1 // Copyright Joyent, Inc. and other Node contributors.
2 //
3 // Permission is hereby granted, free of charge, to any person obtaining a
4 // copy of this software and associated documentation files (the
5 // "Software"), to deal in the Software without restriction, including
6 // without limitation the rights to use, copy, modify, merge, publish,
7 // distribute, sublicense, and/or sell copies of the Software, and to permit
8 // persons to whom the Software is furnished to do so, subject to the
9 // following conditions:
10 //
11 // The above copyright notice and this permission notice shall be included
12 // in all copies or substantial portions of the Software.
13 //
14 // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
15 // OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
16 // MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
17 // NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
18 // DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
19 // OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
20 // USE OR OTHER DEALINGS IN THE SOFTWARE.
21
22 #include "string_bytes.h"
23
24 #include "base64-inl.h"
25 #include "env-inl.h"
26 #include "node_buffer.h"
27 #include "node_errors.h"
28 #include "util.h"
29
30 #include <climits>
31 #include <cstring> // memcpy
32
33 #include <algorithm>
34
35 // When creating strings >= this length v8's gc spins up and consumes
36 // most of the execution time. For these cases it's more performant to
37 // use external string resources.
38 #define EXTERN_APEX 0xFBEE9
39
40 namespace node {
41
42 using v8::HandleScope;
43 using v8::Isolate;
44 using v8::Just;
45 using v8::Local;
46 using v8::Maybe;
47 using v8::MaybeLocal;
48 using v8::Nothing;
49 using v8::String;
50 using v8::Value;
51
52 namespace {
53
54 template <typename ResourceType, typename TypeName>
55 class ExternString: public ResourceType {
56 public:
~ExternString()57 ~ExternString() override {
58 free(const_cast<TypeName*>(data_));
59 isolate()->AdjustAmountOfExternalAllocatedMemory(-byte_length());
60 }
61
data() const62 const TypeName* data() const override {
63 return data_;
64 }
65
length() const66 size_t length() const override {
67 return length_;
68 }
69
byte_length() const70 int64_t byte_length() const {
71 return length() * sizeof(*data());
72 }
73
NewFromCopy(Isolate * isolate,const TypeName * data,size_t length,Local<Value> * error)74 static MaybeLocal<Value> NewFromCopy(Isolate* isolate,
75 const TypeName* data,
76 size_t length,
77 Local<Value>* error) {
78 if (length == 0)
79 return String::Empty(isolate);
80
81 if (length < EXTERN_APEX)
82 return NewSimpleFromCopy(isolate, data, length, error);
83
84 TypeName* new_data = node::UncheckedMalloc<TypeName>(length);
85 if (new_data == nullptr) {
86 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
87 return MaybeLocal<Value>();
88 }
89 memcpy(new_data, data, length * sizeof(*new_data));
90
91 return ExternString<ResourceType, TypeName>::New(isolate,
92 new_data,
93 length,
94 error);
95 }
96
97 // uses "data" for external resource, and will be free'd on gc
New(Isolate * isolate,TypeName * data,size_t length,Local<Value> * error)98 static MaybeLocal<Value> New(Isolate* isolate,
99 TypeName* data,
100 size_t length,
101 Local<Value>* error) {
102 if (length == 0)
103 return String::Empty(isolate);
104
105 if (length < EXTERN_APEX) {
106 MaybeLocal<Value> str = NewSimpleFromCopy(isolate, data, length, error);
107 free(data);
108 return str;
109 }
110
111 ExternString* h_str = new ExternString<ResourceType, TypeName>(isolate,
112 data,
113 length);
114 MaybeLocal<Value> str = NewExternal(isolate, h_str);
115 isolate->AdjustAmountOfExternalAllocatedMemory(h_str->byte_length());
116
117 if (str.IsEmpty()) {
118 delete h_str;
119 *error = node::ERR_STRING_TOO_LONG(isolate);
120 return MaybeLocal<Value>();
121 }
122
123 return str.ToLocalChecked();
124 }
125
isolate() const126 inline Isolate* isolate() const { return isolate_; }
127
128 private:
ExternString(Isolate * isolate,const TypeName * data,size_t length)129 ExternString(Isolate* isolate, const TypeName* data, size_t length)
130 : isolate_(isolate), data_(data), length_(length) { }
131 static MaybeLocal<Value> NewExternal(Isolate* isolate,
132 ExternString* h_str);
133
134 // This method does not actually create ExternString instances.
