Utility.java revision 8845:4be14673b9bf
1/* 2 * Copyright (c) 2002, 2003, Oracle and/or its affiliates. All rights reserved. 3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. 4 * 5 * This code is free software; you can redistribute it and/or modify it 6 * under the terms of the GNU General Public License version 2 only, as 7 * published by the Free Software Foundation. Oracle designates this 8 * particular file as subject to the "Classpath" exception as provided 9 * by Oracle in the LICENSE file that accompanied this code. 10 * 11 * This code is distributed in the hope that it will be useful, but WITHOUT 12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or 13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 14 * version 2 for more details (a copy is included in the LICENSE file that 15 * accompanied this code). 16 * 17 * You should have received a copy of the GNU General Public License version 18 * 2 along with this work; if not, write to the Free Software Foundation, 19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. 20 * 21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA 22 * or visit www.oracle.com if you need additional information or have any 23 * questions. 24 */ 25 26package build.tools.generatecharacter; 27 28import java.text.*; 29import java.util.*; 30 31public class Utility { 32 static byte peekByte(String s, int index) { 33 char c = s.charAt(index/2); 34 return ((index&1)==0)?(byte)(c>>8):(byte)c; 35 } 36 37 static short peekShort(String s, int index) { 38 return (short)s.charAt(index); 39 } 40 41 static int peekInt(String s, int index) { 42 index *= 2; 43 return (((int)s.charAt(index)) << 16) | s.charAt(index+1); 44 } 45 46 static void poke(String s, int index, byte value) { 47 int mask = 0xFF00; 48 int ivalue = value; 49 if ((index&1)==0) { 50 ivalue <<= 8; 51 mask = 0x00FF; 52 } 53 index /= 2; 54 if (index == s.length()) { 55 s = s + (char)ivalue; 56 } 57 else if (index == 0) { 58 s = (char)(ivalue|(s.charAt(0)&mask)) + s.substring(1); 59 } 60 else { 61 s = s.substring(0, index) + (char)(ivalue|(s.charAt(index)&mask)) 62 + s.substring(index+1); 63 } 64 } 65 66 static void poke(String s, int index, short value) { 67 if (index == s.length()) { 68 s = s + (char)value; 69 } 70 else if (index == 0) { 71 s = (char)value + s.substring(1); 72 } 73 else { 74 s = s.substring(0, index) + (char)value + s.substring(index+1); 75 } 76 } 77 78 static void poke(String s, int index, int value) { 79 index *= 2; 80 char hi = (char)(value >> 16); 81 if (index == s.length()) { 82 s = s + hi + (char)value; 83 } 84 else if (index == 0) { 85 s = hi + (char)value + s.substring(2); 86 } 87 else { 88 s = s.substring(0, index) + hi + (char)value + s.substring(index+2); 89 } 90 } 91 92 /** 93 * The ESCAPE character is used during run-length encoding. It signals 94 * a run of identical chars. 95 */ 96 static final char ESCAPE = '\uA5A5'; 97 98 /** 99 * The ESCAPE_BYTE character is used during run-length encoding. It signals 100 * a run of identical bytes. 101 */ 102 static final byte ESCAPE_BYTE = (byte)0xA5; 103 104 /** 105 * Construct a string representing a short array. Use run-length encoding. 106 * A character represents itself, unless it is the ESCAPE character. Then 107 * the following notations are possible: 108 * ESCAPE ESCAPE ESCAPE literal 109 * ESCAPE n c n instances of character c 110 * Since an encoded run occupies 3 characters, we only encode runs of 4 or 111 * more characters. Thus we have n > 0 and n != ESCAPE and n <= 0xFFFF. 112 * If we encounter a run where n == ESCAPE, we represent this as: 113 * c ESCAPE n-1 c 114 * The ESCAPE value is chosen so as not to collide with commonly 115 * seen values. 116 */ 117 static final String arrayToRLEString(short[] a) { 118 StringBuffer buffer = new StringBuffer(); 119 // for (int i=0; i<a.length; ++i) buffer.append((char) a[i]); 120 buffer.append((char) (a.length >> 16)); 121 buffer.append((char) a.length); 122 short runValue = a[0]; 123 int runLength = 1; 124 for (int i=1; i<a.length; ++i) { 125 short s = a[i]; 126 if (s == runValue && runLength < 0xFFFF) ++runLength; 127 else { 128 encodeRun(buffer, runValue, runLength); 129 runValue = s; 130 runLength = 1; 131 } 132 } 133 encodeRun(buffer, runValue, runLength); 134 return buffer.toString(); 135 } 136 137 /** 138 * Construct a string representing a byte array. Use run-length encoding. 