Wed, 20 May 2009 13:36:23 -0700
6827026: Change javac source and target default to 7
Reviewed-by: darcy, ohair
duke@1 | 1 | /* |
xdono@54 | 2 | * Copyright 1999-2008 Sun Microsystems, Inc. All Rights Reserved. |
duke@1 | 3 | * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. |
duke@1 | 4 | * |
duke@1 | 5 | * This code is free software; you can redistribute it and/or modify it |
duke@1 | 6 | * under the terms of the GNU General Public License version 2 only, as |
duke@1 | 7 | * published by the Free Software Foundation. Sun designates this |
duke@1 | 8 | * particular file as subject to the "Classpath" exception as provided |
duke@1 | 9 | * by Sun in the LICENSE file that accompanied this code. |
duke@1 | 10 | * |
duke@1 | 11 | * This code is distributed in the hope that it will be useful, but WITHOUT |
duke@1 | 12 | * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
duke@1 | 13 | * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
duke@1 | 14 | * version 2 for more details (a copy is included in the LICENSE file that |
duke@1 | 15 | * accompanied this code). |
duke@1 | 16 | * |
duke@1 | 17 | * You should have received a copy of the GNU General Public License version |
duke@1 | 18 | * 2 along with this work; if not, write to the Free Software Foundation, |
duke@1 | 19 | * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. |
duke@1 | 20 | * |
duke@1 | 21 | * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, |
duke@1 | 22 | * CA 95054 USA or visit www.sun.com if you need additional information or |
duke@1 | 23 | * have any questions. |
duke@1 | 24 | */ |
duke@1 | 25 | |
duke@1 | 26 | package com.sun.tools.javac.parser; |
duke@1 | 27 | |
duke@1 | 28 | import java.nio.*; |
duke@1 | 29 | |
jjg@50 | 30 | import com.sun.tools.javac.code.Source; |
jjg@50 | 31 | import com.sun.tools.javac.file.JavacFileManager; |
duke@1 | 32 | import com.sun.tools.javac.util.*; |
duke@1 | 33 | |
duke@1 | 34 | |
duke@1 | 35 | import static com.sun.tools.javac.parser.Token.*; |
duke@1 | 36 | import static com.sun.tools.javac.util.LayoutCharacters.*; |
duke@1 | 37 | |
duke@1 | 38 | /** The lexical analyzer maps an input stream consisting of |
duke@1 | 39 | * ASCII characters and Unicode escapes into a token sequence. |
duke@1 | 40 | * |
duke@1 | 41 | * <p><b>This is NOT part of any API supported by Sun Microsystems. If |
duke@1 | 42 | * you write code that depends on this, you do so at your own risk. |
duke@1 | 43 | * This code and its internal interfaces are subject to change or |
duke@1 | 44 | * deletion without notice.</b> |
duke@1 | 45 | */ |
duke@1 | 46 | public class Scanner implements Lexer { |
duke@1 | 47 | |
duke@1 | 48 | private static boolean scannerDebug = false; |
duke@1 | 49 | |
duke@1 | 50 | /** A factory for creating scanners. */ |
duke@1 | 51 | public static class Factory { |
duke@1 | 52 | /** The context key for the scanner factory. */ |
duke@1 | 53 | public static final Context.Key<Scanner.Factory> scannerFactoryKey = |
duke@1 | 54 | new Context.Key<Scanner.Factory>(); |
duke@1 | 55 | |
duke@1 | 56 | /** Get the Factory instance for this context. */ |
duke@1 | 57 | public static Factory instance(Context context) { |
duke@1 | 58 | Factory instance = context.get(scannerFactoryKey); |
duke@1 | 59 | if (instance == null) |
duke@1 | 60 | instance = new Factory(context); |
duke@1 | 61 | return instance; |
duke@1 | 62 | } |
duke@1 | 63 | |
duke@1 | 64 | final Log log; |
jjg@113 | 65 | final Names names; |
duke@1 | 66 | final Source source; |
duke@1 | 67 | final Keywords keywords; |
duke@1 | 68 | |
duke@1 | 69 | /** Create a new scanner factory. */ |
duke@1 | 70 | protected Factory(Context context) { |
duke@1 | 71 | context.put(scannerFactoryKey, this); |
duke@1 | 72 | this.log = Log.instance(context); |
jjg@113 | 73 | this.names = Names.instance(context); |
duke@1 | 74 | this.source = Source.instance(context); |
duke@1 | 75 | this.keywords = Keywords.instance(context); |
duke@1 | 76 | } |
duke@1 | 77 | |
duke@1 | 78 | public Scanner newScanner(CharSequence input) { |
duke@1 | 79 | if (input instanceof CharBuffer) { |
duke@1 | 80 | return new Scanner(this, (CharBuffer)input); |
duke@1 | 81 | } else { |
duke@1 | 82 | char[] array = input.toString().toCharArray(); |
duke@1 | 83 | return newScanner(array, array.length); |
duke@1 | 84 | } |
duke@1 | 85 | } |
duke@1 | 86 | |
duke@1 | 87 | public Scanner newScanner(char[] input, int inputLength) { |
duke@1 | 88 | return new Scanner(this, input, inputLength); |
duke@1 | 89 | } |
duke@1 | 90 | } |
duke@1 | 91 | |
duke@1 | 92 | /* Output variables; set by nextToken(): |
duke@1 | 93 | */ |
duke@1 | 94 | |
duke@1 | 95 | /** The token, set by nextToken(). |
duke@1 | 96 | */ |
duke@1 | 97 | private Token token; |
duke@1 | 98 | |
duke@1 | 99 | /** Allow hex floating-point literals. |
duke@1 | 100 | */ |
duke@1 | 101 | private boolean allowHexFloats; |
duke@1 | 102 | |
duke@1 | 103 | /** The token's position, 0-based offset from beginning of text. |
duke@1 | 104 | */ |
duke@1 | 105 | private int pos; |
duke@1 | 106 | |
duke@1 | 107 | /** Character position just after the last character of the token. |
duke@1 | 108 | */ |
duke@1 | 109 | private int endPos; |
duke@1 | 110 | |
duke@1 | 111 | /** The last character position of the previous token. |
duke@1 | 112 | */ |
duke@1 | 113 | private int prevEndPos; |
duke@1 | 114 | |
duke@1 | 115 | /** The position where a lexical error occurred; |
duke@1 | 116 | */ |
duke@1 | 117 | private int errPos = Position.NOPOS; |
duke@1 | 118 | |
duke@1 | 119 | /** The name of an identifier or token: |
duke@1 | 120 | */ |
duke@1 | 121 | private Name name; |
duke@1 | 122 | |
duke@1 | 123 | /** The radix of a numeric literal token. |
duke@1 | 124 | */ |
