File: | out/../src/node_url.cc |
Warning: | line 1007, column 11 Value stored to 'special_back_slash' is never read |
Press '?' to see keyboard shortcuts
Keyboard shortcuts:
1 | #include "node_url.h" |
2 | #include "base_object-inl.h" |
3 | #include "node_errors.h" |
4 | #include "node_external_reference.h" |
5 | #include "node_i18n.h" |
6 | #include "util-inl.h" |
7 | |
8 | #include <algorithm> |
9 | #include <cmath> |
10 | #include <cstdio> |
11 | #include <numeric> |
12 | #include <string> |
13 | #include <vector> |
14 | |
15 | namespace node { |
16 | |
17 | using errors::TryCatchScope; |
18 | |
19 | using url::table_data::hex; |
20 | using url::table_data::C0_CONTROL_ENCODE_SET; |
21 | using url::table_data::FRAGMENT_ENCODE_SET; |
22 | using url::table_data::PATH_ENCODE_SET; |
23 | using url::table_data::USERINFO_ENCODE_SET; |
24 | using url::table_data::QUERY_ENCODE_SET_NONSPECIAL; |
25 | using url::table_data::QUERY_ENCODE_SET_SPECIAL; |
26 | |
27 | using v8::Array; |
28 | using v8::Context; |
29 | using v8::Function; |
30 | using v8::FunctionCallbackInfo; |
31 | using v8::HandleScope; |
32 | using v8::Int32; |
33 | using v8::Integer; |
34 | using v8::Isolate; |
35 | using v8::Local; |
36 | using v8::MaybeLocal; |
37 | using v8::NewStringType; |
38 | using v8::Null; |
39 | using v8::Object; |
40 | using v8::String; |
41 | using v8::Undefined; |
42 | using v8::Value; |
43 | |
44 | Local<String> Utf8String(Isolate* isolate, const std::string& str) { |
45 | return String::NewFromUtf8(isolate, |
46 | str.data(), |
47 | NewStringType::kNormal, |
48 | str.length()).ToLocalChecked(); |
49 | } |
50 | |
51 | namespace url { |
52 | namespace { |
53 | |
54 | // https://url.spec.whatwg.org/#eof-code-point |
55 | constexpr char kEOL = -1; |
56 | |
57 | // https://url.spec.whatwg.org/#concept-host |
58 | class URLHost { |
59 | public: |
60 | ~URLHost(); |
61 | |
62 | void ParseIPv4Host(const char* input, size_t length); |
63 | void ParseIPv6Host(const char* input, size_t length); |
64 | void ParseOpaqueHost(const char* input, size_t length); |
65 | void ParseHost(const char* input, |
66 | size_t length, |
67 | bool is_special, |
68 | bool unicode = false); |
69 | |
70 | bool ParsingFailed() const { return type_ == HostType::H_FAILED; } |
71 | std::string ToString() const; |
72 | // Like ToString(), but avoids a copy in exchange for invalidating `*this`. |
73 | std::string ToStringMove(); |
74 | |
75 | private: |
76 | enum class HostType { |
77 | H_FAILED, |
78 | H_DOMAIN, |
79 | H_IPV4, |
80 | H_IPV6, |
81 | H_OPAQUE, |
82 | }; |
83 | |
84 | union Value { |
85 | std::string domain_or_opaque; |
86 | uint32_t ipv4; |
87 | uint16_t ipv6[8]; |
88 | |
89 | ~Value() {} |
90 | Value() : ipv4(0) {} |
91 | }; |
92 | |
93 | Value value_; |
94 | HostType type_ = HostType::H_FAILED; |
95 | |
96 | void Reset() { |
97 | using string = std::string; |
98 | switch (type_) { |
99 | case HostType::H_DOMAIN: |
100 | case HostType::H_OPAQUE: |
101 | value_.domain_or_opaque.~string(); |
102 | break; |
103 | default: |
104 | break; |
105 | } |
106 | type_ = HostType::H_FAILED; |
107 | } |
108 | |
109 | // Setting the string members of the union with = is brittle because |
110 | // it relies on them being initialized to a state that requires no |
111 | // destruction of old data. |
112 | // For a long time, that worked well enough because ParseIPv6Host() happens |
113 | // to zero-fill `value_`, but that really is relying on standard library |
114 | // internals too much. |
115 | // These helpers are the easiest solution but we might want to consider |
116 | // just not forcing strings into an union. |
117 | void SetOpaque(std::string&& string) { |
118 | Reset(); |
119 | type_ = HostType::H_OPAQUE; |
120 | new(&value_.domain_or_opaque) std::string(std::move(string)); |
121 | } |
122 | |
123 | void SetDomain(std::string&& string) { |
124 | Reset(); |
125 | type_ = HostType::H_DOMAIN; |
126 | new(&value_.domain_or_opaque) std::string(std::move(string)); |
127 | } |
128 | }; |
129 | |
130 | URLHost::~URLHost() { |
131 | Reset(); |
132 | } |
133 | |
134 | #define ARGS(XX)XX(ARG_FLAGS) XX(ARG_PROTOCOL) XX(ARG_USERNAME) XX(ARG_PASSWORD ) XX(ARG_HOST) XX(ARG_PORT) XX(ARG_PATH) XX(ARG_QUERY) XX(ARG_FRAGMENT ) XX(ARG_COUNT) \ |
135 | XX(ARG_FLAGS) \ |
136 | XX(ARG_PROTOCOL) \ |
137 | XX(ARG_USERNAME) \ |
138 | XX(ARG_PASSWORD) \ |
139 | XX(ARG_HOST) \ |
140 | XX(ARG_PORT) \ |
141 | XX(ARG_PATH) \ |
142 | XX(ARG_QUERY) \ |
143 | XX(ARG_FRAGMENT) \ |
144 | XX(ARG_COUNT) // This one has to be last. |
145 | |
146 | enum url_cb_args { |
147 | #define XX(name) name, |
148 | ARGS(XX)XX(ARG_FLAGS) XX(ARG_PROTOCOL) XX(ARG_USERNAME) XX(ARG_PASSWORD ) XX(ARG_HOST) XX(ARG_PORT) XX(ARG_PATH) XX(ARG_QUERY) XX(ARG_FRAGMENT ) XX(ARG_COUNT) |
149 | #undef XX |
150 | }; |
151 | |
152 | #define TWO_CHAR_STRING_TEST(bits, name, expr) \ |
153 | template <typename T> \ |
154 | bool name(const T ch1, const T ch2) { \ |
155 | static_assert(sizeof(ch1) >= (bits) / 8, \ |
156 | "Character must be wider than " #bits " bits"); \ |
157 | return (expr); \ |
158 | } \ |
159 | template <typename T> \ |
160 | bool name(const std::basic_string<T>& str) { \ |
161 | static_assert(sizeof(str[0]) >= (bits) / 8, \ |
162 | "Character must be wider than " #bits " bits"); \ |
163 | return str.length() >= 2 && name(str[0], str[1]); \ |
164 | } |
165 | |
166 | // https://infra.spec.whatwg.org/#ascii-tab-or-newline |
167 | CHAR_TEST(8, IsASCIITabOrNewline, (ch == '\t' || ch == '\n' || ch == '\r'))template <typename T> bool IsASCIITabOrNewline(const T ch ) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((ch == '\t' || ch == '\n' || ch == '\r' )); } |
168 | |
169 | // https://infra.spec.whatwg.org/#c0-control |
170 | CHAR_TEST(8, IsC0Control, (ch >= '\0' && ch <= '\x1f'))template <typename T> bool IsC0Control(const T ch) { static_assert (sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((ch >= '\0' && ch <= '\x1f') ); } |
171 | |
172 | // https://infra.spec.whatwg.org/#c0-control-or-space |
173 | CHAR_TEST(8, IsC0ControlOrSpace, (ch >= '\0' && ch <= ' '))template <typename T> bool IsC0ControlOrSpace(const T ch ) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((ch >= '\0' && ch <= ' ' )); } |
174 | |
175 | // https://infra.spec.whatwg.org/#ascii-digit |
176 | CHAR_TEST(8, IsASCIIDigit, (ch >= '0' && ch <= '9'))template <typename T> bool IsASCIIDigit(const T ch) { static_assert (sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((ch >= '0' && ch <= '9')); } |
177 | |
178 | // https://infra.spec.whatwg.org/#ascii-hex-digit |
179 | CHAR_TEST(8, IsASCIIHexDigit, (IsASCIIDigit(ch) ||template <typename T> bool IsASCIIHexDigit(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((IsASCIIDigit(ch) || (ch >= 'A' && ch <= 'F') || (ch >= 'a' && ch <= 'f'))); } |
180 | (ch >= 'A' && ch <= 'F') ||template <typename T> bool IsASCIIHexDigit(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((IsASCIIDigit(ch) || (ch >= 'A' && ch <= 'F') || (ch >= 'a' && ch <= 'f'))); } |
181 | (ch >= 'a' && ch <= 'f')))template <typename T> bool IsASCIIHexDigit(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((IsASCIIDigit(ch) || (ch >= 'A' && ch <= 'F') || (ch >= 'a' && ch <= 'f'))); } |
182 | |
183 | // https://infra.spec.whatwg.org/#ascii-alpha |
184 | CHAR_TEST(8, IsASCIIAlpha, ((ch >= 'A' && ch <= 'Z') ||template <typename T> bool IsASCIIAlpha(const T ch) { static_assert (sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (((ch >= 'A' && ch <= 'Z') || (ch >= 'a' && ch <= 'z'))); } |
185 | (ch >= 'a' && ch <= 'z')))template <typename T> bool IsASCIIAlpha(const T ch) { static_assert (sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (((ch >= 'A' && ch <= 'Z') || (ch >= 'a' && ch <= 'z'))); } |
186 | |
187 | // https://infra.spec.whatwg.org/#ascii-alphanumeric |
188 | CHAR_TEST(8, IsASCIIAlphanumeric, (IsASCIIDigit(ch) || IsASCIIAlpha(ch)))template <typename T> bool IsASCIIAlphanumeric(const T ch ) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return ((IsASCIIDigit(ch) || IsASCIIAlpha(ch)) ); } |
189 | |
190 | // https://infra.spec.whatwg.org/#ascii-lowercase |
191 | template <typename T> |
192 | T ASCIILowercase(T ch) { |
193 | return IsASCIIAlpha(ch) ? (ch | 0x20) : ch; |
194 | } |
195 | |