139 * Two bytes are packed into a single char, with a single extra zero byte at 140 * the end if needed. A byte represents itself, unless it is the 141 * ESCAPE_BYTE. Then the following notations are possible: 142 * ESCAPE_BYTE ESCAPE_BYTE ESCAPE_BYTE literal 143 * ESCAPE_BYTE n b n instances of byte b 144 * Since an encoded run occupies 3 bytes, we only encode runs of 4 or 145 * more bytes. Thus we have n > 0 and n != ESCAPE_BYTE and n <= 0xFF. 146 * If we encounter a run where n == ESCAPE_BYTE, we represent this as: 147 * b ESCAPE_BYTE n-1 b 148 * The ESCAPE_BYTE value is chosen so as not to collide with commonly 149 * seen values. 150 */ 151 static final String arrayToRLEString(byte[] a) { 152 StringBuffer buffer = new StringBuffer(); 153 buffer.append((char) (a.length >> 16)); 154 buffer.append((char) a.length); 155 byte runValue = a[0]; 156 int runLength = 1; 157 byte[] state = new byte[2]; 158 for (int i=1; i<a.length; ++i) { 159 byte b = a[i]; 160 if (b == runValue && runLength < 0xFF) ++runLength; 161 else { 162 encodeRun(buffer, runValue, runLength, state); 163 runValue = b; 164 runLength = 1; 165 } 166 } 167 encodeRun(buffer, runValue, runLength, state); 168 169 // We must save the final byte, if there is one, by padding 170 // an extra zero. 171 if (state[0] != 0) appendEncodedByte(buffer, (byte)0, state); 172 173 return buffer.toString(); 174 } 175 176 /** 177 * Encode a run, possibly a degenerate run (of < 4 values). 178 * @param length The length of the run; must be > 0 && <= 0xFFFF. 179 */ 180 private static final void encodeRun(StringBuffer buffer, short value, int length) { 181 if (length < 4) { 182 for (int j=0; j<length; ++j) { 183 if (value == (int) ESCAPE) buffer.append(ESCAPE); 184 buffer.append((char) value); 185 } 186 } 187 else { 188 if (length == (int) ESCAPE) { 189 if (value == (int) ESCAPE) buffer.append(ESCAPE); 190 buffer.append((char) value); 191 --length; 192 } 193 buffer.append(ESCAPE); 194 buffer.append((char) length); 195 buffer.append((char) value); // Don't need to escape this value 196 } 197 } 198 199 /** 200 * Encode a run, possibly a degenerate run (of < 4 values). 201 * @param length The length of the run; must be > 0 && <= 0xFF. 202 */ 203 private static final void encodeRun(StringBuffer buffer, byte value, int length, 204 byte[] state) { 205 if (length < 4) { 206 for (int j=0; j<length; ++j) { 207 if (value == ESCAPE_BYTE) appendEncodedByte(buffer, ESCAPE_BYTE, state); 208 appendEncodedByte(buffer, value, state); 209 } 210 } 211 else { 212 if (length == ESCAPE_BYTE) { 213 if (value == ESCAPE_BYTE) appendEncodedByte(buffer, ESCAPE_BYTE, state); 214 appendEncodedByte(buffer, value, state); 215 --length; 216 } 217 appendEncodedByte(buffer, ESCAPE_BYTE, state); 218 appendEncodedByte(buffer, (byte)length, state); 219 appendEncodedByte(buffer, value, state); // Don't need to escape this value 220 } 221 } 222 223 /** 224 * Append a byte to the given StringBuffer, packing two bytes into each 225 * character. The state parameter maintains intermediary data between 226 * calls. 227 * @param state A two-element array, with state[0] == 0 if this is the 228 * first byte of a pair, or state[0] != 0 if this is the second byte 229 * of a pair, in which case state[1] is the first byte. 230 */ 231 private static final void appendEncodedByte(StringBuffer buffer, byte value, 232 byte[] state) { 233 if (state[0] != 0) { 234 char c = (char) ((state[1] << 8) | (((int) value) & 0xFF)); 235 buffer.append(c); 236 state[0] = 0; 237 } 238 else { 239 state[0] = 1; 240 state[1] = value; 241 } 242 } 243 244 /** 245 * Construct an array of shorts from a run-length encoded string. 246 */ 247 static final short[] RLEStringToShortArray(String s) { 248 int length = (((int) s.charAt(0)) << 16) | ((int) s.charAt(1)); 249 short[] array = new short[length]; 250 int ai = 0; 251 for (int i=2; i<s.length(); ++i) { 252 char c = s.charAt(i); 253 if (c == ESCAPE) { 254 c = s.charAt(++i); 255 if (c == ESCAPE) array[ai++] = (short) c; 256 else { 257 int runLength = (int) c; 258 short runValue = (short) s.charAt(++i); 259 for (int j=0; j<runLength; ++j) array[ai++] = runValue; 260 } 261 } 262 else { 263 array[ai++] = (short) c; 264 } 265 } 266 267 if (ai != length) 268 throw new InternalError("Bad run-length encoded short array"); 269 270 return array; 271 } 272 273 /** 274 * Construct an array of bytes from a run-length encoded string. 