duke@1 | 125 | private int radix; |
duke@1 | 126 | |
duke@1 | 127 | /** Has a @deprecated been encountered in last doc comment? |
duke@1 | 128 | * this needs to be reset by client. |
duke@1 | 129 | */ |
duke@1 | 130 | protected boolean deprecatedFlag = false; |
duke@1 | 131 | |
duke@1 | 132 | /** A character buffer for literals. |
duke@1 | 133 | */ |
duke@1 | 134 | private char[] sbuf = new char[128]; |
duke@1 | 135 | private int sp; |
duke@1 | 136 | |
duke@1 | 137 | /** The input buffer, index of next chacter to be read, |
duke@1 | 138 | * index of one past last character in buffer. |
duke@1 | 139 | */ |
duke@1 | 140 | private char[] buf; |
duke@1 | 141 | private int bp; |
duke@1 | 142 | private int buflen; |
duke@1 | 143 | private int eofPos; |
duke@1 | 144 | |
duke@1 | 145 | /** The current character. |
duke@1 | 146 | */ |
duke@1 | 147 | private char ch; |
duke@1 | 148 | |
duke@1 | 149 | /** The buffer index of the last converted unicode character |
duke@1 | 150 | */ |
duke@1 | 151 | private int unicodeConversionBp = -1; |
duke@1 | 152 | |
duke@1 | 153 | /** The log to be used for error reporting. |
duke@1 | 154 | */ |
duke@1 | 155 | private final Log log; |
duke@1 | 156 | |
duke@1 | 157 | /** The name table. */ |
jjg@113 | 158 | private final Names names; |
duke@1 | 159 | |
duke@1 | 160 | /** The keyword table. */ |
duke@1 | 161 | private final Keywords keywords; |
duke@1 | 162 | |
duke@1 | 163 | /** Common code for constructors. */ |
duke@1 | 164 | private Scanner(Factory fac) { |
duke@1 | 165 | this.log = fac.log; |
duke@1 | 166 | this.names = fac.names; |
duke@1 | 167 | this.keywords = fac.keywords; |
duke@1 | 168 | this.allowHexFloats = fac.source.allowHexFloats(); |
duke@1 | 169 | } |
duke@1 | 170 | |
duke@1 | 171 | private static final boolean hexFloatsWork = hexFloatsWork(); |
duke@1 | 172 | private static boolean hexFloatsWork() { |
duke@1 | 173 | try { |
duke@1 | 174 | Float.valueOf("0x1.0p1"); |
duke@1 | 175 | return true; |
duke@1 | 176 | } catch (NumberFormatException ex) { |
duke@1 | 177 | return false; |
duke@1 | 178 | } |
duke@1 | 179 | } |
duke@1 | 180 | |
duke@1 | 181 | /** Create a scanner from the input buffer. buffer must implement |
duke@1 | 182 | * array() and compact(), and remaining() must be less than limit(). |
duke@1 | 183 | */ |
duke@1 | 184 | protected Scanner(Factory fac, CharBuffer buffer) { |
duke@1 | 185 | this(fac, JavacFileManager.toArray(buffer), buffer.limit()); |
duke@1 | 186 | } |
duke@1 | 187 | |
duke@1 | 188 | /** |
duke@1 | 189 | * Create a scanner from the input array. This method might |
duke@1 | 190 | * modify the array. To avoid copying the input array, ensure |
duke@1 | 191 | * that {@code inputLength < input.length} or |
duke@1 | 192 | * {@code input[input.length -1]} is a white space character. |
duke@1 | 193 | * |
duke@1 | 194 | * @param fac the factory which created this Scanner |
duke@1 | 195 | * @param input the input, might be modified |
duke@1 | 196 | * @param inputLength the size of the input. |
duke@1 | 197 | * Must be positive and less than or equal to input.length. |
duke@1 | 198 | */ |
duke@1 | 199 | protected Scanner(Factory fac, char[] input, int inputLength) { |
duke@1 | 200 | this(fac); |
duke@1 | 201 | eofPos = inputLength; |
duke@1 | 202 | if (inputLength == input.length) { |
duke@1 | 203 | if (input.length > 0 && Character.isWhitespace(input[input.length - 1])) { |
duke@1 | 204 | inputLength--; |
duke@1 | 205 | } else { |
duke@1 | 206 | char[] newInput = new char[inputLength + 1]; |
duke@1 | 207 | System.arraycopy(input, 0, newInput, 0, input.length); |
duke@1 | 208 | input = newInput; |
duke@1 | 209 | } |
duke@1 | 210 | } |
duke@1 | 211 | buf = input; |
duke@1 | 212 | buflen = inputLength; |
duke@1 | 213 | buf[buflen] = EOI; |
duke@1 | 214 | bp = -1; |
duke@1 | 215 | scanChar(); |
duke@1 | 216 | } |
duke@1 | 217 | |
duke@1 | 218 | /** Report an error at the given position using the provided arguments. |
duke@1 | 219 | */ |
duke@1 | 220 | private void lexError(int pos, String key, Object... args) { |
duke@1 | 221 | log.error(pos, key, args); |
duke@1 | 222 | token = ERROR; |
duke@1 | 223 | errPos = pos; |
duke@1 | 224 | } |
duke@1 | 225 | |
duke@1 | 226 | /** Report an error at the current token position using the provided |
duke@1 | 227 | * arguments. |
duke@1 | 228 | */ |
duke@1 | 229 | private void lexError(String key, Object... args) { |
duke@1 | 230 | lexError(pos, key, args); |
duke@1 | 231 | } |
duke@1 | 232 | |
duke@1 | 233 | /** Convert an ASCII digit from its base (8, 10, or 16) |
duke@1 | 234 | * to its value. |
duke@1 | 235 | */ |
duke@1 | 236 | private int digit(int base) { |
duke@1 | 237 | char c = ch; |
duke@1 | 238 | int result = Character.digit(c, base); |
duke@1 | 239 | if (result >= 0 && c > 0x7f) { |
duke@1 | 240 | lexError(pos+1, "illegal.nonascii.digit"); |
duke@1 | 241 | ch = "0123456789abcdef".charAt(result); |
duke@1 | 242 | } |
duke@1 | 243 | return result; |
duke@1 | 244 | } |
duke@1 | 245 | |
duke@1 | 246 | /** Convert unicode escape; bp points to initial '\' character |
duke@1 | 247 | * (Spec 3.3). |
duke@1 | 248 | */ |
duke@1 | 249 | private void convertUnicode() { |
duke@1 | 250 | if (ch == '\\' && unicodeConversionBp != bp) { |
duke@1 | 251 | bp++; ch = buf[bp]; |
duke@1 | 252 | if (ch == 'u') { |
duke@1 | 253 | do { |
duke@1 | 254 | bp++; ch = buf[bp]; |
duke@1 | 255 | } while (ch == 'u'); |
duke@1 | 256 | int limit = bp + 3; |
duke@1 | 257 | if (limit < buflen) { |
duke@1 | 258 | int d = digit(16); |
duke@1 | 259 | int code = d; |
duke@1 | 260 | while (bp < limit && d >= 0) { |
duke@1 | 261 | bp++; ch = buf[bp]; |
duke@1 | 262 | d = digit(16); |
duke@1 | 263 | code = (code << 4) + d; |
duke@1 | 264 | } |
duke@1 | 265 | if (d >= 0) { |
