| 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 strings in a terminal context. The main |
| 28 | * purpose is to make it easier to work with text that may or may not contain double-width text characters, such as CJK |
| 29 | * (Chinese, Japanese, Korean) and other special symbols. This class assumes those are all double-width and in case the |
| 30 | * terminal (-emulator) chooses to draw them (somehow) as single-column then all the calculations in this class will be |
| 31 | * wrong. It seems safe to assume what this class considers double-width really is taking up two columns though. |
| 32 | * |
| 33 | * @author Martin |
| 34 | */ |
| 35 | public class TerminalTextUtils { |
| 36 | private TerminalTextUtils() { |
| 37 | } |
| 38 | |
| 39 | /** |
| 40 | * Given a character, is this character considered to be a CJK character? |
| 41 | * Shamelessly stolen from |
| 42 | * <a href="http://stackoverflow.com/questions/1499804/how-can-i-detect-japanese-text-in-a-java-string">StackOverflow</a> |
| 43 | * where it was contributed by user Rakesh N |
| 44 | * @param c Character to test |
| 45 | * @return {@code true} if the character is a CJK character |
| 46 | * |
| 47 | */ |
| 48 | public static boolean isCharCJK(final char c) { |
| 49 | Character.UnicodeBlock unicodeBlock = Character.UnicodeBlock.of(c); |
| 50 | return (unicodeBlock == Character.UnicodeBlock.HIRAGANA) |
| 51 | || (unicodeBlock == Character.UnicodeBlock.KATAKANA) |
| 52 | || (unicodeBlock == Character.UnicodeBlock.KATAKANA_PHONETIC_EXTENSIONS) |
| 53 | || (unicodeBlock == Character.UnicodeBlock.HANGUL_COMPATIBILITY_JAMO) |
| 54 | || (unicodeBlock == Character.UnicodeBlock.HANGUL_JAMO) |
| 55 | || (unicodeBlock == Character.UnicodeBlock.HANGUL_SYLLABLES) |
| 56 | || (unicodeBlock == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS) |
| 57 | || (unicodeBlock == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS_EXTENSION_A) |
| 58 | || (unicodeBlock == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS_EXTENSION_B) |
| 59 | || (unicodeBlock == Character.UnicodeBlock.CJK_COMPATIBILITY_FORMS) |
| 60 | || (unicodeBlock == Character.UnicodeBlock.CJK_COMPATIBILITY_IDEOGRAPHS) |
| 61 | || (unicodeBlock == Character.UnicodeBlock.CJK_RADICALS_SUPPLEMENT) |
| 62 | || (unicodeBlock == Character.UnicodeBlock.CJK_SYMBOLS_AND_PUNCTUATION) |
| 63 | || (unicodeBlock == Character.UnicodeBlock.ENCLOSED_CJK_LETTERS_AND_MONTHS) |
| 64 | || (unicodeBlock == Character.UnicodeBlock.HALFWIDTH_AND_FULLWIDTH_FORMS && c < 0xFF61); //The magic number here is the separating index between full-width and half-width |
| 65 | } |
| 66 | |
| 67 | /** |
| 68 | * Checks if a character is expected to be taking up two columns if printed to a terminal. This will generally be |
| 69 | * {@code true} for CJK (Chinese, Japanese and Korean) characters. |
| 70 | * @param c Character to test if it's double-width when printed to a terminal |
| 71 | * @return {@code true} if this character is expected to be taking up two columns when printed to the terminal, |
| 72 | * otherwise {@code false} |
| 73 | */ |
| 74 | public static boolean isCharDoubleWidth(final char c) { |
| 75 | return isCharCJK(c); |
| 76 | } |
| 77 | |
| 78 | /** |
| 79 | * @deprecated Call {@code getColumnWidth(s)} instead |
| 80 | */ |
| 81 | @Deprecated |
| 82 | public static int getTrueWidth(String s) { |
| 83 | return getColumnWidth(s); |
| 84 | } |
| 85 | |
| 86 | /** |
| 87 | * Given a string, returns how many columns this string would need to occupy in a terminal, taking into account that |
| 88 | * CJK characters takes up two columns. |
| 89 | * @param s String to check length |
| 90 | * @return Number of actual terminal columns the string would occupy |
| 91 | */ |
| 92 | public static int getColumnWidth(String s) { |
| 93 | return getColumnIndex(s, s.length()); |
| 94 | } |
| 95 | |
| 96 | /** |
| 97 | * Given a string and a character index inside that string, find out what the column index of that character would |
| 98 | * be if printed in a terminal. If the string only contains non-CJK characters then the returned value will be same |
| 99 | * as {@code stringCharacterIndex}, but if there are CJK characters the value will be different due to CJK |