196 | // https://url.spec.whatwg.org/#forbidden-host-code-point |
197 | CHAR_TEST(8,template <typename T> bool IsForbiddenHostCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (ch == '\0' || ch == '\t' || ch == '\n' || ch == '\r' || ch == ' ' || ch == '#' || ch == '/' || ch == ':' || ch == '?' || ch == '@' || ch == '[' || ch == '<' || ch == '>' || ch == '\\' || ch == ']' || ch == '^' || ch == '|'); } |
198 | IsForbiddenHostCodePoint,template <typename T> bool IsForbiddenHostCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (ch == '\0' || ch == '\t' || ch == '\n' || ch == '\r' || ch == ' ' || ch == '#' || ch == '/' || ch == ':' || ch == '?' || ch == '@' || ch == '[' || ch == '<' || ch == '>' || ch == '\\' || ch == ']' || ch == '^' || ch == '|'); } |
199 | ch == '\0' || ch == '\t' || ch == '\n' || ch == '\r' || ch == ' ' ||template <typename T> bool IsForbiddenHostCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (ch == '\0' || ch == '\t' || ch == '\n' || ch == '\r' || ch == ' ' || ch == '#' || ch == '/' || ch == ':' || ch == '?' || ch == '@' || ch == '[' || ch == '<' || ch == '>' || ch == '\\' || ch == ']' || ch == '^' || ch == '|'); } |
200 | ch == '#' || ch == '/' || ch == ':' || ch == '?' || ch == '@' ||template <typename T> bool IsForbiddenHostCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (ch == '\0' || ch == '\t' || ch == '\n' || ch == '\r' || ch == ' ' || ch == '#' || ch == '/' || ch == ':' || ch == '?' || ch == '@' || ch == '[' || ch == '<' || ch == '>' || ch == '\\' || ch == ']' || ch == '^' || ch == '|'); } |
201 | ch == '[' || ch == '<' || ch == '>' || ch == '\\' || ch == ']' ||template <typename T> bool IsForbiddenHostCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (ch == '\0' || ch == '\t' || ch == '\n' || ch == '\r' || ch == ' ' || ch == '#' || ch == '/' || ch == ':' || ch == '?' || ch == '@' || ch == '[' || ch == '<' || ch == '>' || ch == '\\' || ch == ']' || ch == '^' || ch == '|'); } |
202 | ch == '^' || ch == '|')template <typename T> bool IsForbiddenHostCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (ch == '\0' || ch == '\t' || ch == '\n' || ch == '\r' || ch == ' ' || ch == '#' || ch == '/' || ch == ':' || ch == '?' || ch == '@' || ch == '[' || ch == '<' || ch == '>' || ch == '\\' || ch == ']' || ch == '^' || ch == '|'); } |
203 | |
204 | // https://url.spec.whatwg.org/#forbidden-domain-code-point |
205 | CHAR_TEST(8,template <typename T> bool IsForbiddenDomainCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (IsForbiddenHostCodePoint(ch) || IsC0Control (ch) || ch == '%' || ch == '\x7f'); } |
206 | IsForbiddenDomainCodePoint,template <typename T> bool IsForbiddenDomainCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (IsForbiddenHostCodePoint(ch) || IsC0Control (ch) || ch == '%' || ch == '\x7f'); } |
207 | IsForbiddenHostCodePoint(ch) || IsC0Control(ch) || ch == '%' ||template <typename T> bool IsForbiddenDomainCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (IsForbiddenHostCodePoint(ch) || IsC0Control (ch) || ch == '%' || ch == '\x7f'); } |
208 | ch == '\x7f')template <typename T> bool IsForbiddenDomainCodePoint(const T ch) { static_assert(sizeof(ch) >= (8) / 8, "Character must be wider than " "8" " bits"); return (IsForbiddenHostCodePoint(ch) || IsC0Control (ch) || ch == '%' || ch == '\x7f'); } |
209 | |
210 | // https://url.spec.whatwg.org/#windows-drive-letter |
211 | TWO_CHAR_STRING_TEST(8, IsWindowsDriveLetter, |
212 | (IsASCIIAlpha(ch1) && (ch2 == ':' || ch2 == '|'))) |
213 | |
214 | // https://url.spec.whatwg.org/#normalized-windows-drive-letter |
215 | TWO_CHAR_STRING_TEST(8, IsNormalizedWindowsDriveLetter, |
216 | (IsASCIIAlpha(ch1) && ch2 == ':')) |
217 | |
218 | #undef TWO_CHAR_STRING_TEST |
219 | |
220 | bool BitAt(const uint8_t a[], const uint8_t i) { |
221 | return !!(a[i >> 3] & (1 << (i & 7))); |
222 | } |
223 | |
224 | // Appends ch to str. If ch position in encode_set is set, the ch will |
225 | // be percent-encoded then appended. |
226 | void AppendOrEscape(std::string* str, |
227 | const unsigned char ch, |
228 | const uint8_t encode_set[]) { |
229 | if (BitAt(encode_set, ch)) |
230 | *str += hex + ch * 4; // "%XX\0" has a length of 4 |
231 | else |
232 | *str += ch; |
233 | } |
234 | |
235 | unsigned hex2bin(const char ch) { |
236 | if (ch >= '0' && ch <= '9') |
237 | return ch - '0'; |
238 | if (ch >= 'A' && ch <= 'F') |
239 | return 10 + (ch - 'A'); |
240 | if (ch >= 'a' && ch <= 'f') |
241 | return 10 + (ch - 'a'); |
242 | UNREACHABLE()do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "242", "\"Unreachable code reached\"", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); |
243 | } |
244 | |
245 | std::string PercentDecode(const char* input, size_t len) { |
246 | std::string dest; |
247 | if (len == 0) |
248 | return dest; |
249 | dest.reserve(len); |
250 | const char* pointer = input; |
251 | const char* end = input + len; |
252 | |
253 | while (pointer < end) { |
254 | const char ch = pointer[0]; |
255 | size_t remaining = end - pointer - 1; |
256 | if (ch != '%' || remaining < 2 || |
257 | (ch == '%' && |
258 | (!IsASCIIHexDigit(pointer[1]) || |
259 | !IsASCIIHexDigit(pointer[2])))) { |
260 | dest += ch; |
261 | pointer++; |
262 | continue; |
263 | } else { |
264 | unsigned a = hex2bin(pointer[1]); |
265 | unsigned b = hex2bin(pointer[2]); |
266 | char c = static_cast<char>(a * 16 + b); |
267 | dest += c; |
268 | pointer += 3; |
269 | } |
270 | } |
271 | return dest; |
272 | } |
273 | |
274 | #define SPECIALS(XX)XX(ftp, 21, "ftp:") XX(file, -1, "file:") XX(http, 80, "http:" ) XX(https, 443, "https:") XX(ws, 80, "ws:") XX(wss, 443, "wss:" ) \ |
275 | XX(ftp, 21, "ftp:") \ |
276 | XX(file, -1, "file:") \ |
277 | XX(http, 80, "http:") \ |
278 | XX(https, 443, "https:") \ |
279 | XX(ws, 80, "ws:") \ |
280 | XX(wss, 443, "wss:") |
281 | |
282 | bool IsSpecial(const std::string& scheme) { |
283 | #define V(_, __, name) if (scheme == name) return true; |
284 | SPECIALS(V)V(ftp, 21, "ftp:") V(file, -1, "file:") V(http, 80, "http:") V (https, 443, "https:") V(ws, 80, "ws:") V(wss, 443, "wss:"); |
285 | #undef V |
286 | return false; |
287 | } |
288 | |
289 | Local<String> GetSpecial(Environment* env, const std::string& scheme) { |
290 | #define V(key, _, name) if (scheme == name) \ |
291 | return env->url_special_##key##_string(); |
292 | SPECIALS(V)V(ftp, 21, "ftp:") V(file, -1, "file:") V(http, 80, "http:") V (https, 443, "https:") V(ws, 80, "ws:") V(wss, 443, "wss:") |
293 | #undef V |
294 | UNREACHABLE()do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "294", "\"Unreachable code reached\"", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); |
295 | } |
296 | |
297 | int NormalizePort(const std::string& scheme, int p) { |
298 | #define V(_, port, name) if (scheme == name && p == port) return -1; |
299 | SPECIALS(V)V(ftp, 21, "ftp:") V(file, -1, "file:") V(http, 80, "http:") V (https, 443, "https:") V(ws, 80, "ws:") V(wss, 443, "wss:"); |
300 | #undef V |
301 | return p; |
302 | } |
303 | |
304 | // https://url.spec.whatwg.org/#start-with-a-windows-drive-letter |
305 | bool StartsWithWindowsDriveLetter(const char* p, const char* end) { |
306 | size_t length = end - p; |
307 | return length >= 2 && |
308 | IsWindowsDriveLetter(p[0], p[1]) && |
309 | (length == 2 || |
310 | p[2] == '/' || |
311 | p[2] == '\\' || |
312 | p[2] == '?' || |
313 | p[2] == '#'); |
314 | } |
315 | |
316 | #if defined(NODE_HAVE_I18N_SUPPORT1) |
317 | bool ToUnicode(const std::string& input, std::string* output) { |
318 | MaybeStackBuffer<char> buf; |
319 | if (i18n::ToUnicode(&buf, input.c_str(), input.length()) < 0) |
320 | return false; |
321 | output->assign(*buf, buf.length()); |
322 | return true; |
323 | } |
324 | |
325 | bool ToASCII(const std::string& input, std::string* output) { |
326 | MaybeStackBuffer<char> buf; |
327 | if (i18n::ToASCII(&buf, input.c_str(), input.length()) < 0) |
328 | return false; |
329 | if (buf.length() == 0) |
330 | return false; |
331 | output->assign(*buf, buf.length()); |
332 | return true; |
333 | } |
334 | #else // !defined(NODE_HAVE_I18N_SUPPORT) |
335 | // Intentional non-ops if ICU is not present. |
336 | bool ToUnicode(const std::string& input, std::string* output) { |
337 | *output = input; |