duke@1 | 266 | ch = (char)code; |
duke@1 | 267 | unicodeConversionBp = bp; |
duke@1 | 268 | return; |
duke@1 | 269 | } |
duke@1 | 270 | } |
duke@1 | 271 | lexError(bp, "illegal.unicode.esc"); |
duke@1 | 272 | } else { |
duke@1 | 273 | bp--; |
duke@1 | 274 | ch = '\\'; |
duke@1 | 275 | } |
duke@1 | 276 | } |
duke@1 | 277 | } |
duke@1 | 278 | |
duke@1 | 279 | /** Read next character. |
duke@1 | 280 | */ |
duke@1 | 281 | private void scanChar() { |
duke@1 | 282 | ch = buf[++bp]; |
duke@1 | 283 | if (ch == '\\') { |
duke@1 | 284 | convertUnicode(); |
duke@1 | 285 | } |
duke@1 | 286 | } |
duke@1 | 287 | |
duke@1 | 288 | /** Read next character in comment, skipping over double '\' characters. |
duke@1 | 289 | */ |
duke@1 | 290 | private void scanCommentChar() { |
duke@1 | 291 | scanChar(); |
duke@1 | 292 | if (ch == '\\') { |
duke@1 | 293 | if (buf[bp+1] == '\\' && unicodeConversionBp != bp) { |
duke@1 | 294 | bp++; |
duke@1 | 295 | } else { |
duke@1 | 296 | convertUnicode(); |
duke@1 | 297 | } |
duke@1 | 298 | } |
duke@1 | 299 | } |
duke@1 | 300 | |
duke@1 | 301 | /** Append a character to sbuf. |
duke@1 | 302 | */ |
duke@1 | 303 | private void putChar(char ch) { |
duke@1 | 304 | if (sp == sbuf.length) { |
duke@1 | 305 | char[] newsbuf = new char[sbuf.length * 2]; |
duke@1 | 306 | System.arraycopy(sbuf, 0, newsbuf, 0, sbuf.length); |
duke@1 | 307 | sbuf = newsbuf; |
duke@1 | 308 | } |
duke@1 | 309 | sbuf[sp++] = ch; |
duke@1 | 310 | } |
duke@1 | 311 | |
duke@1 | 312 | /** For debugging purposes: print character. |
duke@1 | 313 | */ |
duke@1 | 314 | private void dch() { |
duke@1 | 315 | System.err.print(ch); System.out.flush(); |
duke@1 | 316 | } |
duke@1 | 317 | |
duke@1 | 318 | /** Read next character in character or string literal and copy into sbuf. |
duke@1 | 319 | */ |
jrose@267 | 320 | private void scanLitChar(boolean forBytecodeName) { |
duke@1 | 321 | if (ch == '\\') { |
duke@1 | 322 | if (buf[bp+1] == '\\' && unicodeConversionBp != bp) { |
duke@1 | 323 | bp++; |
duke@1 | 324 | putChar('\\'); |
duke@1 | 325 | scanChar(); |
duke@1 | 326 | } else { |
duke@1 | 327 | scanChar(); |
duke@1 | 328 | switch (ch) { |
duke@1 | 329 | case '0': case '1': case '2': case '3': |
duke@1 | 330 | case '4': case '5': case '6': case '7': |
duke@1 | 331 | char leadch = ch; |
duke@1 | 332 | int oct = digit(8); |
duke@1 | 333 | scanChar(); |
duke@1 | 334 | if ('0' <= ch && ch <= '7') { |
duke@1 | 335 | oct = oct * 8 + digit(8); |
duke@1 | 336 | scanChar(); |
duke@1 | 337 | if (leadch <= '3' && '0' <= ch && ch <= '7') { |
duke@1 | 338 | oct = oct * 8 + digit(8); |
duke@1 | 339 | scanChar(); |
duke@1 | 340 | } |
duke@1 | 341 | } |
duke@1 | 342 | putChar((char)oct); |
duke@1 | 343 | break; |
duke@1 | 344 | case 'b': |
duke@1 | 345 | putChar('\b'); scanChar(); break; |
duke@1 | 346 | case 't': |
duke@1 | 347 | putChar('\t'); scanChar(); break; |
duke@1 | 348 | case 'n': |
duke@1 | 349 | putChar('\n'); scanChar(); break; |
duke@1 | 350 | case 'f': |
duke@1 | 351 | putChar('\f'); scanChar(); break; |
duke@1 | 352 | case 'r': |
duke@1 | 353 | putChar('\r'); scanChar(); break; |
duke@1 | 354 | case '\'': |
duke@1 | 355 | putChar('\''); scanChar(); break; |
duke@1 | 356 | case '\"': |
duke@1 | 357 | putChar('\"'); scanChar(); break; |
duke@1 | 358 | case '\\': |
duke@1 | 359 | putChar('\\'); scanChar(); break; |
jrose@267 | 360 | case '|': case ',': case '?': case '%': |
jrose@267 | 361 | case '^': case '_': case '{': case '}': |
jrose@267 | 362 | case '!': case '-': case '=': |
jrose@267 | 363 | if (forBytecodeName) { |
jrose@267 | 364 | // Accept escape sequences for dangerous bytecode chars. |
jrose@267 | 365 | // This is illegal in normal Java string or character literals. |
jrose@267 | 366 | // Note that the escape sequence itself is passed through. |
jrose@267 | 367 | putChar('\\'); putChar(ch); scanChar(); |
jrose@267 | 368 | } else { |
jrose@267 | 369 | lexError(bp, "illegal.esc.char"); |
jrose@267 | 370 | } |
jrose@267 | 371 | break; |
duke@1 | 372 | default: |
duke@1 | 373 | lexError(bp, "illegal.esc.char"); |
duke@1 | 374 | } |
duke@1 | 375 | } |
duke@1 | 376 | } else if (bp != buflen) { |
duke@1 | 377 | putChar(ch); scanChar(); |
duke@1 | 378 | } |
duke@1 | 379 | } |
jrose@267 | 380 | private void scanLitChar() { |
jrose@267 | 381 | scanLitChar(false); |
jrose@267 | 382 | } |
jrose@267 | 383 | |
jrose@267 | 384 | /** Read next character in an exotic name #"foo" |
jrose@267 | 385 | */ |
jrose@267 | 386 | private void scanBytecodeNameChar() { |
jrose@267 | 387 | switch (ch) { |
jrose@267 | 388 | // reject any "dangerous" char which is illegal somewhere in the JVM spec |
jrose@267 | 389 | // cf. http://blogs.sun.com/jrose/entry/symbolic_freedom_in_the_vm |
jrose@267 | 390 | case '/': case '.': case ';': // illegal everywhere |
jrose@267 | 391 | case '<': case '>': // illegal in methods, dangerous in classes |
jrose@267 | 392 | case '[': // illegal in classes |
jrose@267 | 393 | lexError(bp, "illegal.bytecode.ident.char", String.valueOf((int)ch)); |
jrose@267 | 394 | break; |
jrose@267 | 395 | } |
jrose@267 | 396 | scanLitChar(true); |
jrose@267 | 397 | } |
duke@1 | 398 | |
duke@1 | 399 | /** Read fractional part of hexadecimal floating point number. |
duke@1 | 400 | */ |
duke@1 | 401 | private void scanHexExponentAndSuffix() { |
duke@1 | 402 | if (ch == 'p' || ch == 'P') { |
duke@1 | 403 | putChar(ch); |
duke@1 | 404 | scanChar(); |
duke@1 | 405 | if (ch == '+' || ch == '-') { |
duke@1 | 406 | putChar(ch); |
duke@1 | 407 | scanChar(); |
duke@1 | 408 | } |
duke@1 | 409 | if ('0' <= ch && ch <= '9') { |
duke@1 | 410 | do { |
duke@1 | 411 | putChar(ch); |
duke@1 | 412 | scanChar(); |
duke@1 | 413 | } while ('0' <= ch && ch <= '9'); |
