J. Duke | 319a3b9 | 2007-12-01 00:00:00 +0000 | [diff] [blame^] | 1 | /* |
| 2 | * Copyright 1999-2000 Sun Microsystems, Inc. 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. Sun designates this |
| 8 | * particular file as subject to the "Classpath" exception as provided |
| 9 | * by Sun 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 Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, |
| 22 | * CA 95054 USA or visit www.sun.com if you need additional information or |
| 23 | * have any questions. |
| 24 | */ |
| 25 | package javax.swing.text.html; |
| 26 | |
| 27 | import java.io.*; |
| 28 | |
| 29 | /** |
| 30 | * A CSS parser. This works by way of a delegate that implements the |
| 31 | * CSSParserCallback interface. The delegate is notified of the following |
| 32 | * events: |
| 33 | * <ul> |
| 34 | * <li>Import statement: <code>handleImport</code> |
| 35 | * <li>Selectors <code>handleSelector</code>. This is invoked for each |
| 36 | * string. For example if the Reader contained p, bar , a {}, the delegate |
| 37 | * would be notified 4 times, for 'p,' 'bar' ',' and 'a'. |
| 38 | * <li>When a rule starts, <code>startRule</code> |
| 39 | * <li>Properties in the rule via the <code>handleProperty</code>. This |
| 40 | * is invoked one per property/value key, eg font size: foo;, would |
| 41 | * cause the delegate to be notified once with a value of 'font size'. |
| 42 | * <li>Values in the rule via the <code>handleValue</code>, this is notified |
| 43 | * for the total value. |
| 44 | * <li>When a rule ends, <code>endRule</code> |
| 45 | * </ul> |
| 46 | * This will parse much more than CSS 1, and loosely implements the |
| 47 | * recommendation for <i>Forward-compatible parsing</i> in section |
| 48 | * 7.1 of the CSS spec found at: |
| 49 | * <a href=http://www.w3.org/TR/REC-CSS1>http://www.w3.org/TR/REC-CSS1</a>. |
| 50 | * If an error results in parsing, a RuntimeException will be thrown. |
| 51 | * <p> |
| 52 | * This will preserve case. If the callback wishes to treat certain poritions |
| 53 | * case insensitively (such as selectors), it should use toLowerCase, or |
| 54 | * something similar. |
| 55 | * |
| 56 | * @author Scott Violet |
| 57 | */ |
| 58 | class CSSParser { |
| 59 | // Parsing something like the following: |
| 60 | // (@rule | ruleset | block)* |
| 61 | // |
| 62 | // @rule (block | identifier)*; (block with {} ends @rule) |
| 63 | // block matching [] () {} (that is, [()] is a block, [(){}{[]}] |
| 64 | // is a block, ()[] is two blocks) |
| 65 | // identifier "*" | '*' | anything but a [](){} and whitespace |
| 66 | // |
| 67 | // ruleset selector decblock |
| 68 | // selector (identifier | (block, except block '{}') )* |
| 69 | // declblock declaration* block* |
| 70 | // declaration (identifier* stopping when identifier ends with :) |
| 71 | // (identifier* stopping when identifier ends with ;) |
| 72 | // |
| 73 | // comments /* */ can appear any where, and are stripped. |
| 74 | |
| 75 | |
| 76 | // identifier - letters, digits, dashes and escaped characters |
| 77 | // block starts with { ends with matching }, () [] and {} always occur |
| 78 | // in matching pairs, '' and "" also occur in pairs, except " may be |
| 79 | |
| 80 | |
| 81 | // Indicates the type of token being parsed. |
| 82 | private static final int IDENTIFIER = 1; |