135 static MaybeLocal<Value> NewSimpleFromCopy(Isolate* isolate,
136 const TypeName* data,
137 size_t length,
138 Local<Value>* error);
139
140 Isolate* isolate_;
141 const TypeName* data_;
142 size_t length_;
143 };
144
145
146 typedef ExternString<String::ExternalOneByteStringResource,
147 char> ExternOneByteString;
148 typedef ExternString<String::ExternalStringResource,
149 uint16_t> ExternTwoByteString;
150
151
152 template <>
NewExternal(Isolate * isolate,ExternOneByteString * h_str)153 MaybeLocal<Value> ExternOneByteString::NewExternal(
154 Isolate* isolate, ExternOneByteString* h_str) {
155 return String::NewExternalOneByte(isolate, h_str).FromMaybe(Local<Value>());
156 }
157
158
159 template <>
NewExternal(Isolate * isolate,ExternTwoByteString * h_str)160 MaybeLocal<Value> ExternTwoByteString::NewExternal(
161 Isolate* isolate, ExternTwoByteString* h_str) {
162 return String::NewExternalTwoByte(isolate, h_str).FromMaybe(Local<Value>());
163 }
164
165 template <>
NewSimpleFromCopy(Isolate * isolate,const char * data,size_t length,Local<Value> * error)166 MaybeLocal<Value> ExternOneByteString::NewSimpleFromCopy(Isolate* isolate,
167 const char* data,
168 size_t length,
169 Local<Value>* error) {
170 MaybeLocal<String> str =
171 String::NewFromOneByte(isolate,
172 reinterpret_cast<const uint8_t*>(data),
173 v8::NewStringType::kNormal,
174 length);
175 if (str.IsEmpty()) {
176 *error = node::ERR_STRING_TOO_LONG(isolate);
177 return MaybeLocal<Value>();
178 }
179 return str.ToLocalChecked();
180 }
181
182
183 template <>
NewSimpleFromCopy(Isolate * isolate,const uint16_t * data,size_t length,Local<Value> * error)184 MaybeLocal<Value> ExternTwoByteString::NewSimpleFromCopy(Isolate* isolate,
185 const uint16_t* data,
186 size_t length,
187 Local<Value>* error) {
188 MaybeLocal<String> str =
189 String::NewFromTwoByte(isolate,
190 data,
191 v8::NewStringType::kNormal,
192 length);
193 if (str.IsEmpty()) {
194 *error = node::ERR_STRING_TOO_LONG(isolate);
195 return MaybeLocal<Value>();
196 }
197 return str.ToLocalChecked();
198 }
199
200 } // anonymous namespace
201
202 // supports regular and URL-safe base64
203 const int8_t unbase64_table[256] =
204 { -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -2, -1, -1, -2, -1, -1,
205 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
206 -2, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 62, -1, 62, -1, 63,
207 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, -1, -1, -1, -1,
208 -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
209 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, -1, -1, -1, -1, 63,
210 -1, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
211 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, -1, -1, -1, -1, -1,
212 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
213 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
214 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
215 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
216 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
217 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
218 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
219 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1
220 };
221
222
223 static const int8_t unhex_table[256] =
224 { -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
225 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
226 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
227 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, -1, -1, -1, -1, -1,
228 -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1,