duke@1 | 414 | if (!allowHexFloats) { |
duke@1 | 415 | lexError("unsupported.fp.lit"); |
duke@1 | 416 | allowHexFloats = true; |
duke@1 | 417 | } |
duke@1 | 418 | else if (!hexFloatsWork) |
duke@1 | 419 | lexError("unsupported.cross.fp.lit"); |
duke@1 | 420 | } else |
duke@1 | 421 | lexError("malformed.fp.lit"); |
duke@1 | 422 | } else { |
duke@1 | 423 | lexError("malformed.fp.lit"); |
duke@1 | 424 | } |
duke@1 | 425 | if (ch == 'f' || ch == 'F') { |
duke@1 | 426 | putChar(ch); |
duke@1 | 427 | scanChar(); |
duke@1 | 428 | token = FLOATLITERAL; |
duke@1 | 429 | } else { |
duke@1 | 430 | if (ch == 'd' || ch == 'D') { |
duke@1 | 431 | putChar(ch); |
duke@1 | 432 | scanChar(); |
duke@1 | 433 | } |
duke@1 | 434 | token = DOUBLELITERAL; |
duke@1 | 435 | } |
duke@1 | 436 | } |
duke@1 | 437 | |
duke@1 | 438 | /** Read fractional part of floating point number. |
duke@1 | 439 | */ |
duke@1 | 440 | private void scanFraction() { |
duke@1 | 441 | while (digit(10) >= 0) { |
duke@1 | 442 | putChar(ch); |
duke@1 | 443 | scanChar(); |
duke@1 | 444 | } |
duke@1 | 445 | int sp1 = sp; |
duke@1 | 446 | if (ch == 'e' || ch == 'E') { |
duke@1 | 447 | putChar(ch); |
duke@1 | 448 | scanChar(); |
duke@1 | 449 | if (ch == '+' || ch == '-') { |
duke@1 | 450 | putChar(ch); |
duke@1 | 451 | scanChar(); |
duke@1 | 452 | } |
duke@1 | 453 | if ('0' <= ch && ch <= '9') { |
duke@1 | 454 | do { |
duke@1 | 455 | putChar(ch); |
duke@1 | 456 | scanChar(); |
duke@1 | 457 | } while ('0' <= ch && ch <= '9'); |
duke@1 | 458 | return; |
duke@1 | 459 | } |
duke@1 | 460 | lexError("malformed.fp.lit"); |
duke@1 | 461 | sp = sp1; |
duke@1 | 462 | } |
duke@1 | 463 | } |
duke@1 | 464 | |
duke@1 | 465 | /** Read fractional part and 'd' or 'f' suffix of floating point number. |
duke@1 | 466 | */ |
duke@1 | 467 | private void scanFractionAndSuffix() { |
duke@1 | 468 | this.radix = 10; |
duke@1 | 469 | scanFraction(); |
duke@1 | 470 | if (ch == 'f' || ch == 'F') { |
duke@1 | 471 | putChar(ch); |
duke@1 | 472 | scanChar(); |
duke@1 | 473 | token = FLOATLITERAL; |
duke@1 | 474 | } else { |
duke@1 | 475 | if (ch == 'd' || ch == 'D') { |
duke@1 | 476 | putChar(ch); |
duke@1 | 477 | scanChar(); |
duke@1 | 478 | } |
duke@1 | 479 | token = DOUBLELITERAL; |
duke@1 | 480 | } |
duke@1 | 481 | } |
duke@1 | 482 | |
duke@1 | 483 | /** Read fractional part and 'd' or 'f' suffix of floating point number. |
duke@1 | 484 | */ |
duke@1 | 485 | private void scanHexFractionAndSuffix(boolean seendigit) { |
duke@1 | 486 | this.radix = 16; |
duke@1 | 487 | assert ch == '.'; |
duke@1 | 488 | putChar(ch); |
duke@1 | 489 | scanChar(); |
duke@1 | 490 | while (digit(16) >= 0) { |
duke@1 | 491 | seendigit = true; |
duke@1 | 492 | putChar(ch); |
duke@1 | 493 | scanChar(); |
duke@1 | 494 | } |
duke@1 | 495 | if (!seendigit) |
duke@1 | 496 | lexError("invalid.hex.number"); |
duke@1 | 497 | else |
duke@1 | 498 | scanHexExponentAndSuffix(); |
duke@1 | 499 | } |
duke@1 | 500 | |
duke@1 | 501 | /** Read a number. |
duke@1 | 502 | * @param radix The radix of the number; one of 8, 10, 16. |
duke@1 | 503 | */ |
duke@1 | 504 | private void scanNumber(int radix) { |
duke@1 | 505 | this.radix = radix; |
duke@1 | 506 | // for octal, allow base-10 digit in case it's a float literal |
duke@1 | 507 | int digitRadix = (radix <= 10) ? 10 : 16; |
duke@1 | 508 | boolean seendigit = false; |
duke@1 | 509 | while (digit(digitRadix) >= 0) { |
duke@1 | 510 | seendigit = true; |
duke@1 | 511 | putChar(ch); |
duke@1 | 512 | scanChar(); |
duke@1 | 513 | } |
duke@1 | 514 | if (radix == 16 && ch == '.') { |
duke@1 | 515 | scanHexFractionAndSuffix(seendigit); |
duke@1 | 516 | } else if (seendigit && radix == 16 && (ch == 'p' || ch == 'P')) { |
duke@1 | 517 | scanHexExponentAndSuffix(); |
duke@1 | 518 | } else if (radix <= 10 && ch == '.') { |
duke@1 | 519 | putChar(ch); |
duke@1 | 520 | scanChar(); |
duke@1 | 521 | scanFractionAndSuffix(); |
duke@1 | 522 | } else if (radix <= 10 && |
duke@1 | 523 | (ch == 'e' || ch == 'E' || |
duke@1 | 524 | ch == 'f' || ch == 'F' || |
duke@1 | 525 | ch == 'd' || ch == 'D')) { |
duke@1 | 526 | scanFractionAndSuffix(); |
duke@1 | 527 | } else { |
duke@1 | 528 | if (ch == 'l' || ch == 'L') { |
duke@1 | 529 | scanChar(); |
duke@1 | 530 | token = LONGLITERAL; |
duke@1 | 531 | } else { |
duke@1 | 532 | token = INTLITERAL; |
duke@1 | 533 | } |
duke@1 | 534 | } |
duke@1 | 535 | } |
duke@1 | 536 | |
duke@1 | 537 | /** Read an identifier. |
duke@1 | 538 | */ |
duke@1 | 539 | private void scanIdent() { |
duke@1 | 540 | boolean isJavaIdentifierPart; |
duke@1 | 541 | char high; |
duke@1 | 542 | do { |
duke@1 | 543 | if (sp == sbuf.length) putChar(ch); else sbuf[sp++] = ch; |
duke@1 | 544 | // optimization, was: putChar(ch); |
duke@1 | 545 | |
duke@1 | 546 | scanChar(); |
duke@1 | 547 | switch (ch) { |
duke@1 | 548 | case 'A': case 'B': case 'C': case 'D': case 'E': |
duke@1 | 549 | case 'F': case 'G': case 'H': case 'I': case 'J': |
duke@1 | 550 | case 'K': case 'L': case 'M': case 'N': case 'O': |
duke@1 | 551 | case 'P': case 'Q': case 'R': case 'S': case 'T': |
duke@1 | 552 | case 'U': case 'V': case 'W': case 'X': case 'Y': |
duke@1 | 553 | case 'Z': |
duke@1 | 554 | case 'a': case 'b': case 'c': case 'd': case 'e': |
duke@1 | 555 | case 'f': case 'g': case 'h': case 'i': case 'j': |
duke@1 | 556 | case 'k': case 'l': case 'm': case 'n': case 'o': |
duke@1 | 557 | case 'p': case 'q': case 'r': case 's': case 't': |
duke@1 | 558 | case 'u': case 'v': case 'w': case 'x': case 'y': |
duke@1 | 559 | case 'z': |
duke@1 | 560 | case '$': case '_': |
duke@1 | 561 | case '0': case '1': case '2': case '3': case '4': |
duke@1 | 562 | case '5': case '6': case '7': case '8': case '9': |
duke@1 | 563 | case '\u0000': case '\u0001': case '\u0002': case '\u0003': |