| 83 | private static final int BRACKET_OPEN = 2; |
| 84 | private static final int BRACKET_CLOSE = 3; |
| 85 | private static final int BRACE_OPEN = 4; |
| 86 | private static final int BRACE_CLOSE = 5; |
| 87 | private static final int PAREN_OPEN = 6; |
| 88 | private static final int PAREN_CLOSE = 7; |
| 89 | private static final int END = -1; |
| 90 | |
| 91 | private static final char[] charMapping = { 0, 0, '[', ']', '{', '}', '(', |
| 92 | ')', 0}; |
| 93 | |
| 94 | |
| 95 | /** Set to true if one character has been read ahead. */ |
| 96 | private boolean didPushChar; |
| 97 | /** The read ahead character. */ |
| 98 | private int pushedChar; |
| 99 | /** Temporary place to hold identifiers. */ |
| 100 | private StringBuffer unitBuffer; |
| 101 | /** Used to indicate blocks. */ |
| 102 | private int[] unitStack; |
| 103 | /** Number of valid blocks. */ |
| 104 | private int stackCount; |
| 105 | /** Holds the incoming CSS rules. */ |
| 106 | private Reader reader; |
| 107 | /** Set to true when the first non @ rule is encountered. */ |
| 108 | private boolean encounteredRuleSet; |
| 109 | /** Notified of state. */ |
| 110 | private CSSParserCallback callback; |
| 111 | /** nextToken() inserts the string here. */ |
| 112 | private char[] tokenBuffer; |
| 113 | /** Current number of chars in tokenBufferLength. */ |
| 114 | private int tokenBufferLength; |
| 115 | /** Set to true if any whitespace is read. */ |
| 116 | private boolean readWS; |
| 117 | |
| 118 | |
| 119 | // The delegate interface. |
| 120 | static interface CSSParserCallback { |
| 121 | /** Called when an @import is encountered. */ |
| 122 | void handleImport(String importString); |
| 123 | // There is currently no way to distinguish between '"foo,"' and |
| 124 | // 'foo,'. But this generally isn't valid CSS. If it becomes |
| 125 | // a problem, handleSelector will have to be told if the string is |
| 126 | // quoted. |
| 127 | void handleSelector(String selector); |
| 128 | void startRule(); |
| 129 | // Property names are mapped to lower case before being passed to |
| 130 | // the delegate. |
| 131 | void handleProperty(String property); |
| 132 | void handleValue(String value); |
| 133 | void endRule(); |
| 134 | } |
| 135 | |
| 136 | CSSParser() { |
| 137 | unitStack = new int[2]; |
| 138 | tokenBuffer = new char[80]; |
| 139 | unitBuffer = new StringBuffer(); |
| 140 | } |
| 141 | |
| 142 | void parse(Reader reader, CSSParserCallback callback, |
| 143 | boolean inRule) throws IOException { |
| 144 | this.callback = callback; |
| 145 | stackCount = tokenBufferLength = 0; |
| 146 | this.reader = reader; |
| 147 | encounteredRuleSet = false; |
| 148 | try { |
| 149 | if (inRule) { |
| 150 | parseDeclarationBlock(); |
| 151 | } |
| 152 | else { |
| 153 | while (getNextStatement()); |
| 154 | } |
| 155 | } finally { |
| 156 | callback = null; |
| 157 | reader = null; |
| 158 | } |
| 159 | } |
| 160 | |
| 161 | /** |
| 162 | * Gets the next statement, returning false if the end is reached. A |
| 163 | * statement is either an @rule, or a ruleset. |
| 164 | */ |
| 165 | private boolean getNextStatement() throws IOException { |
| 166 | unitBuffer.setLength(0); |
| 167 | |
| 168 | int token = nextToken((char)0); |
| 169 | |
| 170 | switch (token) { |
| 171 | case IDENTIFIER: |