| 100 | * characters taking up two columns in width. If the character at the index in the string is a CJK character itself, |
| 101 | * the returned value will be the index of the left-side of character. |
| 102 | * @param s String to translate the index from |
| 103 | * @param stringCharacterIndex Index within the string to get the terminal column index of |
| 104 | * @return Index of the character inside the String at {@code stringCharacterIndex} when it has been writted to a |
| 105 | * terminal |
| 106 | * @throws StringIndexOutOfBoundsException if the index given is outside the String length or negative |
| 107 | */ |
| 108 | public static int getColumnIndex(String s, int stringCharacterIndex) throws StringIndexOutOfBoundsException { |
| 109 | int index = 0; |
| 110 | for(int i = 0; i < stringCharacterIndex; i++) { |
| 111 | if(isCharCJK(s.charAt(i))) { |
| 112 | index++; |
| 113 | } |
| 114 | index++; |
| 115 | } |
| 116 | return index; |
| 117 | } |
| 118 | |
| 119 | /** |
| 120 | * This method does the reverse of getColumnIndex, given a String and imagining it has been printed out to the |
| 121 | * top-left corner of a terminal, in the column specified by {@code columnIndex}, what is the index of that |
| 122 | * character in the string. If the string contains no CJK characters, this will always be the same as |
| 123 | * {@code columnIndex}. If the index specified is the right column of a CJK character, the index is the same as if |
| 124 | * the column was the left column. So calling {@code getStringCharacterIndex("英", 0)} and |
| 125 | * {@code getStringCharacterIndex("英", 1)} will both return 0. |
| 126 | * @param s String to translate the index to |
| 127 | * @param columnIndex Column index of the string written to a terminal |
| 128 | * @return The index in the string of the character in terminal column {@code columnIndex} |
| 129 | */ |
| 130 | public static int getStringCharacterIndex(String s, int columnIndex) { |
| 131 | int index = 0; |
| 132 | int counter = 0; |
| 133 | while(counter < columnIndex) { |
| 134 | if(isCharCJK(s.charAt(index++))) { |
| 135 | counter++; |
| 136 | if(counter == columnIndex) { |
| 137 | return index - 1; |
| 138 | } |
| 139 | } |
| 140 | counter++; |
| 141 | } |
| 142 | return index; |
| 143 | } |
| 144 | |
| 145 | /** |
| 146 | * Given a string that may or may not contain CJK characters, returns the substring which will fit inside |
| 147 | * <code>availableColumnSpace</code> columns. This method does not handle special cases like tab or new-line. |
| 148 | * <p> |
| 149 | * Calling this method is the same as calling {@code fitString(string, 0, availableColumnSpace)}. |
| 150 | * @param string The string to fit inside the availableColumnSpace |
| 151 | * @param availableColumnSpace Number of columns to fit the string inside |
| 152 | * @return The whole or part of the input string which will fit inside the supplied availableColumnSpace |
| 153 | */ |
| 154 | public static String fitString(String string, int availableColumnSpace) { |
| 155 | return fitString(string, 0, availableColumnSpace); |
| 156 | } |
| 157 | |
| 158 | /** |
| 159 | * Given a string that may or may not contain CJK characters, returns the substring which will fit inside |
| 160 | * <code>availableColumnSpace</code> columns. This method does not handle special cases like tab or new-line. |
| 161 | * <p> |
| 162 | * This overload has a {@code fromColumn} parameter that specified where inside the string to start fitting. Please |
| 163 | * notice that {@code fromColumn} is not a character index inside the string, but a column index as if the string |
| 164 | * has been printed from the left-most side of the terminal. So if the string is "日本語", fromColumn set to 1 will |
| 165 | * not starting counting from the second character ("本") in the string but from the CJK filler character belonging |
| 166 | * to "日". If you want to count from a particular character index inside the string, please pass in a substring |
| 167 | * and use fromColumn set to 0. |
| 168 | * @param string The string to fit inside the availableColumnSpace |
| 169 | * @param fromColumn From what column of the input string to start fitting (see description above!) |