275 */ 276 static final byte[] RLEStringToByteArray(String s) { 277 int length = (((int) s.charAt(0)) << 16) | ((int) s.charAt(1)); 278 byte[] array = new byte[length]; 279 boolean nextChar = true; 280 char c = 0; 281 int node = 0; 282 int runLength = 0; 283 int i = 2; 284 for (int ai=0; ai<length; ) { 285 // This part of the loop places the next byte into the local 286 // variable 'b' each time through the loop. It keeps the 287 // current character in 'c' and uses the boolean 'nextChar' 288 // to see if we've taken both bytes out of 'c' yet. 289 byte b; 290 if (nextChar) { 291 c = s.charAt(i++); 292 b = (byte) (c >> 8); 293 nextChar = false; 294 } 295 else { 296 b = (byte) (c & 0xFF); 297 nextChar = true; 298 } 299 300 // This part of the loop is a tiny state machine which handles 301 // the parsing of the run-length encoding. This would be simpler 302 // if we could look ahead, but we can't, so we use 'node' to 303 // move between three nodes in the state machine. 304 switch (node) { 305 case 0: 306 // Normal idle node 307 if (b == ESCAPE_BYTE) { 308 node = 1; 309 } 310 else { 311 array[ai++] = b; 312 } 313 break; 314 case 1: 315 // We have seen one ESCAPE_BYTE; we expect either a second 316 // one, or a run length and value. 317 if (b == ESCAPE_BYTE) { 318 array[ai++] = ESCAPE_BYTE; 319 node = 0; 320 } 321 else { 322 runLength = b; 323 // Interpret signed byte as unsigned 324 if (runLength < 0) runLength += 0x100; 325 node = 2; 326 } 327 break; 328 case 2: 329 // We have seen an ESCAPE_BYTE and length byte. We interpret 330 // the next byte as the value to be repeated. 331 for (int j=0; j<runLength; ++j) array[ai++] = b; 332 node = 0; 333 break; 334 } 335 } 336 337 if (node != 0) 338 throw new InternalError("Bad run-length encoded byte array"); 339 340 if (i != s.length()) 341 throw new InternalError("Excess data in RLE byte array string"); 342 343 return array; 344 } 345 346 /** 347 * Format a String for representation in a source file. This includes 348 * breaking it into lines escaping characters using octal notation 349 * when necessary (control characters and double quotes). 350 */ 351 static final String formatForSource(String s) { 352 return formatForSource(s, " "); 353 } 354 355 /** 356 * Format a String for representation in a source file. This includes 357 * breaking it into lines escaping characters using octal notation 358 * when necessary (control characters and double quotes). 359 */ 360 static final String formatForSource(String s, String indent) { 361 StringBuffer buffer = new StringBuffer(); 362 for (int i=0; i<s.length();) { 363 if (i > 0) buffer.append("+\n"); 364 int limit = buffer.length() + 78; // Leave 2 for trailing <"+> 365 buffer.append(indent + '"'); 366 while (i<s.length() && buffer.length()<limit) { 367 char c = s.charAt(i++); 368 /* This works too but it's kind of unnecessary; might as 369 well keep things simple. 370 if (c == '\\' || c == '"') { 371 // Escape backslash and double-quote. Don't need to 372 // escape single-quote. 373 buffer.append("\\" + c); 374 } 375 else if (c >= '\u0020' && c <= '\u007E') { 376 // Printable ASCII ranges from ' ' to '~' 377 buffer.append(c); 378 } 379 else 380 */ 381 if (c <= '\377') { 382 // Represent control characters 383 // using octal notation; otherwise the string we form 384 // won't compile, since Unicode escape sequences are 385 // processed before tokenization. 386 buffer.append('\\'); 387 buffer.append(HEX_DIGIT[(c & 0700) >> 6]); // HEX_DIGIT works for octal 388 buffer.append(HEX_DIGIT[(c & 0070) >> 3]); 389 buffer.append(HEX_DIGIT[(c & 0007)]); 390 } 391 else { 392 // Handle the rest with Unicode 393 buffer.append("\\u"); 394 buffer.append(HEX_DIGIT[(c & 0xF000) >> 12]); 395 buffer.append(HEX_DIGIT[(c & 0x0F00) >> 8]); 396 buffer.append(HEX_DIGIT[(c & 0x00F0) >> 4]); 397 buffer.append(HEX_DIGIT[(c & 0x000F)]); 398 } 399 } 400 buffer.append('"'); 401 } 402 return buffer.toString(); 403 } 404 405 static final char[] HEX_DIGIT = {'0','1','2','3','4','5','6','7', 406 '8','9','A','B','C','D','E','F'}; 407} 408