The data contained in this repository can be downloaded to your computer using one of several clients.
Please see the documentation of your version control software client for more information.
Please select the desired protocol below to get the URL.
This URL has Read-Only access.
main_repo / deps / v8 / src / interpreter-irregexp.cc @ 40c0f755
History | View | Annotate | Download (19.9 KB)
| 1 |
// Copyright 2008 the V8 project authors. All rights reserved.
|
|---|---|
| 2 |
// Redistribution and use in source and binary forms, with or without
|
| 3 |
// modification, are permitted provided that the following conditions are
|
| 4 |
// met:
|
| 5 |
//
|
| 6 |
// * Redistributions of source code must retain the above copyright
|
| 7 |
// notice, this list of conditions and the following disclaimer.
|
| 8 |
// * Redistributions in binary form must reproduce the above
|
| 9 |
// copyright notice, this list of conditions and the following
|
| 10 |
// disclaimer in the documentation and/or other materials provided
|
| 11 |
// with the distribution.
|
| 12 |
// * Neither the name of Google Inc. nor the names of its
|
| 13 |
// contributors may be used to endorse or promote products derived
|
| 14 |
// from this software without specific prior written permission.
|
| 15 |
//
|
| 16 |
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
| 17 |
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
| 18 |
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
| 19 |
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
| 20 |
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
| 21 |
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
| 22 |
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
| 23 |
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
| 24 |
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
| 25 |
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
| 26 |
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
| 27 |
|
| 28 |
// A simple interpreter for the Irregexp byte code.
|
| 29 |
|
| 30 |
|
| 31 |
#include "v8.h" |
| 32 |
#include "unicode.h" |
| 33 |
#include "utils.h" |
| 34 |
#include "ast.h" |
| 35 |
#include "bytecodes-irregexp.h" |
| 36 |
#include "interpreter-irregexp.h" |
| 37 |
|
| 38 |
|
| 39 |
namespace v8 { namespace internal { |
| 40 |
|
| 41 |
|
| 42 |
static unibrow::Mapping<unibrow::Ecma262Canonicalize> interp_canonicalize;
|
| 43 |
|
| 44 |
|
| 45 |
static bool BackRefMatchesNoCase(int from, |
| 46 |
int current,
|
| 47 |
int len,
|
| 48 |
Vector<const uc16> subject) {
|
| 49 |
for (int i = 0; i < len; i++) { |
| 50 |
unibrow::uchar old_char = subject[from++]; |
| 51 |
