instead |
| 99 | */ |
| 100 | @Deprecated |
| 101 | public static int getTrueWidth(String s) { |
| 102 | return getColumnWidth(s); |
| 103 | } |
| 104 | |
| 105 | /** |
| 106 | * Given a string, returns how many columns this string would need to occupy |
| 107 | * in a terminal, taking into account that CJK characters takes up two |
| 108 | * columns. |
| 109 | * |
| 110 | * @param s |
| 111 | * String to check length |
| 112 | * @return Number of actual terminal columns the string would occupy |
| 113 | */ |
| 114 | public static int getColumnWidth(String s) { |
| 115 | return getColumnIndex(s, s.length()); |
| 116 | } |
| 117 | |
| 118 | /** |
| 119 | * Given a string and a character index inside that string, find out what |
| 120 | * the column index of that character would be if printed in a terminal. If |
| 121 | * the string only contains non-CJK characters then the returned value will |
| 122 | * be same as {@code stringCharacterIndex}, but if there are CJK characters |
| 123 | * the value will be different due to CJK characters taking up two columns |
| 124 | * in width. If the character at the index in the string is a CJK character |
| 125 | * itself, the returned value will be the index of the left-side of |
| 126 | * character. |
| 127 | * |
| 128 | * @param s |
| 129 | * String to translate the index from |
| 130 | * @param stringCharacterIndex |
| 131 | * Index within the string to get the terminal column index of |
| 132 | * @return Index of the character inside the String at {@code |
| 133 | * stringCharacterIndex} when it has been writted to a terminal |
| 134 | * @throws StringIndexOutOfBoundsException |
| 135 | * if the index given is outside the String length or negative |
| 136 | */ |
| 137 | public static int getColumnIndex(String s, int stringCharacterIndex) |
| 138 | throws StringIndexOutOfBoundsException { |
| 139 | int index = 0; |
| 140 | for (int i = 0; i < stringCharacterIndex; i++) { |
| 141 | if (isCharCJK(s.charAt(i))) { |
| 142 | index++; |
| 143 | } |
| 144 | index++; |
| 145 | } |
| 146 | return index; |
| 147 | } |
| 148 | |
| 149 | /** |
| 150 | * This method does the reverse of getColumnIndex, given a String and |
| 151 | * imagining it has been printed out to the top-left corner of a terminal, |
| 152 | * in the column specified by {@code columnIndex}, what is the index of that |
| 153 | * character in the string. If the string contains no CJK characters, this |
| 154 | * will always be the same as {@code columnIndex}. If the index specified is |
| 155 | * the right column of a CJK character, the index is the same as if the |
| 156 | * column was the left column. So calling {@code |
| 157 | * getStringCharacterIndex("英", 0)} and {@code getStringCharacterIndex("英", |
| 158 | * 1)} will both return 0. |
| 159 | * |
| 160 | * @param s |
| 161 | * String to translate the index to |
| 162 | * @param columnIndex |
| 163 | * Column index of the string written to a terminal |
| 164 | * @return The index in the string of the character in terminal column |
| 165 | * {@code columnIndex} |
| 166 | */ |
| 167 | public static int getStringCharacterIndex(String s, int columnIndex) { |
| 168 | int index = 0; |
| 169 | int counter = 0; |
| 170 | while (counter < columnIndex) { |
| 171 | if (isCharCJK(s.charAt(index++))) { |
| 172 | counter++; |
| 173 | if (counter == columnIndex) { |
| 174 | return index - 1; |
| 175 | } |
| 176 | } |
| 177 | counter++; |
| 178 | } |
| 179 | return index; |
| 180 | } |
| 181 | |
| 182 | /** |
| 183 | * Given a string that may or may not contain CJK characters, returns the |
| 184 | * substring which will fit inside <code>availableColumnSpace</code> |
| 185 | * columns. This method does not handle special cases like tab or new-line. |
| 186 | * <p> |
| 187 | * Calling this method is the same as calling {@code fitString(string, 0, |
| 188 | * availableColumnSpace)}. |
| 189 | * |
| 190 | * @param string |
| 191 | * The string to fit inside the availableColumnSpace |
| 192 | * @param availableColumnSpace |
| 193 | * Number of columns to fit the string inside |
| 194 | * @return The whole or part of the input string which will fit inside the |
| 195 | * supplied availableColumnSpace |
| 196 | */ |
| 197 | public static String fitString(String string, int availableColumnSpace) { |
| 198 | return fitString(string, 0, availableColumnSpace); |
| 199 | } |
| 200 | |
| 201 | /** |
| 202 | * Given a string that may or may not contain CJK characters, returns the |