229 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
230 -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1,
231 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
232 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
233 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
234 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
235 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
236 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
237 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
238 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
239 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1
240 };
241
unhex(uint8_t x)242 static inline unsigned unhex(uint8_t x) {
243 return unhex_table[x];
244 }
245
246 template <typename TypeName>
hex_decode(char * buf,size_t len,const TypeName * src,const size_t srcLen)247 static size_t hex_decode(char* buf,
248 size_t len,
249 const TypeName* src,
250 const size_t srcLen) {
251 size_t i;
252 for (i = 0; i < len && i * 2 + 1 < srcLen; ++i) {
253 unsigned a = unhex(static_cast<uint8_t>(src[i * 2 + 0]));
254 unsigned b = unhex(static_cast<uint8_t>(src[i * 2 + 1]));
255 if (!~a || !~b)
256 return i;
257 buf[i] = (a << 4) | b;
258 }
259
260 return i;
261 }
262
WriteUCS2(Isolate * isolate,char * buf,size_t buflen,Local<String> str,int flags,size_t * chars_written)263 size_t StringBytes::WriteUCS2(Isolate* isolate,
264 char* buf,
265 size_t buflen,
266 Local<String> str,
267 int flags,
268 size_t* chars_written) {
269 uint16_t* const dst = reinterpret_cast<uint16_t*>(buf);
270
271 size_t max_chars = buflen / sizeof(*dst);
272 if (max_chars == 0) {
273 return 0;
274 }
275
276 uint16_t* const aligned_dst = AlignUp(dst, sizeof(*dst));
277 size_t nchars;
278 if (aligned_dst == dst) {
279 nchars = str->Write(isolate, dst, 0, max_chars, flags);
280 *chars_written = nchars;
281 return nchars * sizeof(*dst);
282 }
283
284 CHECK_EQ(reinterpret_cast<uintptr_t>(aligned_dst) % sizeof(*dst), 0);
285
286 // Write all but the last char
287 max_chars = std::min(max_chars, static_cast<size_t>(str->Length()));
288 if (max_chars == 0) return 0;
289 nchars = str->Write(isolate, aligned_dst, 0, max_chars - 1, flags);
290 CHECK_EQ(nchars, max_chars - 1);
291
292 // Shift everything to unaligned-left
293 memmove(dst, aligned_dst, nchars * sizeof(*dst));
294
295 // One more char to be written
296 uint16_t last;
297 CHECK_EQ(str->Write(isolate, &last, nchars, 1, flags), 1);
298 memcpy(buf + nchars * sizeof(*dst), &last, sizeof(last));
299 nchars++;
300
301 *chars_written = nchars;
302 return nchars * sizeof(*dst);
303 }
304
305
Write(Isolate * isolate,char * buf,size_t buflen,Local<Value> val,enum encoding encoding,int * chars_written)306 size_t StringBytes::Write(Isolate* isolate,
307 char* buf,
308 size_t buflen,
309 Local<Value> val,
310 enum encoding encoding,
311 int* chars_written) {
312 HandleScope scope(isolate);
313 size_t nbytes;
314 int nchars;
315
316 if (chars_written == nullptr)
317 chars_written = &nchars;
318
319 CHECK(val->IsString() == true);
320 Local<String> str = val.As<String>();
321
322 int flags = String::HINT_MANY_WRITES_EXPECTED |
323 String::NO_NULL_TERMINATION |
324 String::REPLACE_INVALID_UTF8;
325
326 switch (encoding) {
327 case ASCII:
328 case LATIN1:
329 if (str->IsExternalOneByte()) {
330 auto ext = str->GetExternalOneByteStringResource();
331 nbytes = std::min(buflen, ext->length());
332 memcpy(buf, ext->data(), nbytes);
333 } else {
334 uint8_t* const dst = reinterpret_cast<uint8_t*>(buf);
335 nbytes = str->WriteOneByte(isolate, dst, 0, buflen, flags);