unibrow::uchar new_char = subject[current++]; |
| 52 |
if (old_char == new_char) continue; |
| 53 |
interp_canonicalize.get(old_char, '\0', &old_char);
|
| 54 |
interp_canonicalize.get(new_char, '\0', &new_char);
|
| 55 |
if (old_char != new_char) {
|
| 56 |
return false; |
| 57 |
} |
| 58 |
} |
| 59 |
return true; |
| 60 |
} |
| 61 |
|
| 62 |
|
| 63 |
static bool BackRefMatchesNoCase(int from, |
| 64 |
int current,
|
| 65 |
int len,
|
| 66 |
Vector<const char> subject) { |
| 67 |
for (int i = 0; i < len; i++) { |
| 68 |
unsigned int old_char = subject[from++]; |
| 69 |
unsigned int new_char = subject[current++]; |
| 70 |
if (old_char == new_char) continue; |
| 71 |
if (old_char - 'A' <= 'Z' - 'A') old_char |= 0x20; |
| 72 |
if (new_char - 'A' <= 'Z' - 'A') new_char |= 0x20; |
| 73 |
if (old_char != new_char) return false; |
| 74 |
} |
| 75 |
return true; |
| 76 |
} |
| 77 |
|
| 78 |
|
| 79 |
#ifdef DEBUG
|
| 80 |
static void TraceInterpreter(const byte* code_base, |
| 81 |
const byte* pc,
|
| 82 |
int stack_depth,
|
| 83 |
int current_position,
|
| 84 |
uint32_t current_char, |
| 85 |
int bytecode_length,
|
| 86 |
const char* bytecode_name) { |
| 87 |
if (FLAG_trace_regexp_bytecodes) {
|
| 88 |
bool printable = (current_char < 127 && current_char >= 32); |
| 89 |
const char* format = |
| 90 |
printable ? |
| 91 |
"pc = %02x, sp = %d, curpos = %d, curchar = %08x (%c), bc = %s" :
|
| 92 |
"pc = %02x, sp = %d, curpos = %d, curchar = %08x .%c., bc = %s";
|
| 93 |
PrintF(format, |
| 94 |
pc - code_base, |
| 95 |
stack_depth, |
| 96 |
current_position, |
| 97 |
current_char, |
| 98 |
printable ? current_char : '.',
|
| 99 |
bytecode_name); |
| 100 |
for (int i = 0; i < bytecode_length; i++) { |
| 101 |
printf(", %02x", pc[i]);
|
| 102 |
} |
| 103 |
printf(" ");
|
| 104 |
for (int i = 1; i < bytecode_length; i++) { |
| 105 |
unsigned char b = pc[i]; |
| 106 |
if (b < 127 && b >= 32) { |
| 107 |
printf("%c", b);
|
| 108 |
} else {
|
| 109 |
printf(".");
|
| 110 |
} |
| 111 |
} |
| 112 |
printf("\n");
|
| 113 |
} |
| 114 |
} |
| 115 |
|
| 116 |
|
| 117 |
#define BYTECODE(name) \
|
| 118 |
case BC_##name: \ |
| 119 |
TraceInterpreter(code_base, \ |
| 120 |
pc, \ |
| 121 |
backtrack_sp - backtrack_stack, \ |
| 122 |
current, \ |
| 123 |
current_char, \ |
| 124 |
BC_##name##_LENGTH, \ |
| 125 |
#name);
|
| 126 |
#else
|
| 127 |
#define BYTECODE(name) \
|
| 128 |
case BC_##name: |
| 129 |
#endif
|
| 130 |
|
| 131 |
|
| 132 |
static int32_t Load32Aligned(const byte* pc) { |
| 133 |
ASSERT((reinterpret_cast<int>(pc) & 3) == 0); |
| 134 |
return *reinterpret_cast<const int32_t *>(pc); |
| 135 |
} |
| 136 |
|
| 137 |
|
| 138 |
static int32_t Load16Aligned(const byte* pc) { |
