/src/node/src/string_bytes.cc
Line | Count | Source (jump to first uncovered line) |
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 "simdutf.h" |
29 | | #include "util.h" |
30 | | |
31 | | #include <climits> |
32 | | #include <cstring> // memcpy |
33 | | |
34 | | #include <algorithm> |
35 | | |
36 | | // When creating strings >= this length v8's gc spins up and consumes |
37 | | // most of the execution time. For these cases it's more performant to |
38 | | // use external string resources. |
39 | 961 | #define EXTERN_APEX 0xFBEE9 |
40 | | |
41 | | namespace node { |
42 | | |
43 | | using v8::HandleScope; |
44 | | using v8::Isolate; |
45 | | using v8::Just; |
46 | | using v8::Local; |
47 | | using v8::Maybe; |
48 | | using v8::MaybeLocal; |
49 | | using v8::Nothing; |
50 | | using v8::String; |
51 | | using v8::Value; |
52 | | |
53 | | namespace { |
54 | | |
55 | | template <typename ResourceType, typename TypeName> |
56 | | class ExternString: public ResourceType { |
57 | | public: |
58 | 0 | ~ExternString() override { |
59 | 0 | free(const_cast<TypeName*>(data_)); |
60 | 0 | isolate()->AdjustAmountOfExternalAllocatedMemory(-byte_length()); |
61 | 0 | } Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::~ExternString() Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::~ExternString() |
62 | | |
63 | 0 | const TypeName* data() const override { |
64 | 0 | return data_; |
65 | 0 | } Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::data() const Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::data() const |
66 | | |
67 | 0 | size_t length() const override { |
68 | 0 | return length_; |
69 | 0 | } Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::length() const Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::length() const |
70 | | |
71 | 0 | int64_t byte_length() const { |
72 | 0 | return length() * sizeof(*data()); |
73 | 0 | } Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::byte_length() const Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::byte_length() const |
74 | | |
75 | | static MaybeLocal<Value> NewFromCopy(Isolate* isolate, |
76 | | const TypeName* data, |
77 | | size_t length, |
78 | 5 | Local<Value>* error) { |
79 | 5 | if (length == 0) |
80 | 0 | return String::Empty(isolate); |
81 | | |
82 | 5 | if (length < EXTERN_APEX) |
83 | 5 | return NewSimpleFromCopy(isolate, data, length, error); |
84 | | |
85 | 0 | TypeName* new_data = node::UncheckedMalloc<TypeName>(length); |
86 | 0 | if (new_data == nullptr) { |
87 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
88 | 0 | return MaybeLocal<Value>(); |
89 | 0 | } |
90 | 0 | memcpy(new_data, data, length * sizeof(*new_data)); |
91 | |
|
92 | 0 | return ExternString<ResourceType, TypeName>::New(isolate, |
93 | 0 | new_data, |
94 | 0 | length, |
95 | 0 | error); |
96 | 0 | } string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::NewFromCopy(v8::Isolate*, char const*, unsigned long, v8::Local<v8::Value>*) Line | Count | Source | 78 | 5 | Local<Value>* error) { | 79 | 5 | if (length == 0) | 80 | 0 | return String::Empty(isolate); | 81 | | | 82 | 5 | if (length < EXTERN_APEX) | 83 | 5 | return NewSimpleFromCopy(isolate, data, length, error); | 84 | | | 85 | 0 | TypeName* new_data = node::UncheckedMalloc<TypeName>(length); | 86 | 0 | if (new_data == nullptr) { | 87 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); | 88 | 0 | return MaybeLocal<Value>(); | 89 | 0 | } | 90 | 0 | memcpy(new_data, data, length * sizeof(*new_data)); | 91 | |
| 92 | 0 | return ExternString<ResourceType, TypeName>::New(isolate, | 93 | 0 | new_data, | 94 | 0 | length, | 95 | 0 | error); | 96 | 0 | } |
Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::NewFromCopy(v8::Isolate*, unsigned short const*, unsigned long, v8::Local<v8::Value>*) |
97 | | |
98 | | // uses "data" for external resource, and will be free'd on gc |
99 | | static MaybeLocal<Value> New(Isolate* isolate, |
100 | | TypeName* data, |
101 | | size_t length, |
102 | 956 | Local<Value>* error) { |
103 | 956 | if (length == 0) |
104 | 0 | return String::Empty(isolate); |
105 | | |
106 | 956 | if (length < EXTERN_APEX) { |
107 | 956 | MaybeLocal<Value> str = NewSimpleFromCopy(isolate, data, length, error); |
108 | 956 | free(data); |
109 | 956 | return str; |
110 | 956 | } |
111 | | |
112 | 0 | ExternString* h_str = new ExternString<ResourceType, TypeName>(isolate, |
113 | 0 | data, |
114 | 0 | length); |
115 | 0 | MaybeLocal<Value> str = NewExternal(isolate, h_str); |
116 | 0 | isolate->AdjustAmountOfExternalAllocatedMemory(h_str->byte_length()); |
117 | |
|
118 | 0 | if (str.IsEmpty()) { |
119 | 0 | delete h_str; |
120 | 0 | *error = node::ERR_STRING_TOO_LONG(isolate); |
121 | 0 | return MaybeLocal<Value>(); |
122 | 0 | } |
123 | | |
124 | 0 | return str.ToLocalChecked(); |
125 | 0 | } string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::New(v8::Isolate*, char*, unsigned long, v8::Local<v8::Value>*) Line | Count | Source | 102 | 956 | Local<Value>* error) { | 103 | 956 | if (length == 0) | 104 | 0 | return String::Empty(isolate); | 105 | | | 106 | 956 | if (length < EXTERN_APEX) { | 107 | 956 | MaybeLocal<Value> str = NewSimpleFromCopy(isolate, data, length, error); | 108 | 956 | free(data); | 109 | 956 | return str; | 110 | 956 | } | 111 | | | 112 | 0 | ExternString* h_str = new ExternString<ResourceType, TypeName>(isolate, | 113 | 0 | data, | 114 | 0 | length); | 115 | 0 | MaybeLocal<Value> str = NewExternal(isolate, h_str); | 116 | 0 | isolate->AdjustAmountOfExternalAllocatedMemory(h_str->byte_length()); | 117 | |
| 118 | 0 | if (str.IsEmpty()) { | 119 | 0 | delete h_str; | 120 | 0 | *error = node::ERR_STRING_TOO_LONG(isolate); | 121 | 0 | return MaybeLocal<Value>(); | 122 | 0 | } | 123 | | | 124 | 0 | return str.ToLocalChecked(); | 125 | 0 | } |
Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::New(v8::Isolate*, unsigned short*, unsigned long, v8::Local<v8::Value>*) |
126 | | |
127 | 0 | inline Isolate* isolate() const { return isolate_; } Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::isolate() const Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::isolate() const |
128 | | |
129 | | private: |
130 | | ExternString(Isolate* isolate, const TypeName* data, size_t length) |
131 | 0 | : isolate_(isolate), data_(data), length_(length) { } Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalOneByteStringResource, char>::ExternString(v8::Isolate*, char const*, unsigned long) Unexecuted instantiation: string_bytes.cc:node::(anonymous namespace)::ExternString<v8::String::ExternalStringResource, unsigned short>::ExternString(v8::Isolate*, unsigned short const*, unsigned long) |
132 | | static MaybeLocal<Value> NewExternal(Isolate* isolate, |
133 | | ExternString* h_str); |
134 | | |
135 | | // This method does not actually create ExternString instances. |
136 | | static MaybeLocal<Value> NewSimpleFromCopy(Isolate* isolate, |
137 | | const TypeName* data, |
138 | | size_t length, |
139 | | Local<Value>* error); |
140 | | |
141 | | Isolate* isolate_; |
142 | | const TypeName* data_; |
143 | | size_t length_; |
144 | | }; |
145 | | |
146 | | |
147 | | typedef ExternString<String::ExternalOneByteStringResource, |