| 172 | if (tokenBufferLength > 0) { |
| 173 | if (tokenBuffer[0] == '@') { |
| 174 | parseAtRule(); |
| 175 | } |
| 176 | else { |
| 177 | encounteredRuleSet = true; |
| 178 | parseRuleSet(); |
| 179 | } |
| 180 | } |
| 181 | return true; |
| 182 | case BRACKET_OPEN: |
| 183 | case BRACE_OPEN: |
| 184 | case PAREN_OPEN: |
| 185 | parseTillClosed(token); |
| 186 | return true; |
| 187 | |
| 188 | case BRACKET_CLOSE: |
| 189 | case BRACE_CLOSE: |
| 190 | case PAREN_CLOSE: |
| 191 | // Shouldn't happen... |
| 192 | throw new RuntimeException("Unexpected top level block close"); |
| 193 | |
| 194 | case END: |
| 195 | return false; |
| 196 | } |
| 197 | return true; |
| 198 | } |
| 199 | |
| 200 | /** |
| 201 | * Parses an @ rule, stopping at a matching brace pair, or ;. |
| 202 | */ |
| 203 | private void parseAtRule() throws IOException { |
| 204 | // PENDING: make this more effecient. |
| 205 | boolean done = false; |
| 206 | boolean isImport = (tokenBufferLength == 7 && |
| 207 | tokenBuffer[0] == '@' && tokenBuffer[1] == 'i' && |
| 208 | tokenBuffer[2] == 'm' && tokenBuffer[3] == 'p' && |
| 209 | tokenBuffer[4] == 'o' && tokenBuffer[5] == 'r' && |
| 210 | tokenBuffer[6] == 't'); |
| 211 | |
| 212 | unitBuffer.setLength(0); |
| 213 | while (!done) { |
| 214 | int nextToken = nextToken(';'); |
| 215 | |
| 216 | switch (nextToken) { |
| 217 | case IDENTIFIER: |
| 218 | if (tokenBufferLength > 0 && |
| 219 | tokenBuffer[tokenBufferLength - 1] == ';') { |
| 220 | --tokenBufferLength; |
| 221 | done = true; |
| 222 | } |
| 223 | if (tokenBufferLength > 0) { |
| 224 | if (unitBuffer.length() > 0 && readWS) { |
| 225 | unitBuffer.append(' '); |
| 226 | } |
| 227 | unitBuffer.append(tokenBuffer, 0, tokenBufferLength); |
| 228 | } |
| 229 | break; |
| 230 | |
| 231 | case BRACE_OPEN: |
| 232 | if (unitBuffer.length() > 0 && readWS) { |
| 233 | unitBuffer.append(' '); |
| 234 | } |
| 235 | unitBuffer.append(charMapping[nextToken]); |
| 236 | parseTillClosed(nextToken); |
| 237 | done = true; |
| 238 | // Skip a tailing ';', not really to spec. |
| 239 | { |
| 240 | int nextChar = readWS(); |
| 241 | if (nextChar != -1 && nextChar != ';') { |
| 242 | pushChar(nextChar); |
| 243 | } |
| 244 | } |
| 245 | break; |
| 246 | |
| 247 | case BRACKET_OPEN: case PAREN_OPEN: |
| 248 | unitBuffer.append(charMapping[nextToken]); |
| 249 | parseTillClosed(nextToken); |
| 250 | break; |
| 251 | |
| 252 | case BRACKET_CLOSE: case BRACE_CLOSE: case PAREN_CLOSE: |
| 253 | throw new RuntimeException("Unexpected close in @ rule"); |
| 254 | |
| 255 | case END: |
| 256 | done = true; |
| 257 | break; |
| 258 | } |
| 259 | } |
| 260 | if (isImport && !encounteredRuleSet) { |
| 261 | callback.handleImport(unitBuffer.toString()); |
| 262 | } |
| 263 | } |
| 264 | |
| 265 | /** |
| 266 | * Parses the next rule set, which is a selector followed by a |
| 267 | * declaration block. |
| 268 | */ |
| 269 | private void parseRuleSet() throws IOException { |
| 270 | if (parseSelectors()) { |
| 271 | callback.startRule(); |
| 272 | parseDeclarationBlock(); |
| 273 | callback.endRule(); |
| 274 | } |
| 275 | } |
| 276 | |
| 277 | /** |
| 278 | * Parses a set of selectors, returning false if the end of the stream |
| 279 | * is reached. |
| 280 | */ |
| 281 | private boolean parseSelectors() throws IOException { |