| 139 |
ASSERT((reinterpret_cast<int>(pc) & 1) == 0); |
| 140 |
return *reinterpret_cast<const uint16_t *>(pc); |
| 141 |
} |
| 142 |
|
| 143 |
|
| 144 |
template <typename Char> |
| 145 |
static bool RawMatch(const byte* code_base, |
| 146 |
Vector<const Char> subject,
|
| 147 |
int* registers,
|
| 148 |
int current,
|
| 149 |
uint32_t current_char) {
|
| 150 |
const byte* pc = code_base;
|
| 151 |
static const int kBacktrackStackSize = 10000; |
| 152 |
int backtrack_stack[kBacktrackStackSize];
|
| 153 |
int backtrack_stack_space = kBacktrackStackSize;
|
| 154 |
int* backtrack_sp = backtrack_stack;
|
| 155 |
#ifdef DEBUG
|
| 156 |
if (FLAG_trace_regexp_bytecodes) {
|
| 157 |
PrintF("\n\nStart bytecode interpreter\n\n");
|
| 158 |
} |
| 159 |
#endif
|
| 160 |
while (true) { |
| 161 |
int32_t insn = Load32Aligned(pc); |
| 162 |
switch (insn & BYTECODE_MASK) {
|
| 163 |
BYTECODE(BREAK) |
| 164 |
UNREACHABLE(); |
| 165 |
return false; |
| 166 |
BYTECODE(PUSH_CP) |
| 167 |
if (--backtrack_stack_space < 0) { |
| 168 |
return false; // No match on backtrack stack overflow. |
| 169 |
} |
| 170 |
*backtrack_sp++ = current; |
| 171 |
pc += BC_PUSH_CP_LENGTH; |
| 172 |
break;
|
| 173 |
BYTECODE(PUSH_BT) |
| 174 |
if (--backtrack_stack_space < 0) { |
| 175 |
return false; // No match on backtrack stack overflow. |
| 176 |
} |
| 177 |
*backtrack_sp++ = Load32Aligned(pc + 4);
|
| 178 |
pc += BC_PUSH_BT_LENGTH; |
| 179 |
break;
|
| 180 |
BYTECODE(PUSH_REGISTER) |
| 181 |
if (--backtrack_stack_space < 0) { |
| 182 |
return false; // No match on backtrack stack overflow. |
| 183 |
} |
| 184 |
*backtrack_sp++ = registers[insn >> BYTECODE_SHIFT]; |
| 185 |
pc += BC_PUSH_REGISTER_LENGTH; |
| 186 |
break;
|
| 187 |
BYTECODE(SET_REGISTER) |
| 188 |
registers[insn >> BYTECODE_SHIFT] = Load32Aligned(pc + 4);
|
| 189 |
pc += BC_SET_REGISTER_LENGTH; |
| 190 |
break;
|
| 191 |
BYTECODE(ADVANCE_REGISTER) |
| 192 |
registers[insn >> BYTECODE_SHIFT] += Load32Aligned(pc + 4);
|
| 193 |
pc += BC_ADVANCE_REGISTER_LENGTH; |
| 194 |
break;
|
| 195 |
BYTECODE(SET_REGISTER_TO_CP) |
| 196 |
registers[insn >> BYTECODE_SHIFT] = current + Load32Aligned(pc + 4);
|
| 197 |
pc += BC_SET_REGISTER_TO_CP_LENGTH; |
| 198 |
break;
|
| 199 |
BYTECODE(SET_CP_TO_REGISTER) |
| 200 |
current = registers[insn >> BYTECODE_SHIFT]; |
| 201 |
pc += BC_SET_CP_TO_REGISTER_LENGTH; |
| 202 |
break;
|
| 203 |
BYTECODE(SET_REGISTER_TO_SP) |
| 204 |
registers[insn >> BYTECODE_SHIFT] = backtrack_sp - backtrack_stack; |
| 205 |
pc += BC_SET_REGISTER_TO_SP_LENGTH; |
| 206 |
break;
|
| 207 |
BYTECODE(SET_SP_TO_REGISTER) |
| 208 |
backtrack_sp = backtrack_stack + registers[insn >> BYTECODE_SHIFT]; |