148 | | char> ExternOneByteString; |
149 | | typedef ExternString<String::ExternalStringResource, |
150 | | uint16_t> ExternTwoByteString; |
151 | | |
152 | | |
153 | | template <> |
154 | | MaybeLocal<Value> ExternOneByteString::NewExternal( |
155 | 0 | Isolate* isolate, ExternOneByteString* h_str) { |
156 | 0 | return String::NewExternalOneByte(isolate, h_str).FromMaybe(Local<Value>()); |
157 | 0 | } |
158 | | |
159 | | |
160 | | template <> |
161 | | MaybeLocal<Value> ExternTwoByteString::NewExternal( |
162 | 0 | Isolate* isolate, ExternTwoByteString* h_str) { |
163 | 0 | return String::NewExternalTwoByte(isolate, h_str).FromMaybe(Local<Value>()); |
164 | 0 | } |
165 | | |
166 | | template <> |
167 | | MaybeLocal<Value> ExternOneByteString::NewSimpleFromCopy(Isolate* isolate, |
168 | | const char* data, |
169 | | size_t length, |
170 | 961 | Local<Value>* error) { |
171 | 961 | MaybeLocal<String> str = |
172 | 961 | String::NewFromOneByte(isolate, |
173 | 961 | reinterpret_cast<const uint8_t*>(data), |
174 | 961 | v8::NewStringType::kNormal, |
175 | 961 | length); |
176 | 961 | if (str.IsEmpty()) { |
177 | 0 | *error = node::ERR_STRING_TOO_LONG(isolate); |
178 | 0 | return MaybeLocal<Value>(); |
179 | 0 | } |
180 | 961 | return str.ToLocalChecked(); |
181 | 961 | } |
182 | | |
183 | | |
184 | | template <> |
185 | | MaybeLocal<Value> ExternTwoByteString::NewSimpleFromCopy(Isolate* isolate, |
186 | | const uint16_t* data, |
187 | | size_t length, |
188 | 0 | Local<Value>* error) { |
189 | 0 | MaybeLocal<String> str = |
190 | 0 | String::NewFromTwoByte(isolate, |
191 | 0 | data, |
192 | 0 | v8::NewStringType::kNormal, |
193 | 0 | length); |
194 | 0 | if (str.IsEmpty()) { |
195 | 0 | *error = node::ERR_STRING_TOO_LONG(isolate); |
196 | 0 | return MaybeLocal<Value>(); |
197 | 0 | } |
198 | 0 | return str.ToLocalChecked(); |
199 | 0 | } |
200 | | |
201 | | } // anonymous namespace |
202 | | |
203 | | // supports regular and URL-safe base64 |
204 | | const int8_t unbase64_table[256] = |
205 | | { -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -2, -1, -1, -2, -1, -1, |
206 | | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, |
207 | | -2, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 62, -1, 62, -1, 63, |
208 | | 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, -1, -1, -1, -1, |
209 | | -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, |
210 | | 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, -1, -1, -1, -1, 63, |
211 | | -1, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, |
212 | | 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, -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 | | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1 |
221 | | }; |
222 | | |
223 | | |
224 | | static const int8_t unhex_table[256] = |
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 | | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, |
228 | | 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, -1, -1, -1, -1, -1, |
229 | | -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1, |
230 | | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, |
231 | | -1, 10, 11, 12, 13, 14, 15, -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 | | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1 |
241 | | }; |
242 | | |
243 | 478 | static inline unsigned unhex(uint8_t x) { |
244 | 478 | return unhex_table[x]; |
245 | 478 | } |
246 | | |
247 | | template <typename TypeName> |
248 | | static size_t hex_decode(char* buf, |
249 | | size_t len, |
250 | | const TypeName* src, |