336 }
337 *chars_written = nbytes;
338 break;
339
340 case BUFFER:
341 case UTF8:
342 nbytes = str->WriteUtf8(isolate, buf, buflen, chars_written, flags);
343 break;
344
345 case UCS2: {
346 size_t nchars;
347
348 nbytes = WriteUCS2(isolate, buf, buflen, str, flags, &nchars);
349 *chars_written = static_cast<int>(nchars);
350
351 // Node's "ucs2" encoding wants LE character data stored in
352 // the Buffer, so we need to reorder on BE platforms. See
353 // https://nodejs.org/api/buffer.html regarding Node's "ucs2"
354 // encoding specification
355 if (IsBigEndian())
356 SwapBytes16(buf, nbytes);
357
358 break;
359 }
360
361 case BASE64URL:
362 // Fall through
363 case BASE64:
364 if (str->IsExternalOneByte()) {
365 auto ext = str->GetExternalOneByteStringResource();
366 nbytes = base64_decode(buf, buflen, ext->data(), ext->length());
367 } else {
368 String::Value value(isolate, str);
369 nbytes = base64_decode(buf, buflen, *value, value.length());
370 }
371 *chars_written = nbytes;
372 break;
373
374 case HEX:
375 if (str->IsExternalOneByte()) {
376 auto ext = str->GetExternalOneByteStringResource();
377 nbytes = hex_decode(buf, buflen, ext->data(), ext->length());
378 } else {
379 String::Value value(isolate, str);
380 nbytes = hex_decode(buf, buflen, *value, value.length());
381 }
382 *chars_written = nbytes;
383 break;
384
385 default:
386 CHECK(0 && "unknown encoding");
387 break;
388 }
389
390 return nbytes;
391 }
392
393
394 // Quick and dirty size calculation
395 // Will always be at least big enough, but may have some extra
396 // UTF8 can be as much as 3x the size, Base64 can have 1-2 extra bytes
StorageSize(Isolate * isolate,Local<Value> val,enum encoding encoding)397 Maybe<size_t> StringBytes::StorageSize(Isolate* isolate,
398 Local<Value> val,
399 enum encoding encoding) {
400 HandleScope scope(isolate);
401 size_t data_size = 0;
402 bool is_buffer = Buffer::HasInstance(val);
403
404 if (is_buffer && (encoding == BUFFER || encoding == LATIN1)) {
405 return Just(Buffer::Length(val));
406 }
407
408 Local<String> str;
409 if (!val->ToString(isolate->GetCurrentContext()).ToLocal(&str))
410 return Nothing<size_t>();
411
412 switch (encoding) {
413 case ASCII:
414 case LATIN1:
415 data_size = str->Length();
416 break;
417
418 case BUFFER:
419 case UTF8:
420 // A single UCS2 codepoint never takes up more than 3 utf8 bytes.
421 // It is an exercise for the caller to decide when a string is
422 // long enough to justify calling Size() instead of StorageSize()
423 data_size = 3 * str->Length();
424 break;
425
426 case UCS2:
427 data_size = str->Length() * sizeof(uint16_t);
428 break;
429
430 case BASE64URL:
431 // Fall through
432 case BASE64:
433 data_size = base64_decoded_size_fast(str->Length());
434 break;
435
436 case HEX:
437 CHECK(str->Length() % 2 == 0 && "invalid hex string length");
438 data_size = str->Length() / 2;
439 break;
440
441 default:
442 CHECK(0 && "unknown encoding");
443 break;
444 }
445
446 return Just(data_size);
447 }
448
Size(Isolate * isolate,Local<Value> val,enum encoding encoding)449 Maybe<size_t> StringBytes::Size(Isolate* isolate,
450 Local<Value> val,
451 enum encoding encoding) {
452 HandleScope scope(isolate);
453
454 if (Buffer::HasInstance(val) && (encoding == BUFFER || encoding == LATIN1))
455 return Just(Buffer::Length(val));
456
457 Local<String> str;
458 if (!val->ToString(isolate->GetCurrentContext()).ToLocal(&str))
459 return Nothing<size_t>();
460
461 switch (encoding) {
462 case ASCII:
463 case LATIN1:
464 return Just<size_t>(str->Length());
465
466 case BUFFER:
467 case UTF8:
468 return Just<size_t>(str->Utf8Length(isolate));
469
470 case UCS2:
471 return Just(str->Length() * sizeof(uint16_t));
472
473 case BASE64URL:
474 // Fall through
475 case BASE64: {
476 String::Value value(isolate, str);
477 return Just(base64_decoded_size(*value, value.length()));
478 }
479
480 case HEX:
481 return Just<size_t>(str->Length() / 2);
482 }
483
484 UNREACHABLE();
485 }
486
487
488
489
contains_non_ascii_slow(const char * buf,size_t len)490 static bool contains_non_ascii_slow(const char* buf, size_t len) {
491 for (size_t i = 0; i < len; ++i) {
492 if (buf[i] & 0x80)
493 return true;
494 }
495 return false;
496 }
497
498
contains_non_ascii(const char * src,size_t len)499 static bool contains_non_ascii(const char* src, size_t len) {
500 if (len < 16) {
501 return contains_non_ascii_slow(src, len);
502 }
503
504 const unsigned bytes_per_word = sizeof(uintptr_t);
505 const unsigned align_mask = bytes_per_word - 1;
506 const unsigned unaligned = reinterpret_cast<uintptr_t>(src) & align_mask;
507
508 if (unaligned > 0) {
509 const unsigned n = bytes_per_word - unaligned;
510 if (contains_non_ascii_slow(src, n))
511 return true;
512 src += n;
513 len -= n;
514 }
515
516
517 #if defined(_WIN64) || defined(_LP64)
518 const uintptr_t mask = 0x8080808080808080ll;
519 #else
520 const uintptr_t mask = 0x80808080l;
521 #endif
522
523 const uintptr_t* srcw = reinterpret_cast<const uintptr_t*>(src);
524
525 for (size_t i = 0, n = len / bytes_per_word; i < n; ++i) {
526 if (srcw[i] & mask)
527 return true;
528 }
529
530 const unsigned remainder = len & align_mask;
531 if (remainder > 0) {
532 const size_t offset = len - remainder;
533 if (contains_non_ascii_slow(src + offset, remainder))
534 return true;
535 }
536
537 return false;
538 }
539
540
force_ascii_slow(const char * src,char * dst,size_t len)541 static void force_ascii_slow(const char* src, char* dst, size_t len) {
542 for (size_t i = 0; i < len; ++i) {
543 dst[i] = src[i] & 0x7f;
544 }
545 }
546
547
force_ascii(const char * src,char * dst,size_t len)548 static void force_ascii(const char* src, char* dst, size_t len) {
549 if (len < 16) {
550 force_ascii_slow(src, dst, len);
551 return;
552 }
553
554 const unsigned bytes_per_word = sizeof(uintptr_t);
555 const unsigned align_mask = bytes_per_word - 1;
556 const unsigned src_unalign = reinterpret_cast<uintptr_t>(src) & align_mask;
557 const unsigned dst_unalign = reinterpret_cast<uintptr_t>(dst) & align_mask;
558
559 if (src_unalign > 0) {
560 if (src_unalign == dst_unalign) {
561 const unsigned unalign = bytes_per_word - src_unalign;
562 force_ascii_slow(src, dst, unalign);
563 src += unalign;
564 dst += unalign;
565 len -= src_unalign;
566 } else {
567 force_ascii_slow(src, dst, len);
568 return;
569 }
570 }
571
572 #if defined(_WIN64) || defined(_LP64)
573 const uintptr_t mask = ~0x8080808080808080ll;
574 #else
575 const uintptr_t mask = ~0x80808080l;
576 #endif
577
578 const uintptr_t* srcw = reinterpret_cast<const uintptr_t*>(src);
579 uintptr_t* dstw = reinterpret_cast<uintptr_t*>(dst);
580
581 for (size_t i = 0, n = len / bytes_per_word; i < n; ++i) {
582 dstw[i] = srcw[i] & mask;
583 }
584
585 const unsigned remainder = len & align_mask;
586 if (remainder > 0) {
587 const size_t offset = len - remainder;
588 force_ascii_slow(src + offset, dst + offset, remainder);
589 }
590 }
591
592
hex_encode(const char * src,size_t slen,char * dst,size_t dlen)593 size_t StringBytes::hex_encode(
594 const char* src,
595 size_t slen,
596 char* dst,
597 size_t dlen) {
598 // We know how much we'll write, just make sure that there's space.