338 | return true; |
339 | } |
340 | |
341 | bool ToASCII(const std::string& input, std::string* output) { |
342 | *output = input; |
343 | return true; |
344 | } |
345 | #endif // !defined(NODE_HAVE_I18N_SUPPORT) |
346 | |
347 | #define NS_IN6ADDRSZ16 16 |
348 | |
349 | void URLHost::ParseIPv6Host(const char* input, size_t length) { |
350 | CHECK_EQ(type_, HostType::H_FAILED)do { if (__builtin_expect(!!(!((type_) == (HostType::H_FAILED ))), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "350", "(type_) == (HostType::H_FAILED)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
351 | |
352 | unsigned char buf[sizeof(struct in6_addr)]; |
353 | MaybeStackBuffer<char> ipv6(length + 1); |
354 | *(*ipv6 + length) = 0; |
355 | memset(buf, 0, sizeof(buf)); |
356 | memcpy(*ipv6, input, sizeof(const char) * length); |
357 | |
358 | int ret = uv_inet_pton(AF_INET610, *ipv6, buf); |
359 | |
360 | if (ret != 0) { |
361 | return; |
362 | } |
363 | |
364 | // Ref: https://sourceware.org/git/?p=glibc.git;a=blob;f=resolv/inet_ntop.c;h=c4d38c0f951013e51a4fc6eaa8a9b82e146abe5a;hb=HEAD#l119 |
365 | for (int i = 0; i < NS_IN6ADDRSZ16; i += 2) { |
366 | value_.ipv6[i >> 1] = (buf[i] << 8) | buf[i + 1]; |
367 | } |
368 | |
369 | type_ = HostType::H_IPV6; |
370 | } |
371 | |
372 | // https://url.spec.whatwg.org/#ipv4-number-parser |
373 | int64_t ParseIPv4Number(const char* start, const char* end) { |
374 | if (end - start == 0) return -1; |
375 | |
376 | unsigned R = 10; |
377 | if (end - start >= 2 && start[0] == '0' && (start[1] | 0x20) == 'x') { |
378 | start += 2; |
379 | R = 16; |
380 | } else if (end - start >= 2 && start[0] == '0') { |
381 | start++; |
382 | R = 8; |
383 | } |
384 | |
385 | if (end - start == 0) return 0; |
386 | |
387 | const char* p = start; |
388 | |
389 | while (p < end) { |
390 | const char ch = p[0]; |
391 | switch (R) { |
392 | case 8: |
393 | if (ch < '0' || ch > '7') |
394 | return -1; |
395 | break; |
396 | case 10: |
397 | if (!IsASCIIDigit(ch)) |
398 | return -1; |
399 | break; |
400 | case 16: |
401 | if (!IsASCIIHexDigit(ch)) |
402 | return -1; |
403 | break; |
404 | } |
405 | p++; |
406 | } |
407 | return strtoll(start, nullptr, R); |
408 | } |
409 | |
410 | // https://url.spec.whatwg.org/#ends-in-a-number-checker |
411 | bool EndsInANumber(const std::string& input) { |
412 | std::vector<std::string> parts = SplitString(input, '.', false); |
413 | |
414 | if (parts.empty()) return false; |
415 | |
416 | if (parts.back().empty()) { |
417 | if (parts.size() == 1) return false; |
418 | parts.pop_back(); |
419 | } |
420 | |
421 | const std::string& last = parts.back(); |
422 | |
423 | // If last is non-empty and contains only ASCII digits, then return true |
424 | if (!last.empty() && std::all_of(last.begin(), last.end(), ::isdigit)) { |
425 | return true; |
426 | } |
427 | |
428 | const char* last_str = last.c_str(); |
429 | int64_t num = ParseIPv4Number(last_str, last_str + last.size()); |
430 | if (num >= 0) return true; |
431 | |
432 | return false; |
433 | } |
434 | |
435 | void URLHost::ParseIPv4Host(const char* input, size_t length) { |
436 | CHECK_EQ(type_, HostType::H_FAILED)do { if (__builtin_expect(!!(!((type_) == (HostType::H_FAILED ))), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "436", "(type_) == (HostType::H_FAILED)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
437 | const char* pointer = input; |
438 | const char* mark = input; |
439 | const char* end = pointer + length; |
440 | int parts = 0; |
441 | uint32_t val = 0; |
442 | uint64_t numbers[4]; |
443 | int tooBigNumbers = 0; |
444 | if (length == 0) |
445 | return; |
446 | |
447 | while (pointer <= end) { |
448 | const char ch = pointer < end ? pointer[0] : kEOL; |
449 | int64_t remaining = end - pointer - 1; |
450 | if (ch == '.' || ch == kEOL) { |
451 | if (++parts > static_cast<int>(arraysize(numbers))) |
452 | return; |
453 | if (pointer == mark) |
454 | return; |
455 | int64_t n = ParseIPv4Number(mark, pointer); |
456 | if (n < 0) |
457 | return; |
458 | |
459 | if (n > 255) { |
460 | tooBigNumbers++; |
461 | } |
462 | numbers[parts - 1] = n; |
463 | mark = pointer + 1; |
464 | if (ch == '.' && remaining == 0) |
465 | break; |
466 | } |
467 | pointer++; |
468 | } |
469 | CHECK_GT(parts, 0)do { if (__builtin_expect(!!(!((parts) > (0))), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "469", "(parts) > (0)", __PRETTY_FUNCTION__ }; node:: Assert(args); } while (0); } } while (0); |
470 | |
471 | // If any but the last item in numbers is greater than 255, return failure. |
472 | // If the last item in numbers is greater than or equal to |
473 | // 256^(5 - the number of items in numbers), return failure. |
474 | if (tooBigNumbers > 1 || |
475 | (tooBigNumbers == 1 && numbers[parts - 1] <= 255) || |
476 | numbers[parts - 1] >= pow(256, static_cast<double>(5 - parts))) { |
477 | return; |
478 | } |
479 | |
480 | type_ = HostType::H_IPV4; |
481 | val = static_cast<uint32_t>(numbers[parts - 1]); |
482 | for (int n = 0; n < parts - 1; n++) { |
483 | double b = 3 - n; |
484 | val += |
485 | static_cast<uint32_t>(numbers[n]) * static_cast<uint32_t>(pow(256, b)); |
486 | } |
487 | |
488 | value_.ipv4 = val; |
489 | } |
490 | |
491 | void URLHost::ParseOpaqueHost(const char* input, size_t length) { |
492 | CHECK_EQ(type_, HostType::H_FAILED)do { if (__builtin_expect(!!(!((type_) == (HostType::H_FAILED ))), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "492", "(type_) == (HostType::H_FAILED)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
493 | std::string output; |
494 | output.reserve(length); |
495 | for (size_t i = 0; i < length; i++) { |
496 | const char ch = input[i]; |
497 | if (IsForbiddenHostCodePoint(ch)) { |
498 | return; |
499 | } else { |
500 | AppendOrEscape(&output, ch, C0_CONTROL_ENCODE_SET); |
501 | } |
502 | } |
503 | |
504 | SetOpaque(std::move(output)); |
505 | } |
506 | |
507 | void URLHost::ParseHost(const char* input, |
508 | size_t length, |
509 | bool is_special, |
510 | bool unicode) { |
511 | CHECK_EQ(type_, HostType::H_FAILED)do { if (__builtin_expect(!!(!((type_) == (HostType::H_FAILED ))), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "511", "(type_) == (HostType::H_FAILED)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
512 | const char* pointer = input; |
513 | |
514 | if (length == 0) |
515 | return; |
516 | |
517 | if (pointer[0] == '[') { |
518 | if (pointer[length - 1] != ']') |
519 | return; |
520 | return ParseIPv6Host(++pointer, length - 2); |
521 | } |
522 | |
523 | if (!is_special) |
524 | return ParseOpaqueHost(input, length); |
525 | |
526 | // First, we have to percent decode |
527 | std::string decoded = PercentDecode(input, length); |
528 | |
529 | // Then we have to punycode toASCII |
530 | if (!ToASCII(decoded, &decoded)) |
531 | return; |
532 | |
533 | // If any of the following characters are still present, we have to fail |
534 | for (size_t n = 0; n < decoded.size(); n++) { |
535 | const char ch = decoded[n]; |
536 | if (IsForbiddenDomainCodePoint(ch)) { |
537 | return; |
538 | } |
539 | } |
540 | |
541 | // If domain ends in a number, then return the result of IPv4 parsing domain |
542 | if (EndsInANumber(decoded)) { |
543 | return ParseIPv4Host(decoded.c_str(), decoded.length()); |
544 | } |
545 | |
546 | // If the unicode flag is set, run the result through punycode ToUnicode |
547 | if (unicode && !ToUnicode(decoded, &decoded)) |
548 | return; |
549 | |
550 | // It's not an IPv4 or IPv6 address, it must be a domain |
551 | SetDomain(std::move(decoded)); |
552 | } |
553 | |
554 | // Locates the longest sequence of 0 segments in an IPv6 address |
555 | // in order to use the :: compression when serializing |
556 | template <typename T> |
557 | T* FindLongestZeroSequence(T* values, size_t len) { |
558 | T* start = values; |
559 | T* end = start + len; |
560 | T* result = nullptr; |
561 | |
562 | T* current = nullptr; |
563 | unsigned counter = 0, longest = 1; |
564 | |
565 | while (start < end) { |
566 | if (*start == 0) { |
567 | if (current == nullptr) |
568 | current = start; |
569 | counter++; |
570 | } else { |
571 | if (counter > longest) { |
572 | longest = counter; |
573 | result = current; |
574 | } |
575 | counter = 0; |