| 282 | // Parse the selectors |
| 283 | int nextToken; |
| 284 | |
| 285 | if (tokenBufferLength > 0) { |
| 286 | callback.handleSelector(new String(tokenBuffer, 0, |
| 287 | tokenBufferLength)); |
| 288 | } |
| 289 | |
| 290 | unitBuffer.setLength(0); |
| 291 | for (;;) { |
| 292 | while ((nextToken = nextToken((char)0)) == IDENTIFIER) { |
| 293 | if (tokenBufferLength > 0) { |
| 294 | callback.handleSelector(new String(tokenBuffer, 0, |
| 295 | tokenBufferLength)); |
| 296 | } |
| 297 | } |
| 298 | switch (nextToken) { |
| 299 | case BRACE_OPEN: |
| 300 | return true; |
| 301 | |
| 302 | case BRACKET_OPEN: case PAREN_OPEN: |
| 303 | parseTillClosed(nextToken); |
| 304 | // Not too sure about this, how we handle this isn't very |
| 305 | // well spec'd. |
| 306 | unitBuffer.setLength(0); |
| 307 | break; |
| 308 | |
| 309 | case BRACKET_CLOSE: case BRACE_CLOSE: case PAREN_CLOSE: |
| 310 | throw new RuntimeException("Unexpected block close in selector"); |
| 311 | |
| 312 | case END: |
| 313 | // Prematurely hit end. |
| 314 | return false; |
| 315 | } |
| 316 | } |
| 317 | } |
| 318 | |
| 319 | /** |
| 320 | * Parses a declaration block. Which a number of declarations followed |
| 321 | * by a })]. |
| 322 | */ |
| 323 | private void parseDeclarationBlock() throws IOException { |
| 324 | for (;;) { |
| 325 | int token = parseDeclaration(); |
| 326 | switch (token) { |
| 327 | case END: case BRACE_CLOSE: |
| 328 | return; |
| 329 | |
| 330 | case BRACKET_CLOSE: case PAREN_CLOSE: |
| 331 | // Bail |
| 332 | throw new RuntimeException("Unexpected close in declaration block"); |
| 333 | case IDENTIFIER: |
| 334 | break; |
| 335 | } |
| 336 | } |
| 337 | } |
| 338 | |
| 339 | /** |
| 340 | * Parses a single declaration, which is an identifier a : and another |
| 341 | * identifier. This returns the last token seen. |
| 342 | */ |
| 343 | // identifier+: identifier* ;|} |
| 344 | private int parseDeclaration() throws IOException { |
| 345 | int token; |
| 346 | |
| 347 | if ((token = parseIdentifiers(':', false)) != IDENTIFIER) { |
| 348 | return token; |
| 349 | } |
| 350 | // Make the property name to lowercase |
| 351 | for (int counter = unitBuffer.length() - 1; counter >= 0; counter--) { |
| 352 | unitBuffer.setCharAt(counter, Character.toLowerCase |
| 353 | (unitBuffer.charAt(counter))); |
| 354 | } |
| 355 | callback.handleProperty(unitBuffer.toString()); |
| 356 | |
| 357 | token = parseIdentifiers(';', true); |
| 358 | callback.handleValue(unitBuffer.toString()); |
| 359 | return token; |
| 360 | } |
| 361 | |
| 362 | /** |
| 363 | * Parses identifiers until <code>extraChar</code> is encountered, |
| 364 | * returning the ending token, which will be IDENTIFIER if extraChar |
| 365 | * is found. |
| 366 | */ |
| 367 | private int parseIdentifiers(char extraChar, |
| 368 | boolean wantsBlocks) throws IOException { |
| 369 | int nextToken; |
| 370 | int ubl; |
| 371 | |
| 372 | unitBuffer.setLength(0); |
| 373 | for (;;) { |
| 374 | nextToken = nextToken(extraChar); |
| 375 | |
| 376 | switch (nextToken) { |
| 377 | case IDENTIFIER: |
| 378 | if (tokenBufferLength > 0) { |
| 379 | if (tokenBuffer[tokenBufferLength - 1] == extraChar) { |
| 380 | if (--tokenBufferLength > 0) { |
| 381 | if (readWS && unitBuffer.length() > 0) { |
| 382 | unitBuffer.append(' '); |
| 383 | } |