251 | 16 | const size_t srcLen) { |
252 | 16 | size_t i; |
253 | 252 | for (i = 0; i < len && i * 2 + 1 < srcLen; ++i) { |
254 | 239 | unsigned a = unhex(static_cast<uint8_t>(src[i * 2 + 0])); |
255 | 239 | unsigned b = unhex(static_cast<uint8_t>(src[i * 2 + 1])); |
256 | 239 | if (!~a || !~b) |
257 | 3 | return i; |
258 | 236 | buf[i] = (a << 4) | b; |
259 | 236 | } |
260 | | |
261 | 13 | return i; |
262 | 16 | } Unexecuted instantiation: string_bytes.cc:unsigned long node::hex_decode<char>(char*, unsigned long, char const*, unsigned long) string_bytes.cc:unsigned long node::hex_decode<unsigned short>(char*, unsigned long, unsigned short const*, unsigned long) Line | Count | Source | 251 | 16 | const size_t srcLen) { | 252 | 16 | size_t i; | 253 | 252 | for (i = 0; i < len && i * 2 + 1 < srcLen; ++i) { | 254 | 239 | unsigned a = unhex(static_cast<uint8_t>(src[i * 2 + 0])); | 255 | 239 | unsigned b = unhex(static_cast<uint8_t>(src[i * 2 + 1])); | 256 | 239 | if (!~a || !~b) | 257 | 3 | return i; | 258 | 236 | buf[i] = (a << 4) | b; | 259 | 236 | } | 260 | | | 261 | 13 | return i; | 262 | 16 | } |
|
263 | | |
264 | | size_t StringBytes::WriteUCS2( |
265 | 7 | Isolate* isolate, char* buf, size_t buflen, Local<String> str, int flags) { |
266 | 7 | uint16_t* const dst = reinterpret_cast<uint16_t*>(buf); |
267 | | |
268 | 7 | size_t max_chars = buflen / sizeof(*dst); |
269 | 7 | if (max_chars == 0) { |
270 | 0 | return 0; |
271 | 0 | } |
272 | | |
273 | 7 | uint16_t* const aligned_dst = AlignUp(dst, sizeof(*dst)); |
274 | 7 | size_t nchars; |
275 | 7 | if (aligned_dst == dst) { |
276 | 7 | nchars = str->Write(isolate, dst, 0, max_chars, flags); |
277 | 7 | return nchars * sizeof(*dst); |
278 | 7 | } |
279 | | |
280 | 0 | CHECK_EQ(reinterpret_cast<uintptr_t>(aligned_dst) % sizeof(*dst), 0); |
281 | | |
282 | | // Write all but the last char |
283 | 0 | max_chars = std::min(max_chars, static_cast<size_t>(str->Length())); |
284 | 0 | if (max_chars == 0) { |
285 | 0 | return 0; |
286 | 0 | } |
287 | 0 | nchars = str->Write(isolate, aligned_dst, 0, max_chars - 1, flags); |
288 | 0 | CHECK_EQ(nchars, max_chars - 1); |
289 | | |
290 | | // Shift everything to unaligned-left |
291 | 0 | memmove(dst, aligned_dst, nchars * sizeof(*dst)); |
292 | | |
293 | | // One more char to be written |
294 | 0 | uint16_t last; |
295 | 0 | CHECK_EQ(str->Write(isolate, &last, nchars, 1, flags), 1); |
296 | 0 | memcpy(buf + nchars * sizeof(*dst), &last, sizeof(last)); |
297 | 0 | nchars++; |
298 | |
|
299 | 0 | return nchars * sizeof(*dst); |
300 | 0 | } |
301 | | |
302 | | size_t StringBytes::Write(Isolate* isolate, |
303 | | char* buf, |
304 | | size_t buflen, |
305 | | Local<Value> val, |
306 | 149k | enum encoding encoding) { |
307 | 149k | HandleScope scope(isolate); |
308 | 149k | size_t nbytes; |
309 | | |
310 | 149k | CHECK(val->IsString() == true); |
311 | 149k | Local<String> str = val.As<String>(); |
312 | | |
313 | 149k | int flags = String::HINT_MANY_WRITES_EXPECTED | |
314 | 149k | String::NO_NULL_TERMINATION | |
315 | 149k | String::REPLACE_INVALID_UTF8; |
316 | | |
317 | 149k | switch (encoding) { |
318 | 3.60k | case ASCII: |
319 | 3.60k | case LATIN1: |
320 | 3.60k | if (str->IsExternalOneByte()) { |
321 | 0 | auto ext = str->GetExternalOneByteStringResource(); |
322 | 0 | nbytes = std::min(buflen, ext->length()); |
323 | 0 | memcpy(buf, ext->data(), nbytes); |
324 | 3.60k | } else { |
325 | 3.60k | uint8_t* const dst = reinterpret_cast<uint8_t*>(buf); |
326 | 3.60k | nbytes = str->WriteOneByte(isolate, dst, 0, buflen, flags); |