duke@1 | 564 | case '\u0004': case '\u0005': case '\u0006': case '\u0007': |
duke@1 | 565 | case '\u0008': case '\u000E': case '\u000F': case '\u0010': |
duke@1 | 566 | case '\u0011': case '\u0012': case '\u0013': case '\u0014': |
duke@1 | 567 | case '\u0015': case '\u0016': case '\u0017': |
duke@1 | 568 | case '\u0018': case '\u0019': case '\u001B': |
duke@1 | 569 | case '\u007F': |
duke@1 | 570 | break; |
duke@1 | 571 | case '\u001A': // EOI is also a legal identifier part |
duke@1 | 572 | if (bp >= buflen) { |
duke@1 | 573 | name = names.fromChars(sbuf, 0, sp); |
duke@1 | 574 | token = keywords.key(name); |
duke@1 | 575 | return; |
duke@1 | 576 | } |
duke@1 | 577 | break; |
duke@1 | 578 | default: |
duke@1 | 579 | if (ch < '\u0080') { |
duke@1 | 580 | // all ASCII range chars already handled, above |
duke@1 | 581 | isJavaIdentifierPart = false; |
duke@1 | 582 | } else { |
duke@1 | 583 | high = scanSurrogates(); |
duke@1 | 584 | if (high != 0) { |
duke@1 | 585 | if (sp == sbuf.length) { |
duke@1 | 586 | putChar(high); |
duke@1 | 587 | } else { |
duke@1 | 588 | sbuf[sp++] = high; |
duke@1 | 589 | } |
duke@1 | 590 | isJavaIdentifierPart = Character.isJavaIdentifierPart( |
duke@1 | 591 | Character.toCodePoint(high, ch)); |
duke@1 | 592 | } else { |
duke@1 | 593 | isJavaIdentifierPart = Character.isJavaIdentifierPart(ch); |
duke@1 | 594 | } |
duke@1 | 595 | } |
duke@1 | 596 | if (!isJavaIdentifierPart) { |
duke@1 | 597 | name = names.fromChars(sbuf, 0, sp); |
duke@1 | 598 | token = keywords.key(name); |
duke@1 | 599 | return; |
duke@1 | 600 | } |
duke@1 | 601 | } |
duke@1 | 602 | } while (true); |
duke@1 | 603 | } |
duke@1 | 604 | |
duke@1 | 605 | /** Are surrogates supported? |
duke@1 | 606 | */ |
duke@1 | 607 | final static boolean surrogatesSupported = surrogatesSupported(); |
duke@1 | 608 | private static boolean surrogatesSupported() { |
duke@1 | 609 | try { |
duke@1 | 610 | Character.isHighSurrogate('a'); |
duke@1 | 611 | return true; |
duke@1 | 612 | } catch (NoSuchMethodError ex) { |
duke@1 | 613 | return false; |
duke@1 | 614 | } |
duke@1 | 615 | } |
duke@1 | 616 | |
duke@1 | 617 | /** Scan surrogate pairs. If 'ch' is a high surrogate and |
duke@1 | 618 | * the next character is a low surrogate, then put the low |
duke@1 | 619 | * surrogate in 'ch', and return the high surrogate. |
duke@1 | 620 | * otherwise, just return 0. |
duke@1 | 621 | */ |
duke@1 | 622 | private char scanSurrogates() { |
duke@1 | 623 | if (surrogatesSupported && Character.isHighSurrogate(ch)) { |
duke@1 | 624 | char high = ch; |
duke@1 | 625 | |
duke@1 | 626 | scanChar(); |
duke@1 | 627 | |
duke@1 | 628 | if (Character.isLowSurrogate(ch)) { |
duke@1 | 629 | return high; |
duke@1 | 630 | } |
duke@1 | 631 | |
duke@1 | 632 | ch = high; |
duke@1 | 633 | } |
duke@1 | 634 | |
duke@1 | 635 | return 0; |
duke@1 | 636 | } |
duke@1 | 637 | |
duke@1 | 638 | /** Return true if ch can be part of an operator. |
duke@1 | 639 | */ |
duke@1 | 640 | private boolean isSpecial(char ch) { |
duke@1 | 641 | switch (ch) { |
duke@1 | 642 | case '!': case '%': case '&': case '*': case '?': |
duke@1 | 643 | case '+': case '-': case ':': case '<': case '=': |
duke@1 | 644 | case '>': case '^': case '|': case '~': |
duke@1 | 645 | case '@': |
duke@1 | 646 | return true; |
duke@1 | 647 | default: |
duke@1 | 648 | return false; |
duke@1 | 649 | } |
duke@1 | 650 | } |
duke@1 | 651 | |
duke@1 | 652 | /** Read longest possible sequence of special characters and convert |
duke@1 | 653 | * to token. |
duke@1 | 654 | */ |
duke@1 | 655 | private void scanOperator() { |
duke@1 | 656 | while (true) { |
duke@1 | 657 | putChar(ch); |
duke@1 | 658 | Name newname = names.fromChars(sbuf, 0, sp); |
duke@1 | 659 | if (keywords.key(newname) == IDENTIFIER) { |
duke@1 | 660 | sp--; |
duke@1 | 661 | break; |
duke@1 | 662 | } |
duke@1 | 663 | name = newname; |
duke@1 | 664 | token = keywords.key(newname); |
duke@1 | 665 | scanChar(); |
duke@1 | 666 | if (!isSpecial(ch)) break; |
duke@1 | 667 | } |
duke@1 | 668 | } |
duke@1 | 669 | |
duke@1 | 670 | /** |
duke@1 | 671 | * Scan a documention comment; determine if a deprecated tag is present. |
duke@1 | 672 | * Called once the initial /, * have been skipped, positioned at the second * |
duke@1 | 673 | * (which is treated as the beginning of the first line). |
duke@1 | 674 | * Stops positioned at the closing '/'. |
duke@1 | 675 | */ |
duke@1 | 676 | @SuppressWarnings("fallthrough") |
duke@1 | 677 | private void scanDocComment() { |
duke@1 | 678 | boolean deprecatedPrefix = false; |
duke@1 | 679 | |
duke@1 | 680 | forEachLine: |
duke@1 | 681 | while (bp < buflen) { |
duke@1 | 682 | |
duke@1 | 683 | // Skip optional WhiteSpace at beginning of line |
duke@1 | 684 | while (bp < buflen && (ch == ' ' || ch == '\t' || ch == FF)) { |
duke@1 | 685 | scanCommentChar(); |
duke@1 | 686 | } |
duke@1 | 687 | |
duke@1 | 688 | // Skip optional consecutive Stars |
duke@1 | 689 | while (bp < buflen && ch == '*') { |
duke@1 | 690 | scanCommentChar(); |
duke@1 | 691 | if (ch == '/') { |
duke@1 | 692 | return; |
duke@1 | 693 | } |
duke@1 | 694 | } |
duke@1 | 695 | |
duke@1 | 696 | // Skip optional WhiteSpace after Stars |
duke@1 | 697 | while (bp < buflen && (ch == ' ' || ch == '\t' || ch == FF)) { |
duke@1 | 698 | scanCommentChar(); |
duke@1 | 699 | } |
duke@1 | 700 | |
duke@1 | 701 | deprecatedPrefix = false; |
duke@1 | 702 | // At beginning of line in the JavaDoc sense. |
duke@1 | 703 | if (bp < buflen && ch == '@' && !deprecatedFlag) { |
duke@1 | 704 | scanCommentChar(); |
duke@1 | 705 | if (bp < buflen && ch == 'd') { |
duke@1 | 706 | scanCommentChar(); |
duke@1 | 707 | if (bp < buflen && ch == 'e') { |
duke@1 | 708 | scanCommentChar(); |