| 170 | * @param availableColumnSpace Number of columns to fit the string inside |
| 171 | * @return The whole or part of the input string which will fit inside the supplied availableColumnSpace |
| 172 | */ |
| 173 | public static String fitString(String string, int fromColumn, int availableColumnSpace) { |
| 174 | if(availableColumnSpace <= 0) { |
| 175 | return ""; |
| 176 | } |
| 177 | |
| 178 | StringBuilder bob = new StringBuilder(); |
| 179 | int column = 0; |
| 180 | int index = 0; |
| 181 | while(index < string.length() && column < fromColumn) { |
| 182 | char c = string.charAt(index++); |
| 183 | column += TerminalTextUtils.isCharCJK(c) ? 2 : 1; |
| 184 | } |
| 185 | if(column > fromColumn) { |
| 186 | bob.append(" "); |
| 187 | availableColumnSpace--; |
| 188 | } |
| 189 | |
| 190 | while(availableColumnSpace > 0 && index < string.length()) { |
| 191 | char c = string.charAt(index++); |
| 192 | availableColumnSpace -= TerminalTextUtils.isCharCJK(c) ? 2 : 1; |
| 193 | if(availableColumnSpace < 0) { |
| 194 | bob.append(' '); |
| 195 | } |
| 196 | else { |
| 197 | bob.append(c); |
| 198 | } |
| 199 | } |
| 200 | return bob.toString(); |
| 201 | } |
| 202 | |
| 203 | /** |
| 204 | * This method will calculate word wrappings given a number of lines of text and how wide the text can be printed. |
| 205 | * The result is a list of new rows where word-wrapping was applied. |
| 206 | * @param maxWidth Maximum number of columns that can be used before word-wrapping is applied, if <= 0 then the |
| 207 | * lines will be returned unchanged |
| 208 | * @param lines Input text |
| 209 | * @return The input text word-wrapped at {@code maxWidth}; this may contain more rows than the input text |
| 210 | */ |
| 211 | public static List<String> getWordWrappedText(int maxWidth, String... lines) { |
| 212 | //Bounds checking |
| 213 | if(maxWidth <= 0) { |
| 214 | return Arrays.asList(lines); |
| 215 | } |
| 216 | |
| 217 | List<String> result = new ArrayList<String>(); |
| 218 | LinkedList<String> linesToBeWrapped = new LinkedList<String>(Arrays.asList(lines)); |
| 219 | while(!linesToBeWrapped.isEmpty()) { |
| 220 | String row = linesToBeWrapped.removeFirst(); |
| 221 | int rowWidth = getColumnWidth(row); |
| 222 | if(rowWidth <= maxWidth) { |
| 223 | result.add(row); |
| 224 | } |
| 225 | else { |
| 226 | //Now search in reverse and find the first possible line-break |
| 227 | final int characterIndexMax = getStringCharacterIndex(row, maxWidth); |
| 228 | int characterIndex = characterIndexMax; |
| 229 | while(characterIndex >= 0 && |
| 230 | !Character.isSpaceChar(row.charAt(characterIndex)) && |
| 231 | !isCharCJK(row.charAt(characterIndex))) { |
| 232 | characterIndex--; |
| 233 | } |
| 234 | // right *after* a CJK is also a "nice" spot to break the line! |
| 235 | if (characterIndex >= 0 && characterIndex < characterIndexMax && |
| 236 | isCharCJK(row.charAt(characterIndex))) { |
| 237 | characterIndex++; // with these conditions it fits! |
| 238 | } |
| 239 | |
| 240 | if(characterIndex < 0) { |
| 241 | //Failed! There was no 'nice' place to cut so just cut it at maxWidth |
| 242 | characterIndex = Math.max(characterIndexMax, 1); // at least 1 char |
| 243 | result.add(row.substring(0, characterIndex)); |
| 244 | linesToBeWrapped.addFirst(row.substring(characterIndex)); |
| 245 | } |
| 246 | else { |
| 247 | // characterIndex == 0 only happens, if either |
| 248 | // - first char is CJK and maxWidth==1 or |
| 249 | // - first char is whitespace |
| 250 | // either way: put it in row before break to prevent infinite loop. |
| 251 | characterIndex = Math.max( characterIndex, 1); // at least 1 char |
| 252 | |
| 253 | //Ok, split the row, add it to the result and continue processing the second half on a new line |
| 254 | result.add(row.substring(0, characterIndex)); |
| 255 | while(characterIndex < row.length() && |
| 256 | Character.isSpaceChar(row.charAt(characterIndex))) { |
| 257 | characterIndex++; |
| 258 | }; |
| 259 | if (characterIndex < row.length()) { // only if rest contains non-whitespace |
| 260 | linesToBeWrapped.addFirst(row.substring(characterIndex)); |
| 261 | } |
| 262 | } |
| 263 | } |
| 264 | } |
| 265 | return result; |
| 266 | } |
| 267 | } |