327 | 3.60k | } |
328 | 3.60k | break; |
329 | | |
330 | 0 | case BUFFER: |
331 | 145k | case UTF8: |
332 | 145k | nbytes = str->WriteUtf8(isolate, buf, buflen, nullptr, flags); |
333 | 145k | break; |
334 | | |
335 | 7 | case UCS2: { |
336 | 7 | nbytes = WriteUCS2(isolate, buf, buflen, str, flags); |
337 | | |
338 | | // Node's "ucs2" encoding wants LE character data stored in |
339 | | // the Buffer, so we need to reorder on BE platforms. See |
340 | | // https://nodejs.org/api/buffer.html regarding Node's "ucs2" |
341 | | // encoding specification |
342 | 7 | if (IsBigEndian()) |
343 | 0 | SwapBytes16(buf, nbytes); |
344 | | |
345 | 7 | break; |
346 | 0 | } |
347 | | |
348 | 0 | case BASE64URL: |
349 | | // Fall through |
350 | 2 | case BASE64: |
351 | 2 | if (str->IsExternalOneByte()) { |
352 | 0 | auto ext = str->GetExternalOneByteStringResource(); |
353 | 0 | nbytes = base64_decode(buf, buflen, ext->data(), ext->length()); |
354 | 2 | } else { |
355 | 2 | String::Value value(isolate, str); |
356 | 2 | nbytes = base64_decode(buf, buflen, *value, value.length()); |
357 | 2 | } |
358 | 2 | break; |
359 | | |
360 | 16 | case HEX: |
361 | 16 | if (str->IsExternalOneByte()) { |
362 | 0 | auto ext = str->GetExternalOneByteStringResource(); |
363 | 0 | nbytes = hex_decode(buf, buflen, ext->data(), ext->length()); |
364 | 16 | } else { |
365 | 16 | String::Value value(isolate, str); |
366 | 16 | nbytes = hex_decode(buf, buflen, *value, value.length()); |
367 | 16 | } |
368 | 16 | break; |
369 | | |
370 | 0 | default: |
371 | 0 | UNREACHABLE("unknown encoding"); |
372 | 149k | } |
373 | | |
374 | 149k | return nbytes; |
375 | 149k | } |
376 | | |
377 | | // Quick and dirty size calculation |
378 | | // Will always be at least big enough, but may have some extra |
379 | | // UTF8 can be as much as 3x the size, Base64 can have 1-2 extra bytes |
380 | | Maybe<size_t> StringBytes::StorageSize(Isolate* isolate, |
381 | | Local<Value> val, |
382 | 15.9M | enum encoding encoding) { |
383 | 15.9M | HandleScope scope(isolate); |
384 | 15.9M | size_t data_size = 0; |
385 | 15.9M | bool is_buffer = Buffer::HasInstance(val); |
386 | | |
387 | 15.9M | if (is_buffer && (encoding == BUFFER || encoding == LATIN1)) { |
388 | 0 | return Just(Buffer::Length(val)); |
389 | 0 | } |
390 | | |
391 | 15.9M | Local<String> str; |
392 | 15.9M | if (!val->ToString(isolate->GetCurrentContext()).ToLocal(&str)) |
393 | 0 | return Nothing<size_t>(); |
394 | | |
395 | 15.9M | switch (encoding) { |
396 | 0 | case ASCII: |
397 | 0 | case LATIN1: |
398 | 0 | data_size = str->Length(); |
399 | 0 | break; |
400 | | |
401 | 0 | case BUFFER: |
402 | 15.9M | case UTF8: |
403 | | // A single UCS2 codepoint never takes up more than 3 utf8 bytes. |
404 | | // It is an exercise for the caller to decide when a string is |
405 | | // long enough to justify calling Size() instead of StorageSize() |
406 | 15.9M | data_size = 3 * str->Length(); |
407 | 15.9M | break; |
408 | | |
409 | 0 | case UCS2: |
410 | 0 | data_size = str->Length() * sizeof(uint16_t); |
411 | 0 | break; |
412 | | |
413 | 0 | case BASE64URL: |
414 | | // Fall through |
415 | 0 | case BASE64: |
416 | 0 | data_size = base64_decoded_size_fast(str->Length()); |
417 | 0 | break; |
418 | | |
419 | 3 | case HEX: |
420 | 3 | CHECK(str->Length() % 2 == 0 && "invalid hex string length"); |
421 | 3 | data_size = str->Length() / 2; |
422 | 3 | break; |
423 | | |
424 | 0 | default: |
425 | 0 | UNREACHABLE("unknown encoding"); |
426 | 15.9M | } |
427 | | |
428 | 15.9M | return Just(data_size); |
429 | 15.9M | } |
430 | | |