599 CHECK(dlen >= slen * 2 &&
600 "not enough space provided for hex encode");
601
602 dlen = slen * 2;
603 for (uint32_t i = 0, k = 0; k < dlen; i += 1, k += 2) {
604 static const char hex[] = "0123456789abcdef";
605 uint8_t val = static_cast<uint8_t>(src[i]);
606 dst[k + 0] = hex[val >> 4];
607 dst[k + 1] = hex[val & 15];
608 }
609
610 return dlen;
611 }
612
hex_encode(const char * src,size_t slen)613 std::string StringBytes::hex_encode(const char* src, size_t slen) {
614 size_t dlen = slen * 2;
615 std::string dst(dlen, '\0');
616 hex_encode(src, slen, &dst[0], dlen);
617 return dst;
618 }
619
620 #define CHECK_BUFLEN_IN_RANGE(len) \
621 do { \
622 if ((len) > Buffer::kMaxLength) { \
623 *error = node::ERR_BUFFER_TOO_LARGE(isolate); \
624 return MaybeLocal<Value>(); \
625 } \
626 } while (0)
627
628
Encode(Isolate * isolate,const char * buf,size_t buflen,enum encoding encoding,Local<Value> * error)629 MaybeLocal<Value> StringBytes::Encode(Isolate* isolate,
630 const char* buf,
631 size_t buflen,
632 enum encoding encoding,
633 Local<Value>* error) {
634 CHECK_BUFLEN_IN_RANGE(buflen);
635
636 if (!buflen && encoding != BUFFER) {
637 return String::Empty(isolate);
638 }
639
640 MaybeLocal<String> val;
641
642 switch (encoding) {
643 case BUFFER:
644 {
645 if (buflen > node::Buffer::kMaxLength) {
646 *error = node::ERR_BUFFER_TOO_LARGE(isolate);
647 return MaybeLocal<Value>();
648 }
649 auto maybe_buf = Buffer::Copy(isolate, buf, buflen);
650 Local<v8::Object> buf;
651 if (!maybe_buf.ToLocal(&buf)) {
652 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
653 }
654 return buf;
655 }
656
657 case ASCII:
658 if (contains_non_ascii(buf, buflen)) {
659 char* out = node::UncheckedMalloc(buflen);
660 if (out == nullptr) {
661 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
662 return MaybeLocal<Value>();
663 }
664 force_ascii(buf, out, buflen);
665 return ExternOneByteString::New(isolate, out, buflen, error);
666 } else {
667 return ExternOneByteString::NewFromCopy(isolate, buf, buflen, error);
668 }
669
670 case UTF8:
671 {
672 val = String::NewFromUtf8(isolate,
673 buf,
674 v8::NewStringType::kNormal,
675 buflen);
676 Local<String> str;
677 if (!val.ToLocal(&str)) {
678 *error = node::ERR_STRING_TOO_LONG(isolate);
679 }
680 return str;
681 }
682
683 case LATIN1:
684 return ExternOneByteString::NewFromCopy(isolate, buf, buflen, error);
685
686 case BASE64: {
687 size_t dlen = base64_encoded_size(buflen);
688 char* dst = node::UncheckedMalloc(dlen);
689 if (dst == nullptr) {
690 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
691 return MaybeLocal<Value>();
692 }
693
694 size_t written = base64_encode(buf, buflen, dst, dlen);
695 CHECK_EQ(written, dlen);
696
697 return ExternOneByteString::New(isolate, dst, dlen, error);
698 }
699
700 case BASE64URL: {
701 size_t dlen = base64_encoded_size(buflen, Base64Mode::URL);
702 char* dst = node::UncheckedMalloc(dlen);
703 if (dst == nullptr) {
704 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
705 return MaybeLocal<Value>();
706 }
707
708 size_t written = base64_encode(buf, buflen, dst, dlen, Base64Mode::URL);