| 209 |
backtrack_stack_space = kBacktrackStackSize - |
| 210 |
(backtrack_sp - backtrack_stack); |
| 211 |
pc += BC_SET_SP_TO_REGISTER_LENGTH; |
| 212 |
break;
|
| 213 |
BYTECODE(POP_CP) |
| 214 |
backtrack_stack_space++; |
| 215 |
--backtrack_sp; |
| 216 |
current = *backtrack_sp; |
| 217 |
pc += BC_POP_CP_LENGTH; |
| 218 |
break;
|
| 219 |
BYTECODE(POP_BT) |
| 220 |
backtrack_stack_space++; |
| 221 |
--backtrack_sp; |
| 222 |
pc = code_base + *backtrack_sp; |
| 223 |
break;
|
| 224 |
BYTECODE(POP_REGISTER) |
| 225 |
backtrack_stack_space++; |
| 226 |
--backtrack_sp; |
| 227 |
registers[insn >> BYTECODE_SHIFT] = *backtrack_sp; |
| 228 |
pc += BC_POP_REGISTER_LENGTH; |
| 229 |
break;
|
| 230 |
BYTECODE(FAIL) |
| 231 |
return false; |
| 232 |
BYTECODE(SUCCEED) |
| 233 |
return true; |
| 234 |
BYTECODE(ADVANCE_CP) |
| 235 |
current += insn >> BYTECODE_SHIFT; |
| 236 |
pc += BC_ADVANCE_CP_LENGTH; |
| 237 |
break;
|
| 238 |
BYTECODE(GOTO) |
| 239 |
pc = code_base + Load32Aligned(pc + 4);
|
| 240 |
break;
|
| 241 |
BYTECODE(ADVANCE_CP_AND_GOTO) |
| 242 |
current += insn >> BYTECODE_SHIFT; |
| 243 |
pc = code_base + Load32Aligned(pc + 4);
|
| 244 |
break;
|
| 245 |
BYTECODE(CHECK_GREEDY) |
| 246 |
if (current == backtrack_sp[-1]) { |
| 247 |
backtrack_sp--; |
| 248 |
backtrack_stack_space++; |
| 249 |
pc = code_base + Load32Aligned(pc + 4);
|
| 250 |
} else {
|
| 251 |
pc += BC_CHECK_GREEDY_LENGTH; |
| 252 |
} |
| 253 |
break;
|
| 254 |
BYTECODE(LOAD_CURRENT_CHAR) {
|
| 255 |
int pos = current + (insn >> BYTECODE_SHIFT);
|
| 256 |
if (pos >= subject.length()) {
|
| 257 |
pc = code_base + Load32Aligned(pc + 4);
|
| 258 |
} else {
|
| 259 |
current_char = subject[pos]; |
| 260 |
pc += BC_LOAD_CURRENT_CHAR_LENGTH; |
| 261 |
} |
| 262 |
break;
|
| 263 |
} |
| 264 |
BYTECODE(LOAD_CURRENT_CHAR_UNCHECKED) {
|
| 265 |
int pos = current + (insn >> BYTECODE_SHIFT);
|
| 266 |
current_char = subject[pos]; |
| 267 |
pc += BC_LOAD_CURRENT_CHAR_UNCHECKED_LENGTH; |
| 268 |
break;
|
| 269 |
} |
| 270 |
BYTECODE(LOAD_2_CURRENT_CHARS) {
|
| 271 |
int pos = current + (insn >> BYTECODE_SHIFT);
|
| 272 |
if (pos + 2 > subject.length()) { |
| 273 |
pc = code_base + Load32Aligned(pc + 4);
|
| 274 |
} else {
|
| 275 |
Char next = subject[pos + 1];
|
| 276 |
current_char = |
| 277 |
(subject[pos] | (next << (kBitsPerByte * sizeof(Char))));
|
| 278 |
pc += BC_LOAD_2_CURRENT_CHARS_LENGTH; |
| 279 |
} |
| 280 |
break;
|
| 281 |
} |
| 282 |
BYTECODE(LOAD_2_CURRENT_CHARS_UNCHECKED) {
|
| 283 |
int pos = current + (insn >> BYTECODE_SHIFT);
|
| 284 |
Char next = subject[pos + 1];
|
| 285 |
current_char = (subject[pos] | (next << (kBitsPerByte * sizeof(Char))));