| 384 | unitBuffer.append(tokenBuffer, 0, |
| 385 | tokenBufferLength); |
| 386 | } |
| 387 | return IDENTIFIER; |
| 388 | } |
| 389 | if (readWS && unitBuffer.length() > 0) { |
| 390 | unitBuffer.append(' '); |
| 391 | } |
| 392 | unitBuffer.append(tokenBuffer, 0, tokenBufferLength); |
| 393 | } |
| 394 | break; |
| 395 | |
| 396 | case BRACKET_OPEN: |
| 397 | case BRACE_OPEN: |
| 398 | case PAREN_OPEN: |
| 399 | ubl = unitBuffer.length(); |
| 400 | if (wantsBlocks) { |
| 401 | unitBuffer.append(charMapping[nextToken]); |
| 402 | } |
| 403 | parseTillClosed(nextToken); |
| 404 | if (!wantsBlocks) { |
| 405 | unitBuffer.setLength(ubl); |
| 406 | } |
| 407 | break; |
| 408 | |
| 409 | case BRACE_CLOSE: |
| 410 | // No need to throw for these two, we return token and |
| 411 | // caller can do whatever. |
| 412 | case BRACKET_CLOSE: |
| 413 | case PAREN_CLOSE: |
| 414 | case END: |
| 415 | // Hit the end |
| 416 | return nextToken; |
| 417 | } |
| 418 | } |
| 419 | } |
| 420 | |
| 421 | /** |
| 422 | * Parses till a matching block close is encountered. This is only |
| 423 | * appropriate to be called at the top level (no nesting). |
| 424 | */ |
| 425 | private void parseTillClosed(int openToken) throws IOException { |
| 426 | int nextToken; |
| 427 | boolean done = false; |
| 428 | |
| 429 | startBlock(openToken); |
| 430 | while (!done) { |
| 431 | nextToken = nextToken((char)0); |
| 432 | switch (nextToken) { |
| 433 | case IDENTIFIER: |
| 434 | if (unitBuffer.length() > 0 && readWS) { |
| 435 | unitBuffer.append(' '); |
| 436 | } |
| 437 | if (tokenBufferLength > 0) { |
| 438 | unitBuffer.append(tokenBuffer, 0, tokenBufferLength); |
| 439 | } |
| 440 | break; |
| 441 | |
| 442 | case BRACKET_OPEN: case BRACE_OPEN: case PAREN_OPEN: |
| 443 | if (unitBuffer.length() > 0 && readWS) { |
| 444 | unitBuffer.append(' '); |
| 445 | } |
| 446 | unitBuffer.append(charMapping[nextToken]); |
| 447 | startBlock(nextToken); |
| 448 | break; |
| 449 | |
| 450 | case BRACKET_CLOSE: case BRACE_CLOSE: case PAREN_CLOSE: |
| 451 | if (unitBuffer.length() > 0 && readWS) { |
| 452 | unitBuffer.append(' '); |
| 453 | } |
| 454 | unitBuffer.append(charMapping[nextToken]); |
| 455 | endBlock(nextToken); |
| 456 | if (!inBlock()) { |
| 457 | done = true; |
| 458 | } |
| 459 | break; |
| 460 | |
| 461 | case END: |
| 462 | // Prematurely hit end. |
| 463 | throw new RuntimeException("Unclosed block"); |
| 464 | } |
| 465 | } |
| 466 | } |
| 467 | |
| 468 | /** |
| 469 | * Fetches the next token. |
| 470 | */ |
| 471 | private int nextToken(char idChar) throws IOException { |
| 472 | readWS = false; |
| 473 | |
| 474 | int nextChar = readWS(); |
| 475 | |
| 476 | switch (nextChar) { |
| 477 | case '\'': |
| 478 | readTill('\''); |
| 479 | if (tokenBufferLength > 0) { |
| 480 | tokenBufferLength--; |
| 481 | } |
| 482 | return IDENTIFIER; |
| 483 | case '"': |
| 484 | readTill('"'); |
| 485 | if (tokenBufferLength > 0) { |
| 486 | tokenBufferLength--; |
| 487 | } |
| 488 | return IDENTIFIER; |
| 489 | case '[': |
| 490 | return BRACKET_OPEN; |
| 491 | case ']': |
| 492 | return BRACKET_CLOSE; |
| 493 | case '{': |
| 494 | return BRACE_OPEN; |
| 495 | case '}': |
| 496 | return BRACE_CLOSE; |
| 497 | case '(': |
| 498 | return PAREN_OPEN; |
| 499 | case ')': |
| 500 | return PAREN_CLOSE; |
| 501 | case -1: |