576 | current = nullptr; |
577 | } |
578 | start++; |
579 | } |
580 | if (counter > longest) |
581 | result = current; |
582 | return result; |
583 | } |
584 | |
585 | std::string URLHost::ToStringMove() { |
586 | std::string return_value; |
587 | switch (type_) { |
588 | case HostType::H_DOMAIN: |
589 | case HostType::H_OPAQUE: |
590 | return_value = std::move(value_.domain_or_opaque); |
591 | break; |
592 | default: |
593 | return_value = ToString(); |
594 | break; |
595 | } |
596 | Reset(); |
597 | return return_value; |
598 | } |
599 | |
600 | std::string URLHost::ToString() const { |
601 | std::string dest; |
602 | switch (type_) { |
603 | case HostType::H_DOMAIN: |
604 | case HostType::H_OPAQUE: |
605 | return value_.domain_or_opaque; |
606 | case HostType::H_IPV4: { |
607 | dest.reserve(15); |
608 | uint32_t value = value_.ipv4; |
609 | for (int n = 0; n < 4; n++) { |
610 | dest.insert(0, std::to_string(value % 256)); |
611 | if (n < 3) |
612 | dest.insert(0, 1, '.'); |
613 | value /= 256; |
614 | } |
615 | break; |
616 | } |
617 | case HostType::H_IPV6: { |
618 | dest.reserve(41); |
619 | dest += '['; |
620 | const uint16_t* start = &value_.ipv6[0]; |
621 | const uint16_t* compress_pointer = |
622 | FindLongestZeroSequence(start, 8); |
623 | bool ignore0 = false; |
624 | for (int n = 0; n <= 7; n++) { |
625 | const uint16_t* piece = &value_.ipv6[n]; |
626 | if (ignore0 && *piece == 0) |
627 | continue; |
628 | else if (ignore0) |
629 | ignore0 = false; |
630 | if (compress_pointer == piece) { |
631 | dest += n == 0 ? "::" : ":"; |
632 | ignore0 = true; |
633 | continue; |
634 | } |
635 | char buf[5]; |
636 | snprintf(buf, sizeof(buf), "%x", *piece); |
637 | dest += buf; |
638 | if (n < 7) |
639 | dest += ':'; |
640 | } |
641 | dest += ']'; |
642 | break; |
643 | } |
644 | case HostType::H_FAILED: |
645 | break; |
646 | } |
647 | return dest; |
648 | } |
649 | |
650 | bool ParseHost(const std::string& input, |
651 | std::string* output, |
652 | bool is_special, |
653 | bool unicode = false) { |
654 | if (input.empty()) { |
655 | output->clear(); |
656 | return true; |
657 | } |
658 | URLHost host; |
659 | host.ParseHost(input.c_str(), input.length(), is_special, unicode); |
660 | if (host.ParsingFailed()) |
661 | return false; |
662 | *output = host.ToStringMove(); |
663 | return true; |
664 | } |
665 | |
666 | std::vector<std::string> FromJSStringArray(Environment* env, |
667 | Local<Array> array) { |
668 | std::vector<std::string> vec; |
669 | if (array->Length() > 0) |
670 | vec.reserve(array->Length()); |
671 | for (size_t n = 0; n < array->Length(); n++) { |
672 | Local<Value> val = array->Get(env->context(), n).ToLocalChecked(); |
673 | if (val->IsString()) { |
674 | Utf8Value value(env->isolate(), val.As<String>()); |
675 | vec.emplace_back(*value, value.length()); |
676 | } |
677 | } |
678 | return vec; |
679 | } |
680 | |
681 | url_data HarvestBase(Environment* env, Local<Object> base_obj) { |
682 | url_data base; |
683 | Local<Context> context = env->context(); |
684 | |
685 | Local<Value> flags = |
686 | base_obj->Get(env->context(), env->flags_string()).ToLocalChecked(); |
687 | if (flags->IsInt32()) |
688 | base.flags = flags->Int32Value(context).FromJust(); |
689 | |
690 | Local<Value> port = |
691 | base_obj->Get(env->context(), env->port_string()).ToLocalChecked(); |
692 | if (port->IsInt32()) |
693 | base.port = port->Int32Value(context).FromJust(); |
694 | |
695 | Local<Value> scheme = |
696 | base_obj->Get(env->context(), env->scheme_string()).ToLocalChecked(); |
697 | base.scheme = Utf8Value(env->isolate(), scheme).out(); |
698 | |
699 | auto GetStr = [&](std::string url_data::*member, |
700 | int flag, |
701 | Local<String> name, |
702 | bool empty_as_present) { |
703 | Local<Value> value = base_obj->Get(env->context(), name).ToLocalChecked(); |
704 | if (value->IsString()) { |
705 | Utf8Value utf8value(env->isolate(), value.As<String>()); |
706 | (base.*member).assign(*utf8value, utf8value.length()); |
707 | if (empty_as_present || value.As<String>()->Length() != 0) { |
708 | base.flags |= flag; |
709 | } |
710 | } |
711 | }; |
712 | GetStr(&url_data::username, |
713 | URL_FLAGS_HAS_USERNAME, |
714 | env->username_string(), |
715 | false); |
716 | GetStr(&url_data::password, |
717 | URL_FLAGS_HAS_PASSWORD, |
718 | env->password_string(), |
719 | false); |
720 | GetStr(&url_data::host, URL_FLAGS_HAS_HOST, env->host_string(), true); |
721 | GetStr(&url_data::query, URL_FLAGS_HAS_QUERY, env->query_string(), true); |
722 | GetStr(&url_data::fragment, |
723 | URL_FLAGS_HAS_FRAGMENT, |
724 | env->fragment_string(), |
725 | true); |
726 | |
727 | Local<Value> |
728 | path = base_obj->Get(env->context(), env->path_string()).ToLocalChecked(); |
729 | if (path->IsArray()) { |
730 | base.flags |= URL_FLAGS_HAS_PATH; |
731 | base.path = FromJSStringArray(env, path.As<Array>()); |
732 | } |
733 | return base; |
734 | } |
735 | |
736 | url_data HarvestContext(Environment* env, Local<Object> context_obj) { |
737 | url_data context; |
738 | Local<Value> flags = |
739 | context_obj->Get(env->context(), env->flags_string()).ToLocalChecked(); |
740 | if (flags->IsInt32()) { |
741 | static constexpr int32_t kCopyFlagsMask = |
742 | URL_FLAGS_SPECIAL | |
743 | URL_FLAGS_CANNOT_BE_BASE | |
744 | URL_FLAGS_HAS_USERNAME | |
745 | URL_FLAGS_HAS_PASSWORD | |
746 | URL_FLAGS_HAS_HOST; |
747 | context.flags |= flags.As<Int32>()->Value() & kCopyFlagsMask; |
748 | } |
749 | Local<Value> scheme = |
750 | context_obj->Get(env->context(), env->scheme_string()).ToLocalChecked(); |
751 | if (scheme->IsString()) { |
752 | Utf8Value value(env->isolate(), scheme); |
753 | context.scheme.assign(*value, value.length()); |
754 | } |
755 | Local<Value> port = |
756 | context_obj->Get(env->context(), env->port_string()).ToLocalChecked(); |
757 | if (port->IsInt32()) |
758 | context.port = port.As<Int32>()->Value(); |
759 | if (context.flags & URL_FLAGS_HAS_USERNAME) { |
760 | Local<Value> username = |
761 | context_obj->Get(env->context(), |
762 | env->username_string()).ToLocalChecked(); |
763 | CHECK(username->IsString())do { if (__builtin_expect(!!(!(username->IsString())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "763", "username->IsString()", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
764 | Utf8Value value(env->isolate(), username); |
765 | context.username.assign(*value, value.length()); |
766 | } |
767 | if (context.flags & URL_FLAGS_HAS_PASSWORD) { |
768 | Local<Value> password = |
769 | context_obj->Get(env->context(), |
770 | env->password_string()).ToLocalChecked(); |
771 | CHECK(password->IsString())do { if (__builtin_expect(!!(!(password->IsString())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "771", "password->IsString()", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
772 | Utf8Value value(env->isolate(), password); |
773 | context.password.assign(*value, value.length()); |
774 | } |
775 | Local<Value> host = |
776 | context_obj->Get(env->context(), |
777 | env->host_string()).ToLocalChecked(); |
778 | if (host->IsString()) { |
779 | Utf8Value value(env->isolate(), host); |
780 | context.host.assign(*value, value.length()); |
781 | } |
782 | return context; |
783 | } |
784 | |
785 | // Single dot segment can be ".", "%2e", or "%2E" |
786 | bool IsSingleDotSegment(const std::string& str) { |
787 | switch (str.size()) { |
788 | case 1: |
789 | return str == "."; |
790 | case 3: |
791 | return str[0] == '%' && |
792 | str[1] == '2' && |
793 | ASCIILowercase(str[2]) == 'e'; |
794 | default: |
795 | return false; |
796 | } |
797 | } |
798 | |
799 | // Double dot segment can be: |
800 | // "..", ".%2e", ".%2E", "%2e.", "%2E.", |
801 | // "%2e%2e", "%2E%2E", "%2e%2E", or "%2E%2e" |
802 | bool IsDoubleDotSegment(const std::string& str) { |
803 | switch (str.size()) { |
804 | case 2: |
805 | return str == ".."; |
806 | case 4: |
807 | if (str[0] != '.' && str[0] != '%') |
808 | return false; |
809 | return ((str[0] == '.' && |
810 | str[1] == '%' && |
811 | str[2] == '2' && |
812 | ASCIILowercase(str[3]) == 'e') || |
813 | (str[0] == '%' && |
814 | str[1] == '2' && |
815 | ASCIILowercase(str[2]) == 'e' && |
816 | str[3] == '.')); |
817 | case 6: |
818 | return (str[0] == '%' && |
819 | str[1] == '2' && |