|
| 286 |
pc += BC_LOAD_2_CURRENT_CHARS_UNCHECKED_LENGTH; |
| 287 |
break;
|
| 288 |
} |
| 289 |
BYTECODE(LOAD_4_CURRENT_CHARS) {
|
| 290 |
ASSERT(sizeof(Char) == 1); |
| 291 |
int pos = current + (insn >> BYTECODE_SHIFT);
|
| 292 |
if (pos + 4 > subject.length()) { |
| 293 |
pc = code_base + Load32Aligned(pc + 4);
|
| 294 |
} else {
|
| 295 |
Char next1 = subject[pos + 1];
|
| 296 |
Char next2 = subject[pos + 2];
|
| 297 |
Char next3 = subject[pos + 3];
|
| 298 |
current_char = (subject[pos] | |
| 299 |
(next1 << 8) |
|
| 300 |
(next2 << 16) |
|
| 301 |
(next3 << 24));
|
| 302 |
pc += BC_LOAD_4_CURRENT_CHARS_LENGTH; |
| 303 |
} |
| 304 |
break;
|
| 305 |
} |
| 306 |
BYTECODE(LOAD_4_CURRENT_CHARS_UNCHECKED) {
|
| 307 |
ASSERT(sizeof(Char) == 1); |
| 308 |
int pos = current + (insn >> BYTECODE_SHIFT);
|
| 309 |
Char next1 = subject[pos + 1];
|
| 310 |
Char next2 = subject[pos + 2];
|
| 311 |
Char next3 = subject[pos + 3];
|
| 312 |
current_char = (subject[pos] | |
| 313 |
(next1 << 8) |
|
| 314 |
(next2 << 16) |
|
| 315 |
(next3 << 24));
|
| 316 |
pc += BC_LOAD_4_CURRENT_CHARS_UNCHECKED_LENGTH; |
| 317 |
break;
|
| 318 |
} |
| 319 |
BYTECODE(CHECK_4_CHARS) {
|
| 320 |
uint32_t c = Load32Aligned(pc + 4);
|
| 321 |
if (c == current_char) {
|
| 322 |
pc = code_base + Load32Aligned(pc + 8);
|
| 323 |
} else {
|
| 324 |
pc += BC_CHECK_4_CHARS_LENGTH; |
| 325 |
} |
| 326 |
break;
|
| 327 |
} |
| 328 |
BYTECODE(CHECK_CHAR) {
|
| 329 |
uint32_t c = (insn >> BYTECODE_SHIFT); |
| 330 |
if (c == current_char) {
|
| 331 |
pc = code_base + Load32Aligned(pc + 4);
|
| 332 |
} else {
|
| 333 |
pc += BC_CHECK_CHAR_LENGTH; |
| 334 |
} |
| 335 |
break;
|
| 336 |
} |
| 337 |
BYTECODE(CHECK_NOT_4_CHARS) {
|
| 338 |
uint32_t c = Load32Aligned(pc + 4);
|
| 339 |
if (c != current_char) {
|
| 340 |
pc = code_base + Load32Aligned(pc + 8);
|
| 341 |
} else {
|
| 342 |
pc += BC_CHECK_NOT_4_CHARS_LENGTH; |
| 343 |
} |
| 344 |
break;
|
| 345 |
} |
| 346 |
BYTECODE(CHECK_NOT_CHAR) {
|
| 347 |
uint32_t c = (insn >> BYTECODE_SHIFT); |
| 348 |
if (c != current_char) {
|
| 349 |
pc = code_base + Load32Aligned(pc + 4);
|
| 350 |
} else {
|
| 351 |
pc += BC_CHECK_NOT_CHAR_LENGTH; |
| 352 |
} |
| 353 |
break;
|
| 354 |
} |
| 355 |
BYTECODE(AND_CHECK_4_CHARS) {
|
| 356 |
uint32_t c = Load32Aligned(pc + 4);
|
| 357 |
if (c == (current_char & Load32Aligned(pc + 8))) { |
| 358 |
pc = code_base + Load32Aligned(pc + 12);
|
| 359 |
} else {
|
| 360 |
pc += BC_AND_CHECK_4_CHARS_LENGTH; |
| 361 |
} |
| 362 |
break;
|
| 363 |
} |
| 364 |
BYTECODE(AND_CHECK_CHAR) {
|
| 365 |
uint32_t c = (insn >> BYTECODE_SHIFT); |
| 366 |
if (c == (current_char & Load32Aligned(pc + 4))) { |