duke@1 | 709 | if (bp < buflen && ch == 'p') { |
duke@1 | 710 | scanCommentChar(); |
duke@1 | 711 | if (bp < buflen && ch == 'r') { |
duke@1 | 712 | scanCommentChar(); |
duke@1 | 713 | if (bp < buflen && ch == 'e') { |
duke@1 | 714 | scanCommentChar(); |
duke@1 | 715 | if (bp < buflen && ch == 'c') { |
duke@1 | 716 | scanCommentChar(); |
duke@1 | 717 | if (bp < buflen && ch == 'a') { |
duke@1 | 718 | scanCommentChar(); |
duke@1 | 719 | if (bp < buflen && ch == 't') { |
duke@1 | 720 | scanCommentChar(); |
duke@1 | 721 | if (bp < buflen && ch == 'e') { |
duke@1 | 722 | scanCommentChar(); |
duke@1 | 723 | if (bp < buflen && ch == 'd') { |
duke@1 | 724 | deprecatedPrefix = true; |
duke@1 | 725 | scanCommentChar(); |
duke@1 | 726 | }}}}}}}}}}} |
duke@1 | 727 | if (deprecatedPrefix && bp < buflen) { |
duke@1 | 728 | if (Character.isWhitespace(ch)) { |
duke@1 | 729 | deprecatedFlag = true; |
duke@1 | 730 | } else if (ch == '*') { |
duke@1 | 731 | scanCommentChar(); |
duke@1 | 732 | if (ch == '/') { |
duke@1 | 733 | deprecatedFlag = true; |
duke@1 | 734 | return; |
duke@1 | 735 | } |
duke@1 | 736 | } |
duke@1 | 737 | } |
duke@1 | 738 | |
duke@1 | 739 | // Skip rest of line |
duke@1 | 740 | while (bp < buflen) { |
duke@1 | 741 | switch (ch) { |
duke@1 | 742 | case '*': |
duke@1 | 743 | scanCommentChar(); |
duke@1 | 744 | if (ch == '/') { |
duke@1 | 745 | return; |
duke@1 | 746 | } |
duke@1 | 747 | break; |
duke@1 | 748 | case CR: // (Spec 3.4) |
duke@1 | 749 | scanCommentChar(); |
duke@1 | 750 | if (ch != LF) { |
duke@1 | 751 | continue forEachLine; |
duke@1 | 752 | } |
duke@1 | 753 | /* fall through to LF case */ |
duke@1 | 754 | case LF: // (Spec 3.4) |
duke@1 | 755 | scanCommentChar(); |
duke@1 | 756 | continue forEachLine; |
duke@1 | 757 | default: |
duke@1 | 758 | scanCommentChar(); |
duke@1 | 759 | } |
duke@1 | 760 | } // rest of line |
duke@1 | 761 | } // forEachLine |
duke@1 | 762 | return; |
duke@1 | 763 | } |
duke@1 | 764 | |
duke@1 | 765 | /** The value of a literal token, recorded as a string. |
duke@1 | 766 | * For integers, leading 0x and 'l' suffixes are suppressed. |
duke@1 | 767 | */ |
duke@1 | 768 | public String stringVal() { |
duke@1 | 769 | return new String(sbuf, 0, sp); |
duke@1 | 770 | } |
duke@1 | 771 | |
duke@1 | 772 | /** Read token. |
duke@1 | 773 | */ |
duke@1 | 774 | public void nextToken() { |
duke@1 | 775 | |
duke@1 | 776 | try { |
duke@1 | 777 | prevEndPos = endPos; |
duke@1 | 778 | sp = 0; |
duke@1 | 779 | |
duke@1 | 780 | while (true) { |
duke@1 | 781 | pos = bp; |
duke@1 | 782 | switch (ch) { |
duke@1 | 783 | case ' ': // (Spec 3.6) |
duke@1 | 784 | case '\t': // (Spec 3.6) |
duke@1 | 785 | case FF: // (Spec 3.6) |
duke@1 | 786 | do { |
duke@1 | 787 | scanChar(); |
duke@1 | 788 | } while (ch == ' ' || ch == '\t' || ch == FF); |
duke@1 | 789 | endPos = bp; |
duke@1 | 790 | processWhiteSpace(); |
duke@1 | 791 | break; |
duke@1 | 792 | case LF: // (Spec 3.4) |
duke@1 | 793 | scanChar(); |
duke@1 | 794 | endPos = bp; |
duke@1 | 795 | processLineTerminator(); |
duke@1 | 796 | break; |
duke@1 | 797 | case CR: // (Spec 3.4) |
duke@1 | 798 | scanChar(); |
duke@1 | 799 | if (ch == LF) { |
duke@1 | 800 | scanChar(); |
duke@1 | 801 | } |
duke@1 | 802 | endPos = bp; |
duke@1 | 803 | processLineTerminator(); |
duke@1 | 804 | break; |
duke@1 | 805 | case 'A': case 'B': case 'C': case 'D': case 'E': |
duke@1 | 806 | case 'F': case 'G': case 'H': case 'I': case 'J': |
duke@1 | 807 | case 'K': case 'L': case 'M': case 'N': case 'O': |
duke@1 | 808 | case 'P': case 'Q': case 'R': case 'S': case 'T': |
duke@1 | 809 | case 'U': case 'V': case 'W': case 'X': case 'Y': |
duke@1 | 810 | case 'Z': |
duke@1 | 811 | case 'a': case 'b': case 'c': case 'd': case 'e': |
duke@1 | 812 | case 'f': case 'g': case 'h': case 'i': case 'j': |
duke@1 | 813 | case 'k': case 'l': case 'm': case 'n': case 'o': |
duke@1 | 814 | case 'p': case 'q': case 'r': case 's': case 't': |
duke@1 | 815 | case 'u': case 'v': case 'w': case 'x': case 'y': |
duke@1 | 816 | case 'z': |
duke@1 | 817 | case '$': case '_': |
duke@1 | 818 | scanIdent(); |
duke@1 | 819 | return; |
duke@1 | 820 | case '0': |
duke@1 | 821 | scanChar(); |
duke@1 | 822 | if (ch == 'x' || ch == 'X') { |
duke@1 | 823 | scanChar(); |
duke@1 | 824 | if (ch == '.') { |
duke@1 | 825 | scanHexFractionAndSuffix(false); |
duke@1 | 826 | } else if (digit(16) < 0) { |
duke@1 | 827 | lexError("invalid.hex.number"); |
duke@1 | 828 | } else { |
duke@1 | 829 | scanNumber(16); |
duke@1 | 830 | } |
duke@1 | 831 | } else { |
duke@1 | 832 | putChar('0'); |
duke@1 | 833 | scanNumber(8); |
duke@1 | 834 | } |
duke@1 | 835 | return; |
duke@1 | 836 | case '1': case '2': case '3': case '4': |
duke@1 | 837 | case '5': case '6': case '7': case '8': case '9': |
duke@1 | 838 | scanNumber(10); |
duke@1 | 839 | return; |
duke@1 | 840 | case '.': |
duke@1 | 841 | scanChar(); |
duke@1 | 842 | if ('0' <= ch && ch <= '9') { |
duke@1 | 843 | putChar('.'); |
duke@1 | 844 | scanFractionAndSuffix(); |
duke@1 | 845 | } else if (ch == '.') { |
duke@1 | 846 | putChar('.'); putChar('.'); |
duke@1 | 847 | scanChar(); |
duke@1 | 848 | if (ch == '.') { |
duke@1 | 849 | scanChar(); |
duke@1 | 850 | putChar('.'); |
duke@1 | 851 | token = ELLIPSIS; |
duke@1 | 852 | } else { |
duke@1 | 853 | lexError("malformed.fp.lit"); |
duke@1 | 854 | } |
duke@1 | 855 | } else { |
duke@1 | 856 | token = DOT; |
duke@1 | 857 | } |
duke@1 | 858 | return; |
duke@1 | 859 | case ',': |
duke@1 | 860 | scanChar(); token = COMMA; return; |
duke@1 | 861 | case ';': |
duke@1 | 862 | scanChar(); token = SEMI; return; |
duke@1 | 863 | case '(': |
duke@1 | 864 | scanChar(); token = LPAREN; return; |
duke@1 | 865 | case ')': |
duke@1 | 866 | scanChar(); token = RPAREN; return; |