| 502 | return END; |
| 503 | default: |
| 504 | pushChar(nextChar); |
| 505 | getIdentifier(idChar); |
| 506 | return IDENTIFIER; |
| 507 | } |
| 508 | } |
| 509 | |
| 510 | /** |
| 511 | * Gets an identifier, returning true if the length of the string is greater than 0, |
| 512 | * stopping when <code>stopChar</code>, whitespace, or one of {}()[] is |
| 513 | * hit. |
| 514 | */ |
| 515 | // NOTE: this could be combined with readTill, as they contain somewhat |
| 516 | // similiar functionality. |
| 517 | private boolean getIdentifier(char stopChar) throws IOException { |
| 518 | boolean lastWasEscape = false; |
| 519 | boolean done = false; |
| 520 | int escapeCount = 0; |
| 521 | int escapeChar = 0; |
| 522 | int nextChar; |
| 523 | int intStopChar = (int)stopChar; |
| 524 | // 1 for '\', 2 for valid escape char [0-9a-fA-F], 3 for |
| 525 | // stop character (white space, ()[]{}) 0 otherwise |
| 526 | short type; |
| 527 | int escapeOffset = 0; |
| 528 | |
| 529 | tokenBufferLength = 0; |
| 530 | while (!done) { |
| 531 | nextChar = readChar(); |
| 532 | switch (nextChar) { |
| 533 | case '\\': |
| 534 | type = 1; |
| 535 | break; |
| 536 | |
| 537 | case '0': case '1': case '2': case '3': case '4': case '5': |
| 538 | case '6': case '7': case '8': case '9': |
| 539 | type = 2; |
| 540 | escapeOffset = nextChar - '0'; |
| 541 | break; |
| 542 | |
| 543 | case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': |
| 544 | type = 2; |
| 545 | escapeOffset = nextChar - 'a' + 10; |
| 546 | break; |
| 547 | |
| 548 | case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': |
| 549 | type = 2; |
| 550 | escapeOffset = nextChar - 'A' + 10; |
| 551 | break; |
| 552 | |
| 553 | case '\'': case '"': case '[': case ']': case '{': case '}': |
| 554 | case '(': case ')': |
| 555 | case ' ': case '\n': case '\t': case '\r': |
| 556 | type = 3; |
| 557 | break; |
| 558 | |
| 559 | case '/': |
| 560 | type = 4; |
| 561 | break; |
| 562 | |
| 563 | case -1: |
| 564 | // Reached the end |
| 565 | done = true; |
| 566 | type = 0; |
| 567 | break; |
| 568 | |
| 569 | default: |
| 570 | type = 0; |
| 571 | break; |
| 572 | } |
| 573 | if (lastWasEscape) { |
| 574 | if (type == 2) { |
| 575 | // Continue with escape. |
| 576 | escapeChar = escapeChar * 16 + escapeOffset; |
| 577 | if (++escapeCount == 4) { |
| 578 | lastWasEscape = false; |
| 579 | append((char)escapeChar); |
| 580 | } |
| 581 | } |
| 582 | else { |
| 583 | // no longer escaped |
| 584 | lastWasEscape = false; |
| 585 | if (escapeCount > 0) { |
| 586 | append((char)escapeChar); |
| 587 | // Make this simpler, reprocess the character. |
| 588 | pushChar(nextChar); |
| 589 | } |
| 590 | else if (!done) { |
| 591 | append((char)nextChar); |
| 592 | } |
| 593 | } |
| 594 | } |
| 595 | else if (!done) { |
| 596 | if (type == 1) { |
| 597 | lastWasEscape = true; |
| 598 | escapeChar = escapeCount = 0; |
| 599 | } |
| 600 | else if (type == 3) { |
| 601 | done = true; |
| 602 | pushChar(nextChar); |
| 603 | } |
| 604 | else if (type == 4) { |
| 605 | // Potential comment |
| 606 | nextChar = readChar(); |
| 607 | if (nextChar == '*') { |
| 608 | done = true; |
| 609 | readComment(); |
| 610 | readWS = true; |
| 611 | } |
| 612 | else { |
| 613 | append('/'); |
| 614 | if (nextChar == -1) { |
| 615 | done = true; |
| 616 | } |