431 | | Maybe<size_t> StringBytes::Size(Isolate* isolate, |
432 | | Local<Value> val, |
433 | 4.76k | enum encoding encoding) { |
434 | 4.76k | HandleScope scope(isolate); |
435 | | |
436 | 4.76k | if (Buffer::HasInstance(val) && (encoding == BUFFER || encoding == LATIN1)) |
437 | 0 | return Just(Buffer::Length(val)); |
438 | | |
439 | 4.76k | Local<String> str; |
440 | 4.76k | if (!val->ToString(isolate->GetCurrentContext()).ToLocal(&str)) |
441 | 0 | return Nothing<size_t>(); |
442 | | |
443 | 4.76k | switch (encoding) { |
444 | 0 | case ASCII: |
445 | 0 | case LATIN1: |
446 | 0 | return Just<size_t>(str->Length()); |
447 | | |
448 | 0 | case BUFFER: |
449 | 4.76k | case UTF8: |
450 | 4.76k | return Just<size_t>(str->Utf8Length(isolate)); |
451 | | |
452 | 4 | case UCS2: |
453 | 4 | return Just(str->Length() * sizeof(uint16_t)); |
454 | | |
455 | 0 | case BASE64URL: |
456 | | // Fall through |
457 | 0 | case BASE64: { |
458 | 0 | String::Value value(isolate, str); |
459 | 0 | return Just(base64_decoded_size(*value, value.length())); |
460 | 0 | } |
461 | | |
462 | 0 | case HEX: |
463 | 0 | return Just<size_t>(str->Length() / 2); |
464 | 4.76k | } |
465 | | |
466 | 0 | UNREACHABLE(); |
467 | 0 | } |
468 | | |
469 | 0 | static void force_ascii_slow(const char* src, char* dst, size_t len) { |
470 | 0 | for (size_t i = 0; i < len; ++i) { |
471 | 0 | dst[i] = src[i] & 0x7f; |
472 | 0 | } |
473 | 0 | } |
474 | | |
475 | | |
476 | 0 | static void force_ascii(const char* src, char* dst, size_t len) { |
477 | 0 | if (len < 16) { |
478 | 0 | force_ascii_slow(src, dst, len); |
479 | 0 | return; |
480 | 0 | } |
481 | | |
482 | 0 | const unsigned bytes_per_word = sizeof(uintptr_t); |
483 | 0 | const unsigned align_mask = bytes_per_word - 1; |
484 | 0 | const unsigned src_unalign = reinterpret_cast<uintptr_t>(src) & align_mask; |
485 | 0 | const unsigned dst_unalign = reinterpret_cast<uintptr_t>(dst) & align_mask; |
486 | |
|
487 | 0 | if (src_unalign > 0) { |
488 | 0 | if (src_unalign == dst_unalign) { |
489 | 0 | const unsigned unalign = bytes_per_word - src_unalign; |
490 | 0 | force_ascii_slow(src, dst, unalign); |
491 | 0 | src += unalign; |
492 | 0 | dst += unalign; |
493 | 0 | len -= src_unalign; |
494 | 0 | } else { |
495 | 0 | force_ascii_slow(src, dst, len); |
496 | 0 | return; |
497 | 0 | } |
498 | 0 | } |
499 | | |
500 | 0 | #if defined(_WIN64) || defined(_LP64) |
501 | 0 | const uintptr_t mask = ~0x8080808080808080ll; |
502 | | #else |
503 | | const uintptr_t mask = ~0x80808080l; |
504 | | #endif |
505 | |
|
506 | 0 | const uintptr_t* srcw = reinterpret_cast<const uintptr_t*>(src); |
507 | 0 | uintptr_t* dstw = reinterpret_cast<uintptr_t*>(dst); |
508 | |
|
509 | 0 | for (size_t i = 0, n = len / bytes_per_word; i < n; ++i) { |
510 | 0 | dstw[i] = srcw[i] & mask; |
511 | 0 | } |
512 | |
|
513 | 0 | const unsigned remainder = len & align_mask; |
514 | 0 | if (remainder > 0) { |
515 | 0 | const size_t offset = len - remainder; |
516 | 0 | force_ascii_slow(src + offset, dst + offset, remainder); |
517 | 0 | } |
518 | 0 | } |
519 | | |
520 | | |
521 | | size_t StringBytes::hex_encode( |
522 | | const char* src, |
523 | | size_t slen, |
524 | | char* dst, |
525 | 955 | size_t dlen) { |
526 | | // We know how much we'll write, just make sure that there's space. |
527 | 955 | CHECK(dlen >= slen * 2 && |
528 | 955 | "not enough space provided for hex encode"); |
529 | | |
530 | 955 | dlen = slen * 2; |
531 | 20.9k | for (uint32_t i = 0, k = 0; k < dlen; i += 1, k += 2) { |
532 | 20.0k | static const char hex[] = "0123456789abcdef"; |