| 367 |
pc = code_base + Load32Aligned(pc + 8);
|
| 368 |
} else {
|
| 369 |
pc += BC_AND_CHECK_CHAR_LENGTH; |
| 370 |
} |
| 371 |
break;
|
| 372 |
} |
| 373 |
BYTECODE(AND_CHECK_NOT_4_CHARS) {
|
| 374 |
uint32_t c = Load32Aligned(pc + 4);
|
| 375 |
if (c != (current_char & Load32Aligned(pc + 8))) { |
| 376 |
pc = code_base + Load32Aligned(pc + 12);
|
| 377 |
} else {
|
| 378 |
pc += BC_AND_CHECK_NOT_4_CHARS_LENGTH; |
| 379 |
} |
| 380 |
break;
|
| 381 |
} |
| 382 |
BYTECODE(AND_CHECK_NOT_CHAR) {
|
| 383 |
uint32_t c = (insn >> BYTECODE_SHIFT); |
| 384 |
if (c != (current_char & Load32Aligned(pc + 4))) { |
| 385 |
pc = code_base + Load32Aligned(pc + 8);
|
| 386 |
} else {
|
| 387 |
pc += BC_AND_CHECK_NOT_CHAR_LENGTH; |
| 388 |
} |
| 389 |
break;
|
| 390 |
} |
| 391 |
BYTECODE(MINUS_AND_CHECK_NOT_CHAR) {
|
| 392 |
uint32_t c = (insn >> BYTECODE_SHIFT); |
| 393 |
uint32_t minus = Load16Aligned(pc + 4);
|
| 394 |
uint32_t mask = Load16Aligned(pc + 6);
|
| 395 |
if (c != ((current_char - minus) & mask)) {
|
| 396 |
pc = code_base + Load32Aligned(pc + 8);
|
| 397 |
} else {
|
| 398 |
pc += BC_MINUS_AND_CHECK_NOT_CHAR_LENGTH; |
| 399 |
} |
| 400 |
break;
|
| 401 |
} |
| 402 |
BYTECODE(CHECK_LT) {
|
| 403 |
uint32_t limit = (insn >> BYTECODE_SHIFT); |
| 404 |
if (current_char < limit) {
|
| 405 |
pc = code_base + Load32Aligned(pc + 4);
|
| 406 |
} else {
|
| 407 |
pc += BC_CHECK_LT_LENGTH; |
| 408 |
} |
| 409 |
break;
|
| 410 |
} |
| 411 |
BYTECODE(CHECK_GT) {
|
| 412 |
uint32_t limit = (insn >> BYTECODE_SHIFT); |
| 413 |
if (current_char > limit) {
|
| 414 |
pc = code_base + Load32Aligned(pc + 4);
|
| 415 |
} else {
|
| 416 |
pc += BC_CHECK_GT_LENGTH; |
| 417 |
} |
| 418 |
break;
|
| 419 |
} |
| 420 |
BYTECODE(CHECK_REGISTER_LT) |
| 421 |
if (registers[insn >> BYTECODE_SHIFT] < Load32Aligned(pc + 4)) { |
| 422 |
pc = code_base + Load32Aligned(pc + 8);
|
| 423 |
} else {
|
| 424 |
pc += BC_CHECK_REGISTER_LT_LENGTH; |
| 425 |
} |
| 426 |
break;
|
| 427 |
BYTECODE(CHECK_REGISTER_GE) |
| 428 |
if (registers[insn >> BYTECODE_SHIFT] >= Load32Aligned(pc + 4)) { |
| 429 |
pc = code_base + Load32Aligned(pc + 8);
|
| 430 |
} else {
|
| 431 |
pc += BC_CHECK_REGISTER_GE_LENGTH; |
| 432 |
} |
| 433 |
break;
|
| 434 |
BYTECODE(CHECK_REGISTER_EQ_POS) |
| 435 |
if (registers[insn >> BYTECODE_SHIFT] == current) {
|
| 436 |
pc = code_base + Load32Aligned(pc + 4);
|
| 437 |
} else {
|
| 438 |
pc += BC_CHECK_REGISTER_EQ_POS_LENGTH; |
| 439 |
} |
| 440 |
break;
|
| 441 |
BYTECODE(LOOKUP_MAP1) {
|
| 442 |
// Look up character in a bitmap. If we find a 0, then jump to the
|
| 443 |
// location at pc + 8. Otherwise fall through!