duke@1 | 867 | case '[': |
duke@1 | 868 | scanChar(); token = LBRACKET; return; |
duke@1 | 869 | case ']': |
duke@1 | 870 | scanChar(); token = RBRACKET; return; |
duke@1 | 871 | case '{': |
duke@1 | 872 | scanChar(); token = LBRACE; return; |
duke@1 | 873 | case '}': |
duke@1 | 874 | scanChar(); token = RBRACE; return; |
duke@1 | 875 | case '/': |
duke@1 | 876 | scanChar(); |
duke@1 | 877 | if (ch == '/') { |
duke@1 | 878 | do { |
duke@1 | 879 | scanCommentChar(); |
duke@1 | 880 | } while (ch != CR && ch != LF && bp < buflen); |
duke@1 | 881 | if (bp < buflen) { |
duke@1 | 882 | endPos = bp; |
duke@1 | 883 | processComment(CommentStyle.LINE); |
duke@1 | 884 | } |
duke@1 | 885 | break; |
duke@1 | 886 | } else if (ch == '*') { |
duke@1 | 887 | scanChar(); |
duke@1 | 888 | CommentStyle style; |
duke@1 | 889 | if (ch == '*') { |
duke@1 | 890 | style = CommentStyle.JAVADOC; |
duke@1 | 891 | scanDocComment(); |
duke@1 | 892 | } else { |
duke@1 | 893 | style = CommentStyle.BLOCK; |
duke@1 | 894 | while (bp < buflen) { |
duke@1 | 895 | if (ch == '*') { |
duke@1 | 896 | scanChar(); |
duke@1 | 897 | if (ch == '/') break; |
duke@1 | 898 | } else { |
duke@1 | 899 | scanCommentChar(); |
duke@1 | 900 | } |
duke@1 | 901 | } |
duke@1 | 902 | } |
duke@1 | 903 | if (ch == '/') { |
duke@1 | 904 | scanChar(); |
duke@1 | 905 | endPos = bp; |
duke@1 | 906 | processComment(style); |
duke@1 | 907 | break; |
duke@1 | 908 | } else { |
duke@1 | 909 | lexError("unclosed.comment"); |
duke@1 | 910 | return; |
duke@1 | 911 | } |
duke@1 | 912 | } else if (ch == '=') { |
duke@1 | 913 | name = names.slashequals; |
duke@1 | 914 | token = SLASHEQ; |
duke@1 | 915 | scanChar(); |
duke@1 | 916 | } else { |
duke@1 | 917 | name = names.slash; |
duke@1 | 918 | token = SLASH; |
duke@1 | 919 | } |
duke@1 | 920 | return; |
duke@1 | 921 | case '\'': |
duke@1 | 922 | scanChar(); |
duke@1 | 923 | if (ch == '\'') { |
duke@1 | 924 | lexError("empty.char.lit"); |
duke@1 | 925 | } else { |
duke@1 | 926 | if (ch == CR || ch == LF) |
duke@1 | 927 | lexError(pos, "illegal.line.end.in.char.lit"); |
duke@1 | 928 | scanLitChar(); |
duke@1 | 929 | if (ch == '\'') { |
duke@1 | 930 | scanChar(); |
duke@1 | 931 | token = CHARLITERAL; |
duke@1 | 932 | } else { |
duke@1 | 933 | lexError(pos, "unclosed.char.lit"); |
duke@1 | 934 | } |
duke@1 | 935 | } |
duke@1 | 936 | return; |
duke@1 | 937 | case '\"': |
duke@1 | 938 | scanChar(); |
duke@1 | 939 | while (ch != '\"' && ch != CR && ch != LF && bp < buflen) |
duke@1 | 940 | scanLitChar(); |
duke@1 | 941 | if (ch == '\"') { |
duke@1 | 942 | token = STRINGLITERAL; |
duke@1 | 943 | scanChar(); |
duke@1 | 944 | } else { |
duke@1 | 945 | lexError(pos, "unclosed.str.lit"); |
duke@1 | 946 | } |
duke@1 | 947 | return; |
jrose@267 | 948 | case '#': |
jrose@267 | 949 | scanChar(); |
jrose@267 | 950 | if (ch == '\"') { |
jrose@267 | 951 | scanChar(); |
jrose@267 | 952 | if (ch == '\"') |
jrose@267 | 953 | lexError(pos, "empty.bytecode.ident"); |
jrose@267 | 954 | while (ch != '\"' && ch != CR && ch != LF && bp < buflen) { |
jrose@267 | 955 | scanBytecodeNameChar(); |
jrose@267 | 956 | } |
jrose@267 | 957 | if (ch == '\"') { |
jrose@267 | 958 | name = names.fromChars(sbuf, 0, sp); |
jrose@267 | 959 | token = IDENTIFIER; // even if #"int" or #"do" |
jrose@267 | 960 | scanChar(); |
jrose@267 | 961 | } else { |
jrose@267 | 962 | lexError(pos, "unclosed.bytecode.ident"); |
jrose@267 | 963 | } |
jrose@267 | 964 | } else { |
jrose@267 | 965 | lexError("illegal.char", String.valueOf((int)'#')); |
jrose@267 | 966 | } |
jrose@267 | 967 | return; |
duke@1 | 968 | default: |
duke@1 | 969 | if (isSpecial(ch)) { |
duke@1 | 970 | scanOperator(); |
duke@1 | 971 | } else { |
duke@1 | 972 | boolean isJavaIdentifierStart; |
duke@1 | 973 | if (ch < '\u0080') { |
duke@1 | 974 | // all ASCII range chars already handled, above |
duke@1 | 975 | isJavaIdentifierStart = false; |
duke@1 | 976 | } else { |
duke@1 | 977 | char high = scanSurrogates(); |
duke@1 | 978 | if (high != 0) { |
duke@1 | 979 | if (sp == sbuf.length) { |
duke@1 | 980 | putChar(high); |
duke@1 | 981 | } else { |
duke@1 | 982 | sbuf[sp++] = high; |
duke@1 | 983 | } |
duke@1 | 984 | |
duke@1 | 985 | isJavaIdentifierStart = Character.isJavaIdentifierStart( |
duke@1 | 986 | Character.toCodePoint(high, ch)); |
duke@1 | 987 | } else { |
duke@1 | 988 | isJavaIdentifierStart = Character.isJavaIdentifierStart(ch); |
duke@1 | 989 | } |
duke@1 | 990 | } |
duke@1 | 991 | if (isJavaIdentifierStart) { |
duke@1 | 992 | scanIdent(); |
duke@1 | 993 | } else if (bp == buflen || ch == EOI && bp+1 == buflen) { // JLS 3.5 |
duke@1 | 994 | token = EOF; |
duke@1 | 995 | pos = bp = eofPos; |
duke@1 | 996 | } else { |
duke@1 | 997 | lexError("illegal.char", String.valueOf((int)ch)); |
duke@1 | 998 | scanChar(); |
duke@1 | 999 | } |
duke@1 | 1000 | } |
duke@1 | 1001 | return; |
duke@1 | 1002 | } |
duke@1 | 1003 | } |
duke@1 | 1004 | } finally { |
duke@1 | 1005 | endPos = bp; |
duke@1 | 1006 | if (scannerDebug) |
duke@1 | 1007 | System.out.println("nextToken(" + pos |
duke@1 | 1008 | + "," + endPos + ")=|" + |
duke@1 | 1009 | new String(getRawCharacters(pos, endPos)) |
duke@1 | 1010 | + "|"); |
duke@1 | 1011 | } |
duke@1 | 1012 | } |
duke@1 | 1013 | |
duke@1 | 1014 | /** Return the current token, set by nextToken(). |
duke@1 | 1015 | */ |
duke@1 | 1016 | public Token token() { |
duke@1 | 1017 | return token; |
duke@1 | 1018 | } |
duke@1 | 1019 | |
duke@1 | 1020 | /** Sets the current token. |
duke@1 | 1021 | */ |
duke@1 | 1022 | public void token(Token token) { |
duke@1 | 1023 | this.token = token; |
duke@1 | 1024 | } |
duke@1 | 1025 | |
duke@1 | 1026 | /** Return the current token's position: a 0-based |
