Ben Murdoch | 4a90d5f | 2016-03-22 12:00:34 +0000 | [diff] [blame] | 1 | // Copyright 2014 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 | #include "src/ast/ast-value-factory.h" |
| 29 | |
| 30 | #include "src/api.h" |
| 31 | #include "src/objects.h" |
| 32 | #include "src/utils.h" |
| 33 | |
| 34 | namespace v8 { |
| 35 | namespace internal { |
| 36 | |
| 37 | namespace { |
| 38 | |
| 39 | // For using StringToArrayIndex. |
| 40 | class OneByteStringStream { |
| 41 | public: |
| 42 | explicit OneByteStringStream(Vector<const byte> lb) : |
| 43 | literal_bytes_(lb), pos_(0) {} |
| 44 | |
| 45 | bool HasMore() { return pos_ < literal_bytes_.length(); } |
| 46 | uint16_t GetNext() { return literal_bytes_[pos_++]; } |
| 47 | |
| 48 | private: |
| 49 | Vector<const byte> literal_bytes_; |
| 50 | int pos_; |
| 51 | }; |
| 52 | |
| 53 | } // namespace |
| 54 | |
| 55 | class AstRawStringInternalizationKey : public HashTableKey { |
| 56 | public: |
| 57 | explicit AstRawStringInternalizationKey(const AstRawString* string) |
| 58 | : string_(string) {} |
| 59 | |
| 60 | bool IsMatch(Object* other) override { |
| 61 | if (string_->is_one_byte_) |
| 62 | return String::cast(other)->IsOneByteEqualTo(string_->literal_bytes_); |
| 63 | return String::cast(other)->IsTwoByteEqualTo( |
| 64 | Vector<const uint16_t>::cast(string_->literal_bytes_)); |
| 65 | } |
| 66 | |
| 67 | uint32_t Hash() override { return string_->hash() >> Name::kHashShift; } |
| 68 | |
| 69 | uint32_t HashForObject(Object* key) override { |
| 70 | return String::cast(key)->Hash(); |
| 71 | } |
| 72 | |
| 73 | Handle<Object> AsHandle(Isolate* isolate) override { |
| 74 | if (string_->is_one_byte_) |
| 75 | return isolate->factory()->NewOneByteInternalizedString( |
| 76 | string_->literal_bytes_, string_->hash()); |
| 77 | return isolate->factory()->NewTwoByteInternalizedString( |
| 78 | Vector<const uint16_t>::cast(string_->literal_bytes_), string_->hash()); |
| 79 | } |
| 80 | |
| 81 | private: |
| 82 | const AstRawString* string_; |
| 83 | }; |
| 84 | |
| 85 | |
| 86 | void AstRawString::Internalize(Isolate* isolate) { |
| 87 | if (!string_.is_null()) return; |
| 88 | if (literal_bytes_.length() == 0) { |
| 89 | string_ = isolate->factory()->empty_string(); |
| 90 | } else { |
| 91 | AstRawStringInternalizationKey key(this); |
| 92 | string_ = StringTable::LookupKey(isolate, &key); |
| 93 | } |
| 94 | } |
| 95 | |
| 96 | |
| 97 | bool AstRawString::AsArrayIndex(uint32_t* index) const { |
| 98 | if (!string_.is_null()) |
| 99 | return string_->AsArrayIndex(index); |
| 100 | if (!is_one_byte_ || literal_bytes_.length() == 0 || |
| 101 | literal_bytes_.length() > String::kMaxArrayIndexSize) |
| 102 | return false; |
| 103 | OneByteStringStream stream(literal_bytes_); |
| 104 | return StringToArrayIndex(&stream, index); |
| 105 | } |
| 106 | |
| 107 | |
| 108 | bool AstRawString::IsOneByteEqualTo(const char* data) const { |
| 109 | int length = static_cast<int>(strlen(data)); |
| 110 | if (is_one_byte_ && literal_bytes_.length() == length) { |
| 111 | const char* token = reinterpret_cast<const char*>(literal_bytes_.start()); |
| 112 | return !strncmp(token, data, length); |
| 113 | } |
| 114 | return false; |
| 115 | } |
| 116 | |
| 117 | |
| 118 | void AstConsString::Internalize(Isolate* isolate) { |
| 119 | // AstRawStrings are internalized before AstConsStrings so left and right are |
| 120 | // already internalized. |
| 121 | string_ = isolate->factory() |