533 | 20.0k | uint8_t val = static_cast<uint8_t>(src[i]); |
534 | 20.0k | dst[k + 0] = hex[val >> 4]; |
535 | 20.0k | dst[k + 1] = hex[val & 15]; |
536 | 20.0k | } |
537 | | |
538 | 955 | return dlen; |
539 | 955 | } |
540 | | |
541 | 0 | std::string StringBytes::hex_encode(const char* src, size_t slen) { |
542 | 0 | size_t dlen = slen * 2; |
543 | 0 | std::string dst(dlen, '\0'); |
544 | 0 | hex_encode(src, slen, dst.data(), dlen); |
545 | 0 | return dst; |
546 | 0 | } |
547 | | |
548 | | #define CHECK_BUFLEN_IN_RANGE(len) \ |
549 | 1.88k | do { \ |
550 | 1.88k | if ((len) > Buffer::kMaxLength) { \ |
551 | 0 | *error = node::ERR_BUFFER_TOO_LARGE(isolate); \ |
552 | 0 | return MaybeLocal<Value>(); \ |
553 | 0 | } \ |
554 | 1.88k | } while (0) |
555 | | |
556 | | |
557 | | MaybeLocal<Value> StringBytes::Encode(Isolate* isolate, |
558 | | const char* buf, |
559 | | size_t buflen, |
560 | | enum encoding encoding, |
561 | 1.88k | Local<Value>* error) { |
562 | 1.88k | CHECK_BUFLEN_IN_RANGE(buflen); |
563 | | |
564 | 1.88k | if (!buflen && encoding != BUFFER) { |
565 | 0 | return String::Empty(isolate); |
566 | 0 | } |
567 | | |
568 | 1.88k | MaybeLocal<String> val; |
569 | | |
570 | 1.88k | switch (encoding) { |
571 | 3 | case BUFFER: |
572 | 3 | { |
573 | 3 | auto maybe_buf = Buffer::Copy(isolate, buf, buflen); |
574 | 3 | Local<v8::Object> buf; |
575 | 3 | if (!maybe_buf.ToLocal(&buf)) { |
576 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
577 | 0 | } |
578 | 3 | return buf; |
579 | 0 | } |
580 | | |
581 | 4 | case ASCII: |
582 | 4 | if (simdutf::validate_ascii_with_errors(buf, buflen).error) { |
583 | | // The input contains non-ASCII bytes. |
584 | 0 | char* out = node::UncheckedMalloc(buflen); |
585 | 0 | if (out == nullptr) { |
586 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
587 | 0 | return MaybeLocal<Value>(); |
588 | 0 | } |
589 | 0 | force_ascii(buf, out, buflen); |
590 | 0 | return ExternOneByteString::New(isolate, out, buflen, error); |
591 | 4 | } else { |
592 | 4 | return ExternOneByteString::NewFromCopy(isolate, buf, buflen, error); |
593 | 4 | } |
594 | | |
595 | 916 | case UTF8: |
596 | 916 | { |
597 | 916 | val = String::NewFromUtf8(isolate, |
598 | 916 | buf, |
599 | 916 | v8::NewStringType::kNormal, |
600 | 916 | buflen); |
601 | 916 | Local<String> str; |
602 | 916 | if (!val.ToLocal(&str)) { |
603 | 0 | *error = node::ERR_STRING_TOO_LONG(isolate); |
604 | 0 | } |
605 | 916 | return str; |
606 | 4 | } |
607 | | |
608 | 1 | case LATIN1: |
609 | 1 | return ExternOneByteString::NewFromCopy(isolate, buf, buflen, error); |
610 | | |
611 | 1 | case BASE64: { |
612 | 1 | size_t dlen = base64_encoded_size(buflen); |
613 | 1 | char* dst = node::UncheckedMalloc(dlen); |
614 | 1 | if (dst == nullptr) { |
615 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
616 | 0 | return MaybeLocal<Value>(); |
617 | 0 | } |
618 | | |
619 | 1 | size_t written = base64_encode(buf, buflen, dst, dlen); |
620 | 1 | CHECK_EQ(written, dlen); |
621 | | |
622 | 1 | return ExternOneByteString::New(isolate, dst, dlen, error); |
623 | 1 | } |
624 | | |
625 | 0 | case BASE64URL: { |
626 | 0 | size_t dlen = base64_encoded_size(buflen, Base64Mode::URL); |
627 | 0 | char* dst = node::UncheckedMalloc(dlen); |
628 | 0 | if (dst == nullptr) { |
629 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
630 | 0 | return MaybeLocal<Value>(); |
631 | 0 | } |
632 | | |
633 | 0 | size_t written = base64_encode(buf, buflen, dst, dlen, Base64Mode::URL); |
634 | 0 | CHECK_EQ(written, dlen); |
635 | | |
636 | 0 | return ExternOneByteString::New(isolate, dst, dlen, error); |