820 | ASCIILowercase(str[2]) == 'e' && |
821 | str[3] == '%' && |
822 | str[4] == '2' && |
823 | ASCIILowercase(str[5]) == 'e'); |
824 | default: |
825 | return false; |
826 | } |
827 | } |
828 | |
829 | void ShortenUrlPath(struct url_data* url) { |
830 | if (url->path.empty()) return; |
831 | if (url->path.size() == 1 && url->scheme == "file:" && |
832 | IsNormalizedWindowsDriveLetter(url->path[0])) return; |
833 | url->path.pop_back(); |
834 | } |
835 | |
836 | } // anonymous namespace |
837 | |
838 | void URL::Parse(const char* input, |
839 | size_t len, |
840 | enum url_parse_state state_override, |
841 | struct url_data* url, |
842 | bool has_url, |
843 | const struct url_data* base, |
844 | bool has_base) { |
845 | const char* p = input; |
846 | const char* end = input + len; |
847 | |
848 | if (!has_url) { |
849 | for (const char* ptr = p; ptr < end; ptr++) { |
850 | if (IsC0ControlOrSpace(*ptr)) |
851 | p++; |
852 | else |
853 | break; |
854 | } |
855 | for (const char* ptr = end - 1; ptr >= p; ptr--) { |
856 | if (IsC0ControlOrSpace(*ptr)) |
857 | end--; |
858 | else |
859 | break; |
860 | } |
861 | input = p; |
862 | len = end - p; |
863 | } |
864 | |
865 | // The spec says we should strip out any ASCII tabs or newlines. |
866 | // In those cases, we create another std::string instance with the filtered |
867 | // contents, but in the general case we avoid the overhead. |
868 | std::string whitespace_stripped; |
869 | for (const char* ptr = p; ptr < end; ptr++) { |
870 | if (!IsASCIITabOrNewline(*ptr)) |
871 | continue; |
872 | // Hit tab or newline. Allocate storage, copy what we have until now, |
873 | // and then iterate and filter all similar characters out. |
874 | whitespace_stripped.reserve(len - 1); |
875 | whitespace_stripped.assign(p, ptr - p); |
876 | // 'ptr + 1' skips the current char, which we know to be tab or newline. |
877 | for (ptr = ptr + 1; ptr < end; ptr++) { |
878 | if (!IsASCIITabOrNewline(*ptr)) |
879 | whitespace_stripped += *ptr; |
880 | } |
881 | |
882 | // Update variables like they should have looked like if the string |
883 | // had been stripped of whitespace to begin with. |
884 | input = whitespace_stripped.c_str(); |
885 | len = whitespace_stripped.size(); |
886 | p = input; |
887 | end = input + len; |
888 | break; |
889 | } |
890 | |
891 | bool atflag = false; // Set when @ has been seen. |
892 | bool square_bracket_flag = false; // Set inside of [...] |
893 | bool password_token_seen_flag = false; // Set after a : after an username. |
894 | |
895 | std::string buffer; |
896 | |
897 | // Set the initial parse state. |
898 | const bool has_state_override = state_override != kUnknownState; |
899 | enum url_parse_state state = has_state_override ? state_override : |
900 | kSchemeStart; |
901 | |
902 | if (state < kSchemeStart || state > kFragment) { |
903 | url->flags |= URL_FLAGS_INVALID_PARSE_STATE; |
904 | return; |
905 | } |
906 | |
907 | while (p <= end) { |
908 | const char ch = p < end ? p[0] : kEOL; |
909 | bool special = (url->flags & URL_FLAGS_SPECIAL); |
910 | bool cannot_be_base; |
911 | bool special_back_slash = (special && ch == '\\'); |
912 | |
913 | switch (state) { |
914 | case kSchemeStart: |
915 | if (IsASCIIAlpha(ch)) { |
916 | buffer += ASCIILowercase(ch); |
917 | state = kScheme; |
918 | } else if (!has_state_override) { |
919 | state = kNoScheme; |
920 | continue; |
921 | } else { |
922 | url->flags |= URL_FLAGS_FAILED; |
923 | return; |
924 | } |
925 | break; |
926 | case kScheme: |
927 | if (IsASCIIAlphanumeric(ch) || ch == '+' || ch == '-' || ch == '.') { |
928 | buffer += ASCIILowercase(ch); |
929 | } else if (ch == ':' || (has_state_override && ch == kEOL)) { |
930 | if (has_state_override && buffer.size() == 0) { |
931 | url->flags |= URL_FLAGS_TERMINATED; |
932 | return; |
933 | } |
934 | buffer += ':'; |
935 | |
936 | bool new_is_special = IsSpecial(buffer); |
937 | |
938 | if (has_state_override) { |
939 | if ((special != new_is_special) || |
940 | ((buffer == "file:") && |
941 | ((url->flags & URL_FLAGS_HAS_USERNAME) || |
942 | (url->flags & URL_FLAGS_HAS_PASSWORD) || |
943 | (url->port != -1))) || |
944 | (url->scheme == "file:" && url->host.empty())) { |
945 | url->flags |= URL_FLAGS_TERMINATED; |
946 | return; |
947 | } |
948 | } |
949 | |
950 | url->scheme = std::move(buffer); |
951 | url->port = NormalizePort(url->scheme, url->port); |
952 | if (new_is_special) { |
953 | url->flags |= URL_FLAGS_SPECIAL; |
954 | special = true; |
955 | } else { |
956 | url->flags &= ~URL_FLAGS_SPECIAL; |
957 | special = false; |
958 | } |
959 | // `special_back_slash` equals to `(special && ch == '\\')` and `ch` |
960 | // here always not equals to `\\`. So `special_back_slash` here always |
961 | // equals to `false`. |
962 | special_back_slash = false; |
963 | buffer.clear(); |
964 | if (has_state_override) |
965 | return; |
966 | if (url->scheme == "file:") { |
967 | state = kFile; |
968 | } else if (special && |
969 | has_base && |
970 | url->scheme == base->scheme) { |
971 | state = kSpecialRelativeOrAuthority; |
972 | } else if (special) { |
973 | state = kSpecialAuthoritySlashes; |
974 | } else if (p + 1 < end && p[1] == '/') { |
975 | state = kPathOrAuthority; |
976 | p++; |
977 | } else { |
978 | url->flags |= URL_FLAGS_CANNOT_BE_BASE; |
979 | url->flags |= URL_FLAGS_HAS_PATH; |
980 | url->path.emplace_back(""); |
981 | state = kCannotBeBase; |
982 | } |
983 | } else if (!has_state_override) { |
984 | buffer.clear(); |
985 | state = kNoScheme; |
986 | p = input; |
987 | continue; |
988 | } else { |
989 | url->flags |= URL_FLAGS_FAILED; |
990 | return; |
991 | } |
992 | break; |
993 | case kNoScheme: |
994 | cannot_be_base = has_base && (base->flags & URL_FLAGS_CANNOT_BE_BASE); |
995 | if (!has_base || (cannot_be_base && ch != '#')) { |
996 | url->flags |= URL_FLAGS_FAILED; |
997 | return; |
998 | } else if (cannot_be_base && ch == '#') { |
999 | url->scheme = base->scheme; |
1000 | if (IsSpecial(url->scheme)) { |
1001 | url->flags |= URL_FLAGS_SPECIAL; |
1002 | special = true; |
1003 | } else { |
1004 | url->flags &= ~URL_FLAGS_SPECIAL; |
1005 | special = false; |
1006 | } |
1007 | special_back_slash = (special && ch == '\\'); |
Value stored to 'special_back_slash' is never read | |
1008 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1009 | url->flags |= URL_FLAGS_HAS_PATH; |
1010 | url->path = base->path; |
1011 | } |
1012 | if (base->flags & URL_FLAGS_HAS_QUERY) { |
1013 | url->flags |= URL_FLAGS_HAS_QUERY; |
1014 | url->query = base->query; |
1015 | } |
1016 | if (base->flags & URL_FLAGS_HAS_FRAGMENT) { |
1017 | url->flags |= URL_FLAGS_HAS_FRAGMENT; |
1018 | url->fragment = base->fragment; |
1019 | } |
1020 | url->flags |= URL_FLAGS_CANNOT_BE_BASE; |
1021 | state = kFragment; |
1022 | } else if (has_base && |
1023 | base->scheme != "file:") { |
1024 | state = kRelative; |
1025 | continue; |
1026 | } else { |
1027 | url->scheme = "file:"; |
1028 | url->flags |= URL_FLAGS_SPECIAL; |
1029 | special = true; |
1030 | state = kFile; |
1031 | special_back_slash = (special && ch == '\\'); |
1032 | continue; |
1033 | } |
1034 | break; |
1035 | case kSpecialRelativeOrAuthority: |
1036 | if (ch == '/' && p + 1 < end && p[1] == '/') { |
1037 | state = kSpecialAuthorityIgnoreSlashes; |
1038 | p++; |
1039 | } else { |
1040 | state = kRelative; |
1041 | continue; |
1042 | } |
1043 | break; |
1044 | case kPathOrAuthority: |
1045 | if (ch == '/') { |
1046 | state = kAuthority; |
1047 | } else { |
1048 | state = kPath; |
1049 | continue; |
1050 | } |
1051 | break; |
1052 | case kRelative: |
1053 | url->scheme = base->scheme; |
1054 | if (IsSpecial(url->scheme)) { |
1055 | url->flags |= URL_FLAGS_SPECIAL; |
1056 | special = true; |
1057 | } else { |
1058 | url->flags &= ~URL_FLAGS_SPECIAL; |
1059 | special = false; |
1060 | } |
1061 | special_back_slash = (special && ch == '\\'); |
1062 | switch (ch) { |
1063 | case kEOL: |
1064 | if (base->flags & URL_FLAGS_HAS_USERNAME) { |
1065 | url->flags |= URL_FLAGS_HAS_USERNAME; |
1066 | url->username = base->username; |
1067 | } |
1068 | if (base->flags & URL_FLAGS_HAS_PASSWORD) { |
1069 | url->flags |= URL_FLAGS_HAS_PASSWORD; |
1070 | url->password = base->password; |
1071 | } |