duke@1 | 1027 | * offset from beginning of the raw input stream |
duke@1 | 1028 | * (before unicode translation) |
duke@1 | 1029 | */ |
duke@1 | 1030 | public int pos() { |
duke@1 | 1031 | return pos; |
duke@1 | 1032 | } |
duke@1 | 1033 | |
duke@1 | 1034 | /** Return the last character position of the current token. |
duke@1 | 1035 | */ |
duke@1 | 1036 | public int endPos() { |
duke@1 | 1037 | return endPos; |
duke@1 | 1038 | } |
duke@1 | 1039 | |
duke@1 | 1040 | /** Return the last character position of the previous token. |
duke@1 | 1041 | */ |
duke@1 | 1042 | public int prevEndPos() { |
duke@1 | 1043 | return prevEndPos; |
duke@1 | 1044 | } |
duke@1 | 1045 | |
duke@1 | 1046 | /** Return the position where a lexical error occurred; |
duke@1 | 1047 | */ |
duke@1 | 1048 | public int errPos() { |
duke@1 | 1049 | return errPos; |
duke@1 | 1050 | } |
duke@1 | 1051 | |
duke@1 | 1052 | /** Set the position where a lexical error occurred; |
duke@1 | 1053 | */ |
duke@1 | 1054 | public void errPos(int pos) { |
duke@1 | 1055 | errPos = pos; |
duke@1 | 1056 | } |
duke@1 | 1057 | |
duke@1 | 1058 | /** Return the name of an identifier or token for the current token. |
duke@1 | 1059 | */ |
duke@1 | 1060 | public Name name() { |
duke@1 | 1061 | return name; |
duke@1 | 1062 | } |
duke@1 | 1063 | |
duke@1 | 1064 | /** Return the radix of a numeric literal token. |
duke@1 | 1065 | */ |
duke@1 | 1066 | public int radix() { |
duke@1 | 1067 | return radix; |
duke@1 | 1068 | } |
duke@1 | 1069 | |
duke@1 | 1070 | /** Has a @deprecated been encountered in last doc comment? |
duke@1 | 1071 | * This needs to be reset by client with resetDeprecatedFlag. |
duke@1 | 1072 | */ |
duke@1 | 1073 | public boolean deprecatedFlag() { |
duke@1 | 1074 | return deprecatedFlag; |
duke@1 | 1075 | } |
duke@1 | 1076 | |
duke@1 | 1077 | public void resetDeprecatedFlag() { |
duke@1 | 1078 | deprecatedFlag = false; |
duke@1 | 1079 | } |
duke@1 | 1080 | |
duke@1 | 1081 | /** |
duke@1 | 1082 | * Returns the documentation string of the current token. |
duke@1 | 1083 | */ |
duke@1 | 1084 | public String docComment() { |
duke@1 | 1085 | return null; |
duke@1 | 1086 | } |
duke@1 | 1087 | |
duke@1 | 1088 | /** |
duke@1 | 1089 | * Returns a copy of the input buffer, up to its inputLength. |
duke@1 | 1090 | * Unicode escape sequences are not translated. |
duke@1 | 1091 | */ |
duke@1 | 1092 | public char[] getRawCharacters() { |
duke@1 | 1093 | char[] chars = new char[buflen]; |
duke@1 | 1094 | System.arraycopy(buf, 0, chars, 0, buflen); |
duke@1 | 1095 | return chars; |
duke@1 | 1096 | } |
duke@1 | 1097 | |
duke@1 | 1098 | /** |
duke@1 | 1099 | * Returns a copy of a character array subset of the input buffer. |
duke@1 | 1100 | * The returned array begins at the <code>beginIndex</code> and |
duke@1 | 1101 | * extends to the character at index <code>endIndex - 1</code>. |
duke@1 | 1102 | * Thus the length of the substring is <code>endIndex-beginIndex</code>. |
duke@1 | 1103 | * This behavior is like |
duke@1 | 1104 | * <code>String.substring(beginIndex, endIndex)</code>. |
duke@1 | 1105 | * Unicode escape sequences are not translated. |
duke@1 | 1106 | * |
duke@1 | 1107 | * @param beginIndex the beginning index, inclusive. |
duke@1 | 1108 | * @param endIndex the ending index, exclusive. |
duke@1 | 1109 | * @throws IndexOutOfBounds if either offset is outside of the |
duke@1 | 1110 | * array bounds |
duke@1 | 1111 | */ |
duke@1 | 1112 | public char[] getRawCharacters(int beginIndex, int endIndex) { |
duke@1 | 1113 | int length = endIndex - beginIndex; |
duke@1 | 1114 | char[] chars = new char[length]; |
duke@1 | 1115 | System.arraycopy(buf, beginIndex, chars, 0, length); |
duke@1 | 1116 | return chars; |
duke@1 | 1117 | } |
duke@1 | 1118 | |
duke@1 | 1119 | public enum CommentStyle { |
duke@1 | 1120 | LINE, |
duke@1 | 1121 | BLOCK, |
duke@1 | 1122 | JAVADOC, |
duke@1 | 1123 | } |
duke@1 | 1124 | |
duke@1 | 1125 | /** |
duke@1 | 1126 | * Called when a complete comment has been scanned. pos and endPos |
duke@1 | 1127 | * will mark the comment boundary. |
duke@1 | 1128 | */ |
duke@1 | 1129 | protected void processComment(CommentStyle style) { |
duke@1 | 1130 | if (scannerDebug) |
duke@1 | 1131 | System.out.println("processComment(" + pos |
duke@1 | 1132 | + "," + endPos + "," + style + ")=|" |
duke@1 | 1133 | + new String(getRawCharacters(pos, endPos)) |
duke@1 | 1134 | + "|"); |
duke@1 | 1135 | } |
duke@1 | 1136 | |
duke@1 | 1137 | /** |
duke@1 | 1138 | * Called when a complete whitespace run has been scanned. pos and endPos |
duke@1 | 1139 | * will mark the whitespace boundary. |
duke@1 | 1140 | */ |
duke@1 | 1141 | protected void processWhiteSpace() { |
duke@1 | 1142 | if (scannerDebug) |
duke@1 | 1143 | System.out.println("processWhitespace(" + pos |
duke@1 | 1144 | + "," + endPos + ")=|" + |
duke@1 | 1145 | new String(getRawCharacters(pos, endPos)) |
duke@1 | 1146 | + "|"); |
duke@1 | 1147 | } |
duke@1 | 1148 | |
duke@1 | 1149 | /** |
duke@1 | 1150 | * Called when a line terminator has been processed. |
duke@1 | 1151 | */ |
duke@1 | 1152 | protected void processLineTerminator() { |
duke@1 | 1153 | if (scannerDebug) |
duke@1 | 1154 | System.out.println("processTerminator(" + pos |
duke@1 | 1155 | + "," + endPos + ")=|" + |
duke@1 | 1156 | new String(getRawCharacters(pos, endPos)) |
duke@1 | 1157 | + "|"); |
duke@1 | 1158 | } |
duke@1 | 1159 | |
duke@1 | 1160 | /** Build a map for translating between line numbers and |
duke@1 | 1161 | * positions in the input. |
duke@1 | 1162 | * |
duke@1 | 1163 | * @return a LineMap */ |
duke@1 | 1164 | public Position.LineMap getLineMap() { |
duke@1 | 1165 | return Position.makeLineMap(buf, buflen, false); |
duke@1 | 1166 | } |
duke@1 | 1167 | |
duke@1 | 1168 | } |