637 | 0 | } |
638 | | |
639 | 955 | case HEX: { |
640 | 955 | size_t dlen = buflen * 2; |
641 | 955 | char* dst = node::UncheckedMalloc(dlen); |
642 | 955 | if (dst == nullptr) { |
643 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
644 | 0 | return MaybeLocal<Value>(); |
645 | 0 | } |
646 | 955 | size_t written = hex_encode(buf, buflen, dst, dlen); |
647 | 955 | CHECK_EQ(written, dlen); |
648 | | |
649 | 955 | return ExternOneByteString::New(isolate, dst, dlen, error); |
650 | 955 | } |
651 | | |
652 | 0 | case UCS2: { |
653 | 0 | size_t str_len = buflen / 2; |
654 | 0 | if (IsBigEndian()) { |
655 | 0 | uint16_t* dst = node::UncheckedMalloc<uint16_t>(str_len); |
656 | 0 | if (str_len != 0 && dst == nullptr) { |
657 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
658 | 0 | return MaybeLocal<Value>(); |
659 | 0 | } |
660 | 0 | for (size_t i = 0, k = 0; k < str_len; i += 2, k += 1) { |
661 | | // The input is in *little endian*, because that's what Node.js |
662 | | // expects, so the high byte comes after the low byte. |
663 | 0 | const uint8_t hi = static_cast<uint8_t>(buf[i + 1]); |
664 | 0 | const uint8_t lo = static_cast<uint8_t>(buf[i + 0]); |
665 | 0 | dst[k] = static_cast<uint16_t>(hi) << 8 | lo; |
666 | 0 | } |
667 | 0 | return ExternTwoByteString::New(isolate, dst, str_len, error); |
668 | 0 | } |
669 | 0 | if (reinterpret_cast<uintptr_t>(buf) % 2 != 0) { |
670 | | // Unaligned data still means we can't directly pass it to V8. |
671 | 0 | char* dst = node::UncheckedMalloc(buflen); |
672 | 0 | if (dst == nullptr) { |
673 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
674 | 0 | return MaybeLocal<Value>(); |
675 | 0 | } |
676 | 0 | memcpy(dst, buf, buflen); |
677 | 0 | return ExternTwoByteString::New( |
678 | 0 | isolate, reinterpret_cast<uint16_t*>(dst), str_len, error); |
679 | 0 | } |
680 | 0 | return ExternTwoByteString::NewFromCopy( |
681 | 0 | isolate, reinterpret_cast<const uint16_t*>(buf), str_len, error); |
682 | 0 | } |
683 | | |
684 | 0 | default: |
685 | 0 | UNREACHABLE("unknown encoding"); |
686 | 1.88k | } |
687 | 1.88k | } |
688 | | |
689 | | |
690 | | MaybeLocal<Value> StringBytes::Encode(Isolate* isolate, |
691 | | const uint16_t* buf, |
692 | | size_t buflen, |
693 | 0 | Local<Value>* error) { |
694 | 0 | if (buflen == 0) return String::Empty(isolate); |
695 | 0 | CHECK_BUFLEN_IN_RANGE(buflen); |
696 | | |
697 | | // Node's "ucs2" encoding expects LE character data inside a |
698 | | // Buffer, so we need to reorder on BE platforms. See |
699 | | // https://nodejs.org/api/buffer.html regarding Node's "ucs2" |
700 | | // encoding specification |
701 | 0 | if (IsBigEndian()) { |
702 | 0 | uint16_t* dst = node::UncheckedMalloc<uint16_t>(buflen); |
703 | 0 | if (dst == nullptr) { |
704 | 0 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); |
705 | 0 | return MaybeLocal<Value>(); |
706 | 0 | } |
707 | 0 | size_t nbytes = buflen * sizeof(uint16_t); |
708 | 0 | memcpy(dst, buf, nbytes); |
709 | 0 | SwapBytes16(reinterpret_cast<char*>(dst), nbytes); |
710 | 0 | return ExternTwoByteString::New(isolate, dst, buflen, error); |
711 | 0 | } else { |
712 | 0 | return ExternTwoByteString::NewFromCopy(isolate, buf, buflen, error); |
713 | 0 | } |
714 | 0 | } |
715 | | |
716 | | MaybeLocal<Value> StringBytes::Encode(Isolate* isolate, |
717 | | const char* buf, |
718 | | enum encoding encoding, |
719 | 0 | Local<Value>* error) { |
720 | 0 | const size_t len = strlen(buf); |
721 | 0 | return Encode(isolate, buf, len, encoding, error); |
722 | 0 | } |
723 | | |
724 | | } // namespace node |