| 617 | else { |
| 618 | pushChar(nextChar); |
| 619 | } |
| 620 | } |
| 621 | } |
| 622 | else { |
| 623 | append((char)nextChar); |
| 624 | if (nextChar == intStopChar) { |
| 625 | done = true; |
| 626 | } |
| 627 | } |
| 628 | } |
| 629 | } |
| 630 | return (tokenBufferLength > 0); |
| 631 | } |
| 632 | |
| 633 | /** |
| 634 | * Reads till a <code>stopChar</code> is encountered, escaping characters |
| 635 | * as necessary. |
| 636 | */ |
| 637 | private void readTill(char stopChar) throws IOException { |
| 638 | boolean lastWasEscape = false; |
| 639 | int escapeCount = 0; |
| 640 | int escapeChar = 0; |
| 641 | int nextChar; |
| 642 | boolean done = false; |
| 643 | int intStopChar = (int)stopChar; |
| 644 | // 1 for '\', 2 for valid escape char [0-9a-fA-F], 0 otherwise |
| 645 | short type; |
| 646 | int escapeOffset = 0; |
| 647 | |
| 648 | tokenBufferLength = 0; |
| 649 | while (!done) { |
| 650 | nextChar = readChar(); |
| 651 | switch (nextChar) { |
| 652 | case '\\': |
| 653 | type = 1; |
| 654 | break; |
| 655 | |
| 656 | case '0': case '1': case '2': case '3': case '4':case '5': |
| 657 | case '6': case '7': case '8': case '9': |
| 658 | type = 2; |
| 659 | escapeOffset = nextChar - '0'; |
| 660 | break; |
| 661 | |
| 662 | case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': |
| 663 | type = 2; |
| 664 | escapeOffset = nextChar - 'a' + 10; |
| 665 | break; |
| 666 | |
| 667 | case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': |
| 668 | type = 2; |
| 669 | escapeOffset = nextChar - 'A' + 10; |
| 670 | break; |
| 671 | |
| 672 | case -1: |
| 673 | // Prematurely reached the end! |
| 674 | throw new RuntimeException("Unclosed " + stopChar); |
| 675 | |
| 676 | default: |
| 677 | type = 0; |
| 678 | break; |
| 679 | } |
| 680 | if (lastWasEscape) { |
| 681 | if (type == 2) { |
| 682 | // Continue with escape. |
| 683 | escapeChar = escapeChar * 16 + escapeOffset; |
| 684 | if (++escapeCount == 4) { |
| 685 | lastWasEscape = false; |
| 686 | append((char)escapeChar); |
| 687 | } |
| 688 | } |
| 689 | else { |
| 690 | // no longer escaped |
| 691 | if (escapeCount > 0) { |
| 692 | append((char)escapeChar); |
| 693 | if (type == 1) { |
| 694 | lastWasEscape = true; |
| 695 | escapeChar = escapeCount = 0; |
| 696 | } |
| 697 | else { |
| 698 | if (nextChar == intStopChar) { |
| 699 | done = true; |
| 700 | } |
| 701 | append((char)nextChar); |
| 702 | lastWasEscape = false; |
| 703 | } |
| 704 | } |
| 705 | else { |
| 706 | append((char)nextChar); |
| 707 | lastWasEscape = false; |
| 708 | } |
| 709 | } |
| 710 | } |
| 711 | else if (type == 1) { |
| 712 | lastWasEscape = true; |
| 713 | escapeChar = escapeCount = 0; |
| 714 | } |
| 715 | else { |
| 716 | if (nextChar == intStopChar) { |
| 717 | done = true; |
| 718 | } |
| 719 | append((char)nextChar); |
| 720 | } |
| 721 | } |
| 722 | } |
| 723 | |
| 724 | private void append(char character) { |
| 725 | if (tokenBufferLength == tokenBuffer.length) { |
| 726 | char[] newBuffer = new char[tokenBuffer.length * 2]; |
| 727 | System.arraycopy(tokenBuffer, 0, newBuffer, 0, tokenBuffer.length); |
| 728 | tokenBuffer = newBuffer; |
| 729 | } |
| 730 | tokenBuffer[tokenBufferLength++] = character; |
| 731 | } |
| 732 | |
| 733 | /** |
| 734 | * Parses a comment block. |
| 735 | */ |