|
| 444 |
int index = current_char - (insn >> BYTECODE_SHIFT);
|
| 445 |
byte map = code_base[Load32Aligned(pc + 4) + (index >> 3)]; |
| 446 |
map = ((map >> (index & 7)) & 1); |
| 447 |
if (map == 0) { |
| 448 |
pc = code_base + Load32Aligned(pc + 8);
|
| 449 |
} else {
|
| 450 |
pc += BC_LOOKUP_MAP1_LENGTH; |
| 451 |
} |
| 452 |
break;
|
| 453 |
} |
| 454 |
BYTECODE(LOOKUP_MAP2) {
|
| 455 |
// Look up character in a half-nibble map. If we find 00, then jump to
|
| 456 |
// the location at pc + 8. If we find 01 then jump to location at
|
| 457 |
// pc + 11, etc.
|
| 458 |
int index = (current_char - (insn >> BYTECODE_SHIFT)) << 1; |
| 459 |
byte map = code_base[Load32Aligned(pc + 3) + (index >> 3)]; |
| 460 |
map = ((map >> (index & 7)) & 3); |
| 461 |
if (map < 2) { |
| 462 |
if (map == 0) { |
| 463 |
pc = code_base + Load32Aligned(pc + 8);
|
| 464 |
} else {
|
| 465 |
pc = code_base + Load32Aligned(pc + 12);
|
| 466 |
} |
| 467 |
} else {
|
| 468 |
if (map == 2) { |
| 469 |
pc = code_base + Load32Aligned(pc + 16);
|
| 470 |
} else {
|
| 471 |
pc = code_base + Load32Aligned(pc + 20);
|
| 472 |
} |
| 473 |
} |
| 474 |
break;
|
| 475 |
} |
| 476 |
BYTECODE(LOOKUP_MAP8) {
|
| 477 |
// Look up character in a byte map. Use the byte as an index into a
|
| 478 |
// table that follows this instruction immediately.
|
| 479 |
int index = current_char - (insn >> BYTECODE_SHIFT);
|
| 480 |
byte map = code_base[Load32Aligned(pc + 4) + index];
|
| 481 |
const byte* new_pc = code_base + Load32Aligned(pc + 8) + (map << 2); |
| 482 |
pc = code_base + Load32Aligned(new_pc); |
| 483 |
break;
|
| 484 |
} |
| 485 |
BYTECODE(LOOKUP_HI_MAP8) {
|
| 486 |
// Look up high byte of this character in a byte map. Use the byte as
|
| 487 |
// an index into a table that follows this instruction immediately.
|
| 488 |
int index = (current_char >> 8) - (insn >> BYTECODE_SHIFT); |
| 489 |
byte map = code_base[Load32Aligned(pc + 4) + index];
|
| 490 |
const byte* new_pc = code_base + Load32Aligned(pc + 8) + (map << 2); |
| 491 |
pc = code_base + Load32Aligned(new_pc); |
| 492 |
break;
|
| 493 |
} |
| 494 |
BYTECODE(CHECK_NOT_REGS_EQUAL) |
| 495 |
if (registers[insn >> BYTECODE_SHIFT] ==
|
| 496 |
registers[Load32Aligned(pc + 4)]) {
|
| 497 |
pc += BC_CHECK_NOT_REGS_EQUAL_LENGTH; |
| 498 |
} else {
|
| 499 |
pc = code_base + Load32Aligned(pc + 8);
|
| 500 |
} |
| 501 |
break;
|
| 502 |
BYTECODE(CHECK_NOT_BACK_REF) {
|
| 503 |
int from = registers[insn >> BYTECODE_SHIFT];
|
| 504 |
int len = registers[(insn >> BYTECODE_SHIFT) + 1] - from; |
| 505 |
if (from < 0 || len <= 0) { |
| 506 |
pc += BC_CHECK_NOT_BACK_REF_LENGTH; |
| 507 |
break;
|
| 508 |
} |
| 509 |
if (current + len > subject.length()) {
|
| 510 |