| 122 | ->NewConsString(left_->string(), right_->string()) |
| 123 | .ToHandleChecked(); |
| 124 | } |
| 125 | |
| 126 | |
| 127 | bool AstValue::IsPropertyName() const { |
| 128 | if (type_ == STRING) { |
| 129 | uint32_t index; |
| 130 | return !string_->AsArrayIndex(&index); |
| 131 | } |
| 132 | return false; |
| 133 | } |
| 134 | |
| 135 | |
| 136 | bool AstValue::BooleanValue() const { |
| 137 | switch (type_) { |
| 138 | case STRING: |
| 139 | DCHECK(string_ != NULL); |
| 140 | return !string_->IsEmpty(); |
| 141 | case SYMBOL: |
| 142 | UNREACHABLE(); |
| 143 | break; |
| 144 | case NUMBER_WITH_DOT: |
| 145 | case NUMBER: |
| 146 | return DoubleToBoolean(number_); |
| 147 | case SMI: |
| 148 | return smi_ != 0; |
| 149 | case BOOLEAN: |
| 150 | return bool_; |
| 151 | case NULL_TYPE: |
| 152 | return false; |
| 153 | case THE_HOLE: |
| 154 | UNREACHABLE(); |
| 155 | break; |
| 156 | case UNDEFINED: |
| 157 | return false; |
| 158 | } |
| 159 | UNREACHABLE(); |
| 160 | return false; |
| 161 | } |
| 162 | |
| 163 | |
| 164 | void AstValue::Internalize(Isolate* isolate) { |
| 165 | switch (type_) { |
| 166 | case STRING: |
| 167 | DCHECK(string_ != NULL); |
| 168 | // Strings are already internalized. |
| 169 | DCHECK(!string_->string().is_null()); |
| 170 | break; |
| 171 | case SYMBOL: |
| 172 | if (symbol_name_[0] == 'i') { |
| 173 | DCHECK_EQ(0, strcmp(symbol_name_, "iterator_symbol")); |
| 174 | value_ = isolate->factory()->iterator_symbol(); |
Ben Murdoch | 097c5b2 | 2016-05-18 11:27:45 +0100 | [diff] [blame] | 175 | } else if (strcmp(symbol_name_, "hasInstance_symbol") == 0) { |
| 176 | value_ = isolate->factory()->has_instance_symbol(); |
Ben Murdoch | 4a90d5f | 2016-03-22 12:00:34 +0000 | [diff] [blame] | 177 | } else { |
| 178 | DCHECK_EQ(0, strcmp(symbol_name_, "home_object_symbol")); |
| 179 | value_ = isolate->factory()->home_object_symbol(); |
| 180 | } |
| 181 | break; |
| 182 | case NUMBER_WITH_DOT: |
| 183 | case NUMBER: |
| 184 | value_ = isolate->factory()->NewNumber(number_, TENURED); |
| 185 | break; |
| 186 | case SMI: |
| 187 | value_ = handle(Smi::FromInt(smi_), isolate); |
| 188 | break; |
| 189 | case BOOLEAN: |
| 190 | if (bool_) { |
| 191 | value_ = isolate->factory()->true_value(); |
| 192 | } else { |
| 193 | value_ = isolate->factory()->false_value(); |
| 194 | } |
| 195 | break; |
| 196 | case NULL_TYPE: |
| 197 | value_ = isolate->factory()->null_value(); |
| 198 | break; |
| 199 | case THE_HOLE: |
| 200 | value_ = isolate->factory()->the_hole_value(); |
| 201 | break; |
| 202 | case UNDEFINED: |
| 203 | value_ = isolate->factory()->undefined_value(); |
| 204 | break; |
| 205 | } |
| 206 | } |
| 207 | |
| 208 | |
| 209 | AstRawString* AstValueFactory::GetOneByteStringInternal( |
| 210 | Vector<const uint8_t> literal) { |
| 211 | uint32_t hash = StringHasher::HashSequentialString<uint8_t>( |
| 212 | literal.start(), literal.length(), hash_seed_); |
| 213 | return GetString(hash, true, literal); |
| 214 | } |
| 215 | |
| 216 | |
| 217 | AstRawString* AstValueFactory::GetTwoByteStringInternal( |
| 218 | Vector<const uint16_t> literal) { |
| 219 | uint32_t hash = StringHasher::HashSequentialString<uint16_t>( |
| 220 | literal.start(), literal.length(), hash_seed_); |
| 221 | return GetString(hash, false, Vector<const byte>::cast(literal)); |
| 222 | } |
| 223 | |
| 224 | |
| 225 | const AstRawString* AstValueFactory::GetString(Handle<String> literal) { |
| 226 | // For the FlatContent to stay valid, we shouldn't do any heap |
| 227 | // allocation. Make sure we won't try to internalize the string in GetString. |
| 228 | AstRawString* result = NULL; |
| 229 | Isolate* saved_isolate = isolate_; |
| 230 | isolate_ = NULL; |
| 231 | { |
| 232 | DisallowHeapAllocation no_gc; |
| 233 | String::FlatContent content = literal->GetFlatContent(); |
| 234 | if (content.IsOneByte()) { |
| 235 | result = GetOneByteStringInternal(content.ToOneByteVector()); |
| 236 | } else { |
| 237 | DCHECK(content.IsTwoByte()); |
| 238 | result = GetTwoByteStringInternal(content.ToUC16Vector()); |
| 239 | } |
| 240 | } |
| 241 | isolate_ = saved_isolate; |
| 242 | if (isolate_) result->Internalize(isolate_); |
| 243 | return result; |
| 244 | } |
| 245 | |
| 246 | |
| 247 | const AstConsString* AstValueFactory::NewConsString( |
| 248 | const AstString* left, const AstString* right) { |
| 249 | // This Vector will be valid as long as the Collector is alive (meaning that |
| 250 | // the AstRawString will not be moved). |
| 251 | AstConsString* new_string = new (zone_) AstConsString(left, right); |
| 252 | strings_.Add(new_string); |
| 253 | if (isolate_) { |
| 254 | new_string->Internalize(isolate_); |
| 255 | } |
| 256 | return new_string; |
| 257 | } |
| 258 | |
| 259 | |
| 260 | void AstValueFactory::Internalize(Isolate* isolate) { |
| 261 | if (isolate_) { |
| 262 | // Everything is already internalized. |
| 263 | return; |
| 264 | } |
| 265 | // Strings need to be internalized before values, because values refer to |
| 266 | // strings. |
| 267 | for (int i = 0; i < strings_.length(); ++i) { |
| 268 | strings_[i]->Internalize(isolate); |
| 269 | } |
| 270 | for (int i = 0; i < values_.length(); ++i) { |
| 271 | values_[i]->Internalize(isolate); |
| 272 | } |
| 273 | isolate_ = isolate; |
| 274 | } |
| 275 | |
| 276 | |
| 277 | const AstValue* AstValueFactory::NewString(const AstRawString* string) { |
| 278 | AstValue* value = new (zone_) AstValue(string); |
| 279 | DCHECK(string != NULL); |
| 280 | if (isolate_) { |
| 281 | value->Internalize(isolate_); |
| 282 | } |
| 283 | values_.Add(value); |
| 284 | return value; |
| 285 | } |
| 286 | |
| 287 | |
| 288 | const AstValue* AstValueFactory::NewSymbol(const char* name) { |
| 289 | AstValue* value = new (zone_) AstValue(name); |
| 290 | if (isolate_) { |
| 291 | value->Internalize(isolate_); |
| 292 | } |
| 293 | values_.Add(value); |
| 294 | return value; |
| 295 | } |
| 296 | |
| 297 | |
| 298 | const AstValue* AstValueFactory::NewNumber(double number, bool with_dot) { |
| 299 | AstValue* value = new (zone_) AstValue(number, with_dot); |
| 300 | if (isolate_) { |
| 301 | value->Internalize(isolate_); |
| 302 | } |
| 303 | values_.Add(value); |
| 304 | return value; |
| 305 | } |
| 306 | |
| 307 | |
| 308 | const AstValue* AstValueFactory::NewSmi(int number) { |
| 309 | AstValue* value = |
| 310 | new (zone_) AstValue(AstValue::SMI, number); |
| 311 | if (isolate_) { |
| 312 | value->Internalize(isolate_); |
| 313 | } |
| 314 | values_.Add(value); |
| 315 | return value; |
| 316 | } |
| 317 | |
| 318 | |
| 319 | #define GENERATE_VALUE_GETTER(value, initializer) \ |
| 320 | if (!value) { \ |
| 321 | value = new (zone_) AstValue(initializer); \ |
| 322 | if (isolate_) { \ |
| 323 | value->Internalize(isolate_); \ |
| 324 | } \ |
| 325 | values_.Add(value); \ |
| 326 | } \ |
| 327 | return value; |
| 328 | |
| 329 | |
| 330 | const AstValue* AstValueFactory::NewBoolean(bool b) { |
| 331 | if (b) { |
| 332 | GENERATE_VALUE_GETTER(true_value_, true); |
| 333 | } else { |
| 334 | GENERATE_VALUE_GETTER(false_value_, false); |
| 335 | } |
| 336 | } |
| 337 | |
| 338 | |
| 339 | const AstValue* AstValueFactory::NewNull() { |
| 340 | GENERATE_VALUE_GETTER(null_value_, AstValue::NULL_TYPE); |
| 341 | } |
| 342 | |
| 343 | |
| 344 | const AstValue* AstValueFactory::NewUndefined() { |
| 345 | GENERATE_VALUE_GETTER(undefined_value_, AstValue::UNDEFINED); |
| 346 | } |
| 347 | |
| 348 | |
| 349 | const AstValue* AstValueFactory::NewTheHole() { |
| 350 | GENERATE_VALUE_GETTER(the_hole_value_, AstValue::THE_HOLE); |
| 351 | } |
| 352 | |
| 353 | |
| 354 | #undef GENERATE_VALUE_GETTER |
| 355 | |
| 356 | AstRawString* AstValueFactory::GetString(uint32_t hash, bool is_one_byte, |
| 357 | Vector<const byte> literal_bytes) { |
| 358 | // literal_bytes here points to whatever the user passed, and this is OK |
| 359 | // because we use vector_compare (which checks the contents) to compare |
| 360 | // against the AstRawStrings which are in the string_table_. We should not |
| 361 | // return this AstRawString. |
| 362 | AstRawString key(is_one_byte, literal_bytes, hash); |
Ben Murdoch | 61f157c | 2016-09-16 13:49:30 +0100 | [diff] [blame] | 363 | base::HashMap::Entry* entry = string_table_.LookupOrInsert(&key, hash); |
Ben Murdoch | 4a90d5f | 2016-03-22 12:00:34 +0000 | [diff] [blame] | 364 | if (entry->value == NULL) { |
| 365 | // Copy literal contents for later comparison. |
| 366 | int length = literal_bytes.length(); |
| 367 | byte* new_literal_bytes = zone_->NewArray<byte>(length); |
| 368 | memcpy(new_literal_bytes, literal_bytes.start(), length); |
| 369 | AstRawString* new_string = new (zone_) AstRawString( |
| 370 | is_one_byte, Vector<const byte>(new_literal_bytes, length), hash); |
| 371 | entry->key = new_string; |
| 372 | strings_.Add(new_string); |
| 373 | if (isolate_) { |
| 374 | new_string->Internalize(isolate_); |
| 375 | } |
| 376 | entry->value = reinterpret_cast<void*>(1); |
| 377 | } |
| 378 | return reinterpret_cast<AstRawString*>(entry->key); |
| 379 | } |
| 380 | |
| 381 | |
| 382 | bool AstValueFactory::AstRawStringCompare(void* a, void* b) { |
| 383 | const AstRawString* lhs = static_cast<AstRawString*>(a); |
| 384 | const AstRawString* rhs = static_cast<AstRawString*>(b); |
| 385 | if (lhs->length() != rhs->length()) return false; |
| 386 | if (lhs->hash() != rhs->hash()) return false; |
| 387 | const unsigned char* l = lhs->raw_data(); |
| 388 | const unsigned char* r = rhs->raw_data(); |
| 389 | size_t length = rhs->length(); |
| 390 | if (lhs->is_one_byte()) { |
| 391 | if (rhs->is_one_byte()) { |
| 392 | return CompareCharsUnsigned(reinterpret_cast<const uint8_t*>(l), |
| 393 | reinterpret_cast<const uint8_t*>(r), |
| 394 | length) == 0; |
| 395 | } else { |
| 396 | return CompareCharsUnsigned(reinterpret_cast<const uint8_t*>(l), |
| 397 | reinterpret_cast<const uint16_t*>(r), |
| 398 | length) == 0; |
| 399 | } |
| 400 | } else { |
| 401 | if (rhs->is_one_byte()) { |
| 402 | return CompareCharsUnsigned(reinterpret_cast<const uint16_t*>(l), |
| 403 | reinterpret_cast<const uint8_t*>(r), |
| 404 | length) == 0; |
| 405 | } else { |
| 406 | return CompareCharsUnsigned(reinterpret_cast<const uint16_t*>(l), |
| 407 | reinterpret_cast<const uint16_t*>(r), |
| 408 | length) == 0; |
| 409 | } |
| 410 | } |
| 411 | } |
| 412 | } // namespace internal |
| 413 | } // namespace v8 |