| 736 | private void readComment() throws IOException { |
| 737 | int nextChar; |
| 738 | |
| 739 | for(;;) { |
| 740 | nextChar = readChar(); |
| 741 | switch (nextChar) { |
| 742 | case -1: |
| 743 | throw new RuntimeException("Unclosed comment"); |
| 744 | case '*': |
| 745 | nextChar = readChar(); |
| 746 | if (nextChar == '/') { |
| 747 | return; |
| 748 | } |
| 749 | else if (nextChar == -1) { |
| 750 | throw new RuntimeException("Unclosed comment"); |
| 751 | } |
| 752 | else { |
| 753 | pushChar(nextChar); |
| 754 | } |
| 755 | break; |
| 756 | default: |
| 757 | break; |
| 758 | } |
| 759 | } |
| 760 | } |
| 761 | |
| 762 | /** |
| 763 | * Called when a block start is encountered ({[. |
| 764 | */ |
| 765 | private void startBlock(int startToken) { |
| 766 | if (stackCount == unitStack.length) { |
| 767 | int[] newUS = new int[stackCount * 2]; |
| 768 | |
| 769 | System.arraycopy(unitStack, 0, newUS, 0, stackCount); |
| 770 | unitStack = newUS; |
| 771 | } |
| 772 | unitStack[stackCount++] = startToken; |
| 773 | } |
| 774 | |
| 775 | /** |
| 776 | * Called when an end block is encountered )]} |
| 777 | */ |
| 778 | private void endBlock(int endToken) { |
| 779 | int startToken; |
| 780 | |
| 781 | switch (endToken) { |
| 782 | case BRACKET_CLOSE: |
| 783 | startToken = BRACKET_OPEN; |
| 784 | break; |
| 785 | case BRACE_CLOSE: |
| 786 | startToken = BRACE_OPEN; |
| 787 | break; |
| 788 | case PAREN_CLOSE: |
| 789 | startToken = PAREN_OPEN; |
| 790 | break; |
| 791 | default: |
| 792 | // Will never happen. |
| 793 | startToken = -1; |
| 794 | break; |
| 795 | } |
| 796 | if (stackCount > 0 && unitStack[stackCount - 1] == startToken) { |
| 797 | stackCount--; |
| 798 | } |
| 799 | else { |
| 800 | // Invalid state, should do something. |
| 801 | throw new RuntimeException("Unmatched block"); |
| 802 | } |
| 803 | } |
| 804 | |
| 805 | /** |
| 806 | * @return true if currently in a block. |
| 807 | */ |
| 808 | private boolean inBlock() { |
| 809 | return (stackCount > 0); |
| 810 | } |
| 811 | |
| 812 | /** |
| 813 | * Skips any white space, returning the character after the white space. |
| 814 | */ |
| 815 | private int readWS() throws IOException { |
| 816 | int nextChar; |
| 817 | while ((nextChar = readChar()) != -1 && |
| 818 | Character.isWhitespace((char)nextChar)) { |
| 819 | readWS = true; |
| 820 | } |
| 821 | return nextChar; |
| 822 | } |
| 823 | |
| 824 | /** |
| 825 | * Reads a character from the stream. |
| 826 | */ |
| 827 | private int readChar() throws IOException { |
| 828 | if (didPushChar) { |
| 829 | didPushChar = false; |
| 830 | return pushedChar; |
| 831 | } |
| 832 | return reader.read(); |
| 833 | // Uncomment the following to do case insensitive parsing. |
| 834 | /* |
| 835 | if (retValue != -1) { |
| 836 | return (int)Character.toLowerCase((char)retValue); |
| 837 | } |
| 838 | return retValue; |
| 839 | */ |
| 840 | } |
| 841 | |
| 842 | /** |
| 843 | * Supports one character look ahead, this will throw if called twice |
| 844 | * in a row. |
| 845 | */ |
| 846 | private void pushChar(int tempChar) { |
| 847 | if (didPushChar) { |
| 848 | // Should never happen. |
| 849 | throw new RuntimeException("Can not handle look ahead of more than one character"); |
| 850 | } |
| 851 | didPushChar = true; |
| 852 | pushedChar = tempChar; |
| 853 | } |
| 854 | } |