1072 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1073 | url->flags |= URL_FLAGS_HAS_HOST; |
1074 | url->host = base->host; |
1075 | } |
1076 | if (base->flags & URL_FLAGS_HAS_QUERY) { |
1077 | url->flags |= URL_FLAGS_HAS_QUERY; |
1078 | url->query = base->query; |
1079 | } |
1080 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1081 | url->flags |= URL_FLAGS_HAS_PATH; |
1082 | url->path = base->path; |
1083 | } |
1084 | url->port = base->port; |
1085 | break; |
1086 | case '/': |
1087 | state = kRelativeSlash; |
1088 | break; |
1089 | case '?': |
1090 | if (base->flags & URL_FLAGS_HAS_USERNAME) { |
1091 | url->flags |= URL_FLAGS_HAS_USERNAME; |
1092 | url->username = base->username; |
1093 | } |
1094 | if (base->flags & URL_FLAGS_HAS_PASSWORD) { |
1095 | url->flags |= URL_FLAGS_HAS_PASSWORD; |
1096 | url->password = base->password; |
1097 | } |
1098 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1099 | url->flags |= URL_FLAGS_HAS_HOST; |
1100 | url->host = base->host; |
1101 | } |
1102 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1103 | url->flags |= URL_FLAGS_HAS_PATH; |
1104 | url->path = base->path; |
1105 | } |
1106 | url->port = base->port; |
1107 | state = kQuery; |
1108 | break; |
1109 | case '#': |
1110 | if (base->flags & URL_FLAGS_HAS_USERNAME) { |
1111 | url->flags |= URL_FLAGS_HAS_USERNAME; |
1112 | url->username = base->username; |
1113 | } |
1114 | if (base->flags & URL_FLAGS_HAS_PASSWORD) { |
1115 | url->flags |= URL_FLAGS_HAS_PASSWORD; |
1116 | url->password = base->password; |
1117 | } |
1118 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1119 | url->flags |= URL_FLAGS_HAS_HOST; |
1120 | url->host = base->host; |
1121 | } |
1122 | if (base->flags & URL_FLAGS_HAS_QUERY) { |
1123 | url->flags |= URL_FLAGS_HAS_QUERY; |
1124 | url->query = base->query; |
1125 | } |
1126 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1127 | url->flags |= URL_FLAGS_HAS_PATH; |
1128 | url->path = base->path; |
1129 | } |
1130 | url->port = base->port; |
1131 | state = kFragment; |
1132 | break; |
1133 | default: |
1134 | if (special_back_slash) { |
1135 | state = kRelativeSlash; |
1136 | } else { |
1137 | if (base->flags & URL_FLAGS_HAS_USERNAME) { |
1138 | url->flags |= URL_FLAGS_HAS_USERNAME; |
1139 | url->username = base->username; |
1140 | } |
1141 | if (base->flags & URL_FLAGS_HAS_PASSWORD) { |
1142 | url->flags |= URL_FLAGS_HAS_PASSWORD; |
1143 | url->password = base->password; |
1144 | } |
1145 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1146 | url->flags |= URL_FLAGS_HAS_HOST; |
1147 | url->host = base->host; |
1148 | } |
1149 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1150 | url->flags |= URL_FLAGS_HAS_PATH; |
1151 | url->path = base->path; |
1152 | ShortenUrlPath(url); |
1153 | } |
1154 | url->port = base->port; |
1155 | state = kPath; |
1156 | continue; |
1157 | } |
1158 | } |
1159 | break; |
1160 | case kRelativeSlash: |
1161 | if (IsSpecial(url->scheme) && (ch == '/' || ch == '\\')) { |
1162 | state = kSpecialAuthorityIgnoreSlashes; |
1163 | } else if (ch == '/') { |
1164 | state = kAuthority; |
1165 | } else { |
1166 | if (base->flags & URL_FLAGS_HAS_USERNAME) { |
1167 | url->flags |= URL_FLAGS_HAS_USERNAME; |
1168 | url->username = base->username; |
1169 | } |
1170 | if (base->flags & URL_FLAGS_HAS_PASSWORD) { |
1171 | url->flags |= URL_FLAGS_HAS_PASSWORD; |
1172 | url->password = base->password; |
1173 | } |
1174 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1175 | url->flags |= URL_FLAGS_HAS_HOST; |
1176 | url->host = base->host; |
1177 | } |
1178 | url->port = base->port; |
1179 | state = kPath; |
1180 | continue; |
1181 | } |
1182 | break; |
1183 | case kSpecialAuthoritySlashes: |
1184 | state = kSpecialAuthorityIgnoreSlashes; |
1185 | if (ch == '/' && p + 1 < end && p[1] == '/') { |
1186 | p++; |
1187 | } else { |
1188 | continue; |
1189 | } |
1190 | break; |
1191 | case kSpecialAuthorityIgnoreSlashes: |
1192 | if (ch != '/' && ch != '\\') { |
1193 | state = kAuthority; |
1194 | continue; |
1195 | } |
1196 | break; |
1197 | case kAuthority: |
1198 | if (ch == '@') { |
1199 | if (atflag) { |
1200 | buffer.reserve(buffer.size() + 3); |
1201 | buffer.insert(0, "%40"); |
1202 | } |
1203 | atflag = true; |
1204 | size_t blen = buffer.size(); |
1205 | if (blen > 0 && buffer[0] != ':') { |
1206 | url->flags |= URL_FLAGS_HAS_USERNAME; |
1207 | } |
1208 | for (size_t n = 0; n < blen; n++) { |
1209 | const char bch = buffer[n]; |
1210 | if (bch == ':') { |
1211 | url->flags |= URL_FLAGS_HAS_PASSWORD; |
1212 | if (!password_token_seen_flag) { |
1213 | password_token_seen_flag = true; |
1214 | continue; |
1215 | } |
1216 | } |
1217 | if (password_token_seen_flag) { |
1218 | AppendOrEscape(&url->password, bch, USERINFO_ENCODE_SET); |
1219 | } else { |
1220 | AppendOrEscape(&url->username, bch, USERINFO_ENCODE_SET); |
1221 | } |
1222 | } |
1223 | buffer.clear(); |
1224 | } else if (ch == kEOL || |
1225 | ch == '/' || |
1226 | ch == '?' || |
1227 | ch == '#' || |
1228 | special_back_slash) { |
1229 | if (atflag && buffer.size() == 0) { |
1230 | url->flags |= URL_FLAGS_FAILED; |
1231 | return; |
1232 | } |
1233 | p -= buffer.size() + 1; |
1234 | buffer.clear(); |
1235 | state = kHost; |
1236 | } else { |
1237 | buffer += ch; |
1238 | } |
1239 | break; |
1240 | case kHost: |
1241 | case kHostname: |
1242 | if (has_state_override && url->scheme == "file:") { |
1243 | state = kFileHost; |
1244 | continue; |
1245 | } else if (ch == ':' && !square_bracket_flag) { |
1246 | if (buffer.size() == 0) { |
1247 | url->flags |= URL_FLAGS_FAILED; |
1248 | return; |
1249 | } |
1250 | if (state_override == kHostname) { |
1251 | return; |
1252 | } |
1253 | url->flags |= URL_FLAGS_HAS_HOST; |
1254 | if (!ParseHost(buffer, &url->host, special)) { |
1255 | url->flags |= URL_FLAGS_FAILED; |
1256 | return; |
1257 | } |
1258 | buffer.clear(); |
1259 | state = kPort; |
1260 | } else if (ch == kEOL || |
1261 | ch == '/' || |
1262 | ch == '?' || |
1263 | ch == '#' || |
1264 | special_back_slash) { |
1265 | p--; |
1266 | if (special && buffer.size() == 0) { |
1267 | url->flags |= URL_FLAGS_FAILED; |
1268 | return; |
1269 | } |
1270 | if (has_state_override && |
1271 | buffer.size() == 0 && |
1272 | ((url->username.size() > 0 || url->password.size() > 0) || |
1273 | url->port != -1)) { |
1274 | url->flags |= URL_FLAGS_TERMINATED; |
1275 | return; |
1276 | } |
1277 | url->flags |= URL_FLAGS_HAS_HOST; |
1278 | if (!ParseHost(buffer, &url->host, special)) { |
1279 | url->flags |= URL_FLAGS_FAILED; |
1280 | return; |
1281 | } |
1282 | buffer.clear(); |
1283 | state = kPathStart; |
1284 | if (has_state_override) { |
1285 | return; |
1286 | } |
1287 | } else { |
1288 | if (ch == '[') |
1289 | square_bracket_flag = true; |
1290 | if (ch == ']') |
1291 | square_bracket_flag = false; |
1292 | buffer += ch; |
1293 | } |
1294 | break; |
1295 | case kPort: |
1296 | if (IsASCIIDigit(ch)) { |
1297 | buffer += ch; |
1298 | } else if (has_state_override || |
1299 | ch == kEOL || |
1300 | ch == '/' || |
1301 | ch == '?' || |
1302 | ch == '#' || |
1303 | special_back_slash) { |
1304 | if (buffer.size() > 0) { |
1305 | unsigned port = 0; |
1306 | // the condition port <= 0xffff prevents integer overflow |
1307 | for (size_t i = 0; port <= 0xffff && i < buffer.size(); i++) |
1308 | port = port * 10 + buffer[i] - '0'; |
1309 | if (port > 0xffff) { |
1310 | // TODO(TimothyGu): This hack is currently needed for the host |
1311 | // setter since it needs access to hostname if it is valid, and |
1312 | // if the FAILED flag is set the entire response to JS layer |
1313 | // will be empty. |
1314 | if (state_override == kHost) |
1315 | url->port = -1; |
1316 | else |
1317 | url->flags |= URL_FLAGS_FAILED; |
1318 | return; |
1319 | } |
1320 | // the port is valid |
1321 | url->port = NormalizePort(url->scheme, static_cast<int>(port)); |
1322 | if (url->port == -1) |
1323 | url->flags |= URL_FLAGS_IS_DEFAULT_SCHEME_PORT; |
1324 | buffer.clear(); |
1325 | } else if (has_state_override) { |
1326 | // TODO(TimothyGu): Similar case as above. |
1327 | if (state_override == kHost) |
1328 | url->port = -1; |
1329 | else |
1330 | url->flags |= URL_FLAGS_TERMINATED; |
1331 | return; |
1332 | } |
1333 | state = kPathStart; |
1334 | continue; |
1335 | } else { |
1336 | url->flags |= URL_FLAGS_FAILED; |
1337 | return; |
1338 | } |
1339 | break; |
1340 | case kFile: |
1341 | url->scheme = "file:"; |
1342 | url->host.clear(); |
1343 | url->flags |= URL_FLAGS_HAS_HOST; |
1344 | if (ch == '/' || ch == '\\') { |
1345 | state = kFileSlash; |
1346 | } else if (has_base && base->scheme == "file:") { |
1347 | switch (ch) { |
1348 | case kEOL: |
1349 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1350 | url->host = base->host; |
1351 | } |
1352 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1353 | url->flags |= URL_FLAGS_HAS_PATH; |
1354 | url->path = base->path; |
1355 | } |
1356 | if (base->flags & URL_FLAGS_HAS_QUERY) { |
1357 | url->flags |= URL_FLAGS_HAS_QUERY; |
1358 | url->query = base->query; |
1359 | } |
1360 | break; |
1361 | case '?': |
1362 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1363 | url->host = base->host; |
1364 | } |
1365 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1366 | url->flags |= URL_FLAGS_HAS_PATH; |
1367 | url->path = base->path; |
1368 | } |
1369 | url->flags |= URL_FLAGS_HAS_QUERY; |
1370 | url->query.clear(); |
1371 | state = kQuery; |
1372 | break; |
1373 | case '#': |
1374 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1375 | url->host = base->host; |
1376 | } |
1377 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1378 | url->flags |= URL_FLAGS_HAS_PATH; |
1379 | url->path = base->path; |
1380 | } |
1381 | if (base->flags & URL_FLAGS_HAS_QUERY) { |
1382 | url->flags |= URL_FLAGS_HAS_QUERY; |
1383 | url->query = base->query; |
1384 | } |
1385 | url->flags |= URL_FLAGS_HAS_FRAGMENT; |
1386 | url->fragment.clear(); |
1387 | state = kFragment; |
1388 | break; |
1389 | default: |
1390 | url->query.clear(); |
1391 | if (base->flags & URL_FLAGS_HAS_HOST) { |
1392 | url->host = base->host; |
1393 | } |
1394 | if (base->flags & URL_FLAGS_HAS_PATH) { |
1395 | url->flags |= URL_FLAGS_HAS_PATH; |
1396 | url->path = base->path; |
1397 | } |
1398 | if (!StartsWithWindowsDriveLetter(p, end)) { |
1399 | ShortenUrlPath(url); |
1400 | } else { |
1401 | url->path.clear(); |
1402 | } |
1403 | state = kPath; |
1404 | continue; |
1405 | } |
1406 | } else { |
1407 | state = kPath; |
1408 | continue; |
1409 | } |
1410 | break; |
1411 | case kFileSlash: |
1412 | if (ch == '/' || ch == '\\') { |
1413 | state = kFileHost; |
1414 | } else { |
1415 | if (has_base && base->scheme == "file:") { |
1416 | url->flags |= URL_FLAGS_HAS_HOST; |
1417 | url->host = base->host; |
1418 | if (!StartsWithWindowsDriveLetter(p, end) && |
1419 | IsNormalizedWindowsDriveLetter(base->path[0])) { |
1420 | url->flags |= URL_FLAGS_HAS_PATH; |
1421 | url->path.push_back(base->path[0]); |
1422 | } |
1423 | } |
1424 | state = kPath; |
1425 | continue; |
1426 | } |
1427 | break; |
1428 | case kFileHost: |
1429 | if (ch == kEOL || |
1430 | ch == '/' || |
1431 | ch == '\\' || |
1432 | ch == '?' || |
1433 | ch == '#') { |
1434 | if (!has_state_override && |
1435 | buffer.size() == 2 && |
1436 | IsWindowsDriveLetter(buffer)) { |
1437 | state = kPath; |
1438 | } else if (buffer.size() == 0) { |
1439 | url->flags |= URL_FLAGS_HAS_HOST; |
1440 | url->host.clear(); |
1441 | if (has_state_override) |
1442 | return; |
1443 | state = kPathStart; |
1444 | } else { |
1445 | std::string host; |
1446 | if (!ParseHost(buffer, &host, special)) { |
1447 | url->flags |= URL_FLAGS_FAILED; |
1448 | return; |
1449 | } |
1450 | if (host == "localhost") |
1451 | host.clear(); |
1452 | url->flags |= URL_FLAGS_HAS_HOST; |
1453 | url->host = host; |
1454 | if (has_state_override) |
1455 | return; |
1456 | buffer.clear(); |
1457 | state = kPathStart; |
1458 | } |
1459 | continue; |
1460 | } else { |
1461 | buffer += ch; |
1462 | } |
1463 | break; |
1464 | case kPathStart: |
1465 | if (IsSpecial(url->scheme)) { |
1466 | state = kPath; |
1467 | if (ch != '/' && ch != '\\') { |
1468 | continue; |
1469 | } |
1470 | } else if (!has_state_override && ch == '?') { |
1471 | url->flags |= URL_FLAGS_HAS_QUERY; |
1472 | url->query.clear(); |
1473 | state = kQuery; |
1474 | } else if (!has_state_override && ch == '#') { |
1475 | url->flags |= URL_FLAGS_HAS_FRAGMENT; |
1476 | url->fragment.clear(); |
1477 | state = kFragment; |
1478 | } else if (ch != kEOL) { |
1479 | state = kPath; |
1480 | if (ch != '/') { |
1481 | continue; |
1482 | } |
1483 | } else if (has_state_override && !(url->flags & URL_FLAGS_HAS_HOST)) { |
1484 | url->flags |= URL_FLAGS_HAS_PATH; |
1485 | url->path.emplace_back(""); |
1486 | } |
1487 | break; |
1488 | case kPath: |
1489 | if (ch == kEOL || |
1490 | ch == '/' || |
1491 | special_back_slash || |
1492 | (!has_state_override && (ch == '?' || ch == '#'))) { |
1493 | if (IsDoubleDotSegment(buffer)) { |
1494 | ShortenUrlPath(url); |
1495 | if (ch != '/' && !special_back_slash) { |
1496 | url->flags |= URL_FLAGS_HAS_PATH; |
1497 | url->path.emplace_back(""); |
1498 | } |
1499 | } else if (IsSingleDotSegment(buffer) && |
1500 | ch != '/' && !special_back_slash) { |
1501 | url->flags |= URL_FLAGS_HAS_PATH; |
1502 | url->path.emplace_back(""); |
1503 | } else if (!IsSingleDotSegment(buffer)) { |
1504 | if (url->scheme == "file:" && |
1505 | url->path.empty() && |
1506 | buffer.size() == 2 && |
1507 | IsWindowsDriveLetter(buffer)) { |
1508 | buffer[1] = ':'; |
1509 | } |
1510 | url->flags |= URL_FLAGS_HAS_PATH; |
1511 | url->path.emplace_back(std::move(buffer)); |
1512 | } |
1513 | buffer.clear(); |
1514 | if (ch == '?') { |
1515 | url->flags |= URL_FLAGS_HAS_QUERY; |
1516 | url->query.clear(); |
1517 | state = kQuery; |
1518 | } else if (ch == '#') { |
1519 | url->flags |= URL_FLAGS_HAS_FRAGMENT; |
1520 | url->fragment.clear(); |
1521 | state = kFragment; |
1522 | } |
1523 | } else { |
1524 | AppendOrEscape(&buffer, ch, PATH_ENCODE_SET); |
1525 | } |
1526 | break; |
1527 | case kCannotBeBase: |
1528 | switch (ch) { |
1529 | case '?': |
1530 | state = kQuery; |
1531 | break; |
1532 | case '#': |
1533 | state = kFragment; |
1534 | break; |
1535 | default: |
1536 | if (url->path.empty()) |
1537 | url->path.emplace_back(""); |
1538 | else if (ch != kEOL) |
1539 | AppendOrEscape(&url->path[0], ch, C0_CONTROL_ENCODE_SET); |
1540 | } |
1541 | break; |
1542 | case kQuery: |
1543 | if (ch == kEOL || (!has_state_override && ch == '#')) { |
1544 | url->flags |= URL_FLAGS_HAS_QUERY; |
1545 | url->query = std::move(buffer); |
1546 | buffer.clear(); |
1547 | if (ch == '#') |
1548 | state = kFragment; |
1549 | } else { |
1550 | AppendOrEscape(&buffer, ch, special ? QUERY_ENCODE_SET_SPECIAL : |
1551 | QUERY_ENCODE_SET_NONSPECIAL); |
1552 | } |
1553 | break; |
1554 | case kFragment: |
1555 | switch (ch) { |
1556 | case kEOL: |
1557 | url->flags |= URL_FLAGS_HAS_FRAGMENT; |
1558 | url->fragment = std::move(buffer); |
1559 | break; |
1560 | default: |
1561 | AppendOrEscape(&buffer, ch, FRAGMENT_ENCODE_SET); |
1562 | } |
1563 | break; |
1564 | default: |
1565 | url->flags |= URL_FLAGS_INVALID_PARSE_STATE; |
1566 | return; |
1567 | } |
1568 | |
1569 | p++; |
1570 | } |
1571 | } // NOLINT(readability/fn_size) |
1572 | |
1573 | // https://url.spec.whatwg.org/#url-serializing |
1574 | std::string URL::SerializeURL(const url_data& url, |
1575 | bool exclude = false) { |
1576 | std::string output; |
1577 | output.reserve( |
1578 | 10 + // We generally insert < 10 separator characters between URL parts |
1579 | url.scheme.size() + |
1580 | url.username.size() + |
1581 | url.password.size() + |
1582 | url.host.size() + |
1583 | url.query.size() + |
1584 | url.fragment.size() + |
1585 | url.href.size() + |
1586 | std::accumulate( |
1587 | url.path.begin(), |
1588 | url.path.end(), |
1589 | 0, |
1590 | [](size_t sum, const auto& str) { return sum + str.size(); })); |
1591 | |
1592 | output += url.scheme; |
1593 | if (url.flags & URL_FLAGS_HAS_HOST) { |
1594 | output += "//"; |
1595 | if (url.flags & URL_FLAGS_HAS_USERNAME || |
1596 | url.flags & URL_FLAGS_HAS_PASSWORD) { |
1597 | if (url.flags & URL_FLAGS_HAS_USERNAME) { |
1598 | output += url.username; |
1599 | } |
1600 | if (url.flags & URL_FLAGS_HAS_PASSWORD) { |
1601 | output += ":" + url.password; |
1602 | } |
1603 | output += "@"; |
1604 | } |
1605 | output += url.host; |
1606 | if (url.port != -1) { |
1607 | output += ":" + std::to_string(url.port); |
1608 | } |
1609 | } |
1610 | if (url.flags & URL_FLAGS_CANNOT_BE_BASE) { |
1611 | output += url.path[0]; |
1612 | } else { |
1613 | if (!(url.flags & URL_FLAGS_HAS_HOST) && |
1614 | url.path.size() > 1 && |
1615 | url.path[0].empty()) { |
1616 | output += "/."; |
1617 | } |
1618 | for (size_t i = 1; i < url.path.size(); i++) { |
1619 | output += "/" + url.path[i]; |
1620 | } |
1621 | } |
1622 | if (url.flags & URL_FLAGS_HAS_QUERY) { |
1623 | output += "?" + url.query; |
1624 | } |
1625 | if (!exclude && (url.flags & URL_FLAGS_HAS_FRAGMENT)) { |
1626 | output += "#" + url.fragment; |
1627 | } |
1628 | output.shrink_to_fit(); |
1629 | return output; |
1630 | } |
1631 | |
1632 | namespace { |
1633 | void SetArgs(Environment* env, |
1634 | Local<Value> argv[ARG_COUNT], |
1635 | const struct url_data& url) { |
1636 | Isolate* isolate = env->isolate(); |
1637 | argv[ARG_FLAGS] = Integer::NewFromUnsigned(isolate, url.flags); |
1638 | argv[ARG_PROTOCOL] = |
1639 | url.flags & URL_FLAGS_SPECIAL ? |
1640 | GetSpecial(env, url.scheme) : |
1641 | OneByteString(isolate, url.scheme.c_str()); |
1642 | if (url.flags & URL_FLAGS_HAS_USERNAME) |
1643 | argv[ARG_USERNAME] = Utf8String(isolate, url.username); |
1644 | if (url.flags & URL_FLAGS_HAS_PASSWORD) |
1645 | argv[ARG_PASSWORD] = Utf8String(isolate, url.password); |
1646 | if (url.flags & URL_FLAGS_HAS_HOST) |
1647 | argv[ARG_HOST] = Utf8String(isolate, url.host); |
1648 | if (url.flags & URL_FLAGS_HAS_QUERY) |
1649 | argv[ARG_QUERY] = Utf8String(isolate, url.query); |
1650 | if (url.flags & URL_FLAGS_HAS_FRAGMENT) |
1651 | argv[ARG_FRAGMENT] = Utf8String(isolate, url.fragment); |
1652 | if (url.port > -1) |
1653 | argv[ARG_PORT] = Integer::New(isolate, url.port); |
1654 | if (url.flags & URL_FLAGS_HAS_PATH) |
1655 | argv[ARG_PATH] = ToV8Value(env->context(), url.path).ToLocalChecked(); |
1656 | } |
1657 | |
1658 | void Parse(Environment* env, |
1659 | Local<Value> recv, |
1660 | const char* input, |
1661 | size_t len, |
1662 | enum url_parse_state state_override, |
1663 | Local<Value> base_obj, |
1664 | Local<Value> context_obj, |
1665 | Local<Function> cb, |
1666 | Local<Value> error_cb) { |
1667 | Isolate* isolate = env->isolate(); |
1668 | Local<Context> context = env->context(); |
1669 | HandleScope handle_scope(isolate); |
1670 | Context::Scope context_scope(context); |
1671 | |
1672 | const bool has_context = context_obj->IsObject(); |
1673 | const bool has_base = base_obj->IsObject(); |
1674 | |
1675 | url_data base; |
1676 | url_data url; |
1677 | if (has_context) |
1678 | url = HarvestContext(env, context_obj.As<Object>()); |
1679 | if (has_base) |
1680 | base = HarvestBase(env, base_obj.As<Object>()); |
1681 | |
1682 | URL::Parse(input, len, state_override, &url, has_context, &base, has_base); |
1683 | if ((url.flags & URL_FLAGS_INVALID_PARSE_STATE) || |
1684 | ((state_override != kUnknownState) && |
1685 | (url.flags & URL_FLAGS_TERMINATED))) |
1686 | return; |
1687 | |
1688 | // Define the return value placeholders |
1689 | const Local<Value> undef = Undefined(isolate); |
1690 | const Local<Value> null = Null(isolate); |
1691 | if (!(url.flags & URL_FLAGS_FAILED)) { |
1692 | Local<Value> argv[] = { |
1693 | undef, |
1694 | undef, |
1695 | undef, |
1696 | undef, |
1697 | null, // host defaults to null |
1698 | null, // port defaults to null |
1699 | undef, |
1700 | null, // query defaults to null |
1701 | null, // fragment defaults to null |
1702 | }; |
1703 | SetArgs(env, argv, url); |
1704 | USE(cb->Call(context, recv, arraysize(argv), argv)); |
1705 | } else if (error_cb->IsFunction()) { |
1706 | Local<Value> flags = Integer::NewFromUnsigned(isolate, url.flags); |
1707 | USE(error_cb.As<Function>()->Call(context, recv, 1, &flags)); |
1708 | } |
1709 | } |
1710 | |
1711 | void Parse(const FunctionCallbackInfo<Value>& args) { |
1712 | Environment* env = Environment::GetCurrent(args); |
1713 | CHECK_GE(args.Length(), 5)do { if (__builtin_expect(!!(!((args.Length()) >= (5))), 0 )) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1713", "(args.Length()) >= (5)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1714 | CHECK(args[0]->IsString())do { if (__builtin_expect(!!(!(args[0]->IsString())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1714", "args[0]->IsString()", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); // input |
1715 | CHECK(args[2]->IsUndefined() || // base contextdo { if (__builtin_expect(!!(!(args[2]->IsUndefined() || args [2]->IsNull() || args[2]->IsObject())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1717" , "args[2]->IsUndefined() || args[2]->IsNull() || args[2]->IsObject()" , __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0) |
1716 | args[2]->IsNull() ||do { if (__builtin_expect(!!(!(args[2]->IsUndefined() || args [2]->IsNull() || args[2]->IsObject())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1717" , "args[2]->IsUndefined() || args[2]->IsNull() || args[2]->IsObject()" , __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0) |
1717 | args[2]->IsObject())do { if (__builtin_expect(!!(!(args[2]->IsUndefined() || args [2]->IsNull() || args[2]->IsObject())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1717" , "args[2]->IsUndefined() || args[2]->IsNull() || args[2]->IsObject()" , __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1718 | CHECK(args[3]->IsUndefined() || // contextdo { if (__builtin_expect(!!(!(args[3]->IsUndefined() || args [3]->IsNull() || args[3]->IsObject())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1720" , "args[3]->IsUndefined() || args[3]->IsNull() || args[3]->IsObject()" , __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0) |
1719 | args[3]->IsNull() ||do { if (__builtin_expect(!!(!(args[3]->IsUndefined() || args [3]->IsNull() || args[3]->IsObject())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1720" , "args[3]->IsUndefined() || args[3]->IsNull() || args[3]->IsObject()" , __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0) |
1720 | args[3]->IsObject())do { if (__builtin_expect(!!(!(args[3]->IsUndefined() || args [3]->IsNull() || args[3]->IsObject())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1720" , "args[3]->IsUndefined() || args[3]->IsNull() || args[3]->IsObject()" , __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1721 | CHECK(args[4]->IsFunction())do { if (__builtin_expect(!!(!(args[4]->IsFunction())), 0) ) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1721", "args[4]->IsFunction()", __PRETTY_FUNCTION__ } ; node::Assert(args); } while (0); } } while (0); // complete callback |
1722 | CHECK(args[5]->IsUndefined() || args[5]->IsFunction())do { if (__builtin_expect(!!(!(args[5]->IsUndefined() || args [5]->IsFunction())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1722", "args[5]->IsUndefined() || args[5]->IsFunction()" , __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); // error callback |
1723 | |
1724 | Utf8Value input(env->isolate(), args[0]); |
1725 | enum url_parse_state state_override = kUnknownState; |
1726 | if (args[1]->IsNumber()) { |
1727 | state_override = static_cast<enum url_parse_state>( |
1728 | args[1]->Uint32Value(env->context()).FromJust()); |
1729 | } |
1730 | |
1731 | Parse(env, args.This(), |
1732 | *input, input.length(), |
1733 | state_override, |
1734 | args[2], |
1735 | args[3], |
1736 | args[4].As<Function>(), |
1737 | args[5]); |
1738 | } |
1739 | |
1740 | void EncodeAuthSet(const FunctionCallbackInfo<Value>& args) { |
1741 | Environment* env = Environment::GetCurrent(args); |
1742 | CHECK_GE(args.Length(), 1)do { if (__builtin_expect(!!(!((args.Length()) >= (1))), 0 )) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1742", "(args.Length()) >= (1)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1743 | CHECK(args[0]->IsString())do { if (__builtin_expect(!!(!(args[0]->IsString())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1743", "args[0]->IsString()", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1744 | Utf8Value value(env->isolate(), args[0]); |
1745 | std::string output; |
1746 | size_t len = value.length(); |
1747 | output.reserve(len); |
1748 | for (size_t n = 0; n < len; n++) { |
1749 | const char ch = (*value)[n]; |
1750 | AppendOrEscape(&output