| 203 | * substring which will fit inside <code>availableColumnSpace</code> |
| 204 | * columns. This method does not handle special cases like tab or new-line. |
| 205 | * <p> |
| 206 | * This overload has a {@code fromColumn} parameter that specified where |
| 207 | * inside the string to start fitting. Please notice that {@code fromColumn} |
| 208 | * is not a character index inside the string, but a column index as if the |
| 209 | * string has been printed from the left-most side of the terminal. So if |
| 210 | * the string is "日本語", fromColumn set to 1 will not starting counting from |
| 211 | * the second character ("本") in the string but from the CJK filler |
| 212 | * character belonging to "日". If you want to count from a particular |
| 213 | * character index inside the string, please pass in a substring and use |
| 214 | * fromColumn set to 0. |
| 215 | * |
| 216 | * @param string |
| 217 | * The string to fit inside the availableColumnSpace |
| 218 | * @param fromColumn |
| 219 | * From what column of the input string to start fitting (see |
| 220 | * description above!) |
| 221 | * @param availableColumnSpace |
| 222 | * Number of columns to fit the string inside |
| 223 | * @return The whole or part of the input string which will fit inside the |
| 224 | * supplied availableColumnSpace |
| 225 | */ |
| 226 | public static String fitString(String string, int fromColumn, |
| 227 | int availableColumnSpace) { |
| 228 | if (availableColumnSpace <= 0) { |
| 229 | return ""; |
| 230 | } |
| 231 | |
| 232 | StringBuilder bob = new StringBuilder(); |
| 233 | int column = 0; |
| 234 | int index = 0; |
| 235 | while (index < string.length() && column < fromColumn) { |
| 236 | char c = string.charAt(index++); |
| 237 | column += TerminalTextUtils.isCharCJK(c) ? 2 : 1; |
| 238 | } |
| 239 | if (column > fromColumn) { |
| 240 | bob.append(" "); |
| 241 | availableColumnSpace--; |
| 242 | } |
| 243 | |
| 244 | while (availableColumnSpace > 0 && index < string.length()) { |
| 245 | char c = string.charAt(index++); |
| 246 | availableColumnSpace -= TerminalTextUtils.isCharCJK(c) ? 2 : 1; |
| 247 | if (availableColumnSpace < 0) { |
| 248 | bob.append(' '); |
| 249 | } else { |
| 250 | bob.append(c); |
| 251 | } |
| 252 | } |
| 253 | return bob.toString(); |
| 254 | } |
| 255 | |
| 256 | /** |
| 257 | * This method will calculate word wrappings given a number of lines of text |
| 258 | * and how wide the text can be printed. The result is a list of new rows |
| 259 | * where word-wrapping was applied. |
| 260 | * |
| 261 | * @param maxWidth |
| 262 | * Maximum number of columns that can be used before |
| 263 | * word-wrapping is applied |
| 264 | * @param lines |
| 265 | * Input text |
| 266 | * @return The input text word-wrapped at {@code maxWidth}; this may contain |
| 267 | * more rows than the input text |
| 268 | */ |
| 269 | public static List<String> getWordWrappedText(int maxWidth, String... lines) { |
| 270 | List<String> result = new ArrayList<String>(); |
| 271 | LinkedList<String> linesToBeWrapped = new LinkedList<String>(Arrays |
| 272 | .asList(lines)); |
| 273 | while (!linesToBeWrapped.isEmpty()) { |
| 274 | String row = linesToBeWrapped.removeFirst(); |
| 275 | int rowWidth = getColumnWidth(row); |
| 276 | if (rowWidth <= maxWidth) { |
| 277 | result.add(row); |
| 278 | } else { |
| 279 | // Now search in reverse and find the first possible line-break |
| 280 | int characterIndex = getStringCharacterIndex(row, maxWidth); |
| 281 | while (!Character.isSpaceChar(row.charAt(characterIndex)) |
| 282 | && !isCharCJK(row.charAt(characterIndex)) |
| 283 | && characterIndex > 0) { |
| 284 | characterIndex--; |
| 285 | } |
| 286 | |
| 287 | if (characterIndex == 0) { |
| 288 | // Failed! There was no 'nice' place to cut so just cut it |
| 289 | // at maxWidth |
| 290 | result.add(row.substring(0, maxWidth)); |
| 291 | linesToBeWrapped.addFirst(row.substring(maxWidth)); |
| 292 | } else { |
| 293 | // Ok, split the row, add it to the result and continue |
| 294 | // processing the second half on a new line |
| 295 | result.add(row.substring(0, characterIndex)); |
| 296 | int spaceCharsToSkip = 0; |
| 297 | while (characterIndex < row.length() |
| 298 | && Character |
| 299 | .isSpaceChar(row.charAt(characterIndex))) { |
| 300 | characterIndex++; |
| 301 | } |
| 302 | ; |
| 303 | linesToBeWrapped.addFirst(row.substring(characterIndex)); |
| 304 | } |
| 305 | } |
| 306 | } |
| 307 | return result; |
| 308 | } |
| 309 | } |