pc = code_base + Load32Aligned(pc + 4);
|
| 511 |
break;
|
| 512 |
} else {
|
| 513 |
int i;
|
| 514 |
for (i = 0; i < len; i++) { |
| 515 |
if (subject[from + i] != subject[current + i]) {
|
| 516 |
pc = code_base + Load32Aligned(pc + 4);
|
| 517 |
break;
|
| 518 |
} |
| 519 |
} |
| 520 |
if (i < len) break; |
| 521 |
current += len; |
| 522 |
} |
| 523 |
pc += BC_CHECK_NOT_BACK_REF_LENGTH; |
| 524 |
break;
|
| 525 |
} |
| 526 |
BYTECODE(CHECK_NOT_BACK_REF_NO_CASE) {
|
| 527 |
int from = registers[insn >> BYTECODE_SHIFT];
|
| 528 |
int len = registers[(insn >> BYTECODE_SHIFT) + 1] - from; |
| 529 |
if (from < 0 || len <= 0) { |
| 530 |
pc += BC_CHECK_NOT_BACK_REF_NO_CASE_LENGTH; |
| 531 |
break;
|
| 532 |
} |
| 533 |
if (current + len > subject.length()) {
|
| 534 |
pc = code_base + Load32Aligned(pc + 4);
|
| 535 |
break;
|
| 536 |
} else {
|
| 537 |
if (BackRefMatchesNoCase(from, current, len, subject)) {
|
| 538 |
current += len; |
| 539 |
pc += BC_CHECK_NOT_BACK_REF_NO_CASE_LENGTH; |
| 540 |
} else {
|
| 541 |
pc = code_base + Load32Aligned(pc + 4);
|
| 542 |
} |
| 543 |
} |
| 544 |
break;
|
| 545 |
} |
| 546 |
BYTECODE(CHECK_AT_START) |
| 547 |
if (current == 0) { |
| 548 |
pc = code_base + Load32Aligned(pc + 4);
|
| 549 |
} else {
|
| 550 |
pc += BC_CHECK_AT_START_LENGTH; |
| 551 |
} |
| 552 |
break;
|
| 553 |
BYTECODE(CHECK_NOT_AT_START) |
| 554 |
if (current == 0) { |
| 555 |
pc += BC_CHECK_NOT_AT_START_LENGTH; |
| 556 |
} else {
|
| 557 |
pc = code_base + Load32Aligned(pc + 4);
|
| 558 |
} |
| 559 |
break;
|
| 560 |
default:
|
| 561 |
UNREACHABLE(); |
| 562 |
break;
|
| 563 |
} |
| 564 |
} |
| 565 |
} |
| 566 |
|
| 567 |
|
| 568 |
bool IrregexpInterpreter::Match(Handle<ByteArray> code_array,
|
| 569 |
Handle<String> subject, |
| 570 |
int* registers,
|
| 571 |
int start_position) {
|
| 572 |
ASSERT(subject->IsFlat()); |
| 573 |
|
| 574 |
AssertNoAllocation a; |
| 575 |
const byte* code_base = code_array->GetDataStartAddress();
|
| 576 |
uc16 previous_char = '\n';
|
| 577 |
if (StringShape(*subject).IsAsciiRepresentation()) {
|
| 578 |
Vector<const char> subject_vector = subject->ToAsciiVector(); |
| 579 |
if (start_position != 0) previous_char = subject_vector[start_position - 1]; |
| 580 |
return RawMatch(code_base,
|
| 581 |
subject_vector, |
| 582 |
registers, |
| 583 |
start_position, |
| 584 |
previous_char); |
| 585 |
} else {
|
| 586 |
Vector<const uc16> subject_vector = subject->ToUC16Vector();
|
| 587 |
if (start_position != 0) previous_char = subject_vector[start_position - 1]; |
| 588 |
return RawMatch(code_base,
|
| 589 |
subject_vector, |
| 590 |
registers, |
| 591 |
start_position, |
| 592 |
previous_char); |
| 593 |
} |
| 594 |
} |
| 595 |
|
| 596 |
} } // namespace v8::internal
|