709 CHECK_EQ(written, dlen);
710
711 return ExternOneByteString::New(isolate, dst, dlen, error);
712 }
713
714 case HEX: {
715 size_t dlen = buflen * 2;
716 char* dst = node::UncheckedMalloc(dlen);
717 if (dst == nullptr) {
718 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
719 return MaybeLocal<Value>();
720 }
721 size_t written = hex_encode(buf, buflen, dst, dlen);
722 CHECK_EQ(written, dlen);
723
724 return ExternOneByteString::New(isolate, dst, dlen, error);
725 }
726
727 case UCS2: {
728 if (IsBigEndian()) {
729 uint16_t* dst = node::UncheckedMalloc<uint16_t>(buflen / 2);
730 if (dst == nullptr) {
731 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
732 return MaybeLocal<Value>();
733 }
734 for (size_t i = 0, k = 0; k < buflen / 2; i += 2, k += 1) {
735 // The input is in *little endian*, because that's what Node.js
736 // expects, so the high byte comes after the low byte.
737 const uint8_t hi = static_cast<uint8_t>(buf[i + 1]);
738 const uint8_t lo = static_cast<uint8_t>(buf[i + 0]);
739 dst[k] = static_cast<uint16_t>(hi) << 8 | lo;
740 }
741 return ExternTwoByteString::New(isolate, dst, buflen / 2, error);
742 }
743 if (reinterpret_cast<uintptr_t>(buf) % 2 != 0) {
744 // Unaligned data still means we can't directly pass it to V8.
745 char* dst = node::UncheckedMalloc(buflen);
746 if (dst == nullptr) {
747 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
748 return MaybeLocal<Value>();
749 }
750 memcpy(dst, buf, buflen);
751 return ExternTwoByteString::New(
752 isolate, reinterpret_cast<uint16_t*>(dst), buflen / 2, error);
753 }
754 return ExternTwoByteString::NewFromCopy(
755 isolate, reinterpret_cast<const uint16_t*>(buf), buflen / 2, error);
756 }
757
758 default:
759 CHECK(0 && "unknown encoding");
760 break;
761 }
762
763 UNREACHABLE();
764 }
765
766
Encode(Isolate * isolate,const uint16_t * buf,size_t buflen,Local<Value> * error)767 MaybeLocal<Value> StringBytes::Encode(Isolate* isolate,
768 const uint16_t* buf,
769 size_t buflen,
770 Local<Value>* error) {
771 CHECK_BUFLEN_IN_RANGE(buflen);
772
773 // Node's "ucs2" encoding expects LE character data inside a
774 // Buffer, so we need to reorder on BE platforms. See
775 // https://nodejs.org/api/buffer.html regarding Node's "ucs2"
776 // encoding specification
777 if (IsBigEndian()) {
778 uint16_t* dst = node::UncheckedMalloc<uint16_t>(buflen);
779 if (dst == nullptr) {
780 *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate);
781 return MaybeLocal<Value>();
782 }
783 size_t nbytes = buflen * sizeof(uint16_t);
784 memcpy(dst, buf, nbytes);
785 SwapBytes16(reinterpret_cast<char*>(dst), nbytes);
786 return ExternTwoByteString::New(isolate, dst, buflen, error);
787 } else {
788 return ExternTwoByteString::NewFromCopy(isolate, buf, buflen, error);
789 }
790 }
791
Encode(Isolate * isolate,const char * buf,enum encoding encoding,Local<Value> * error)792 MaybeLocal<Value> StringBytes::Encode(Isolate* isolate,
793 const char* buf,
794 enum encoding encoding,
795 Local<Value>* error) {
796 const size_t len = strlen(buf);
797 return Encode(isolate, buf, len, encoding, error);
798 }
799
800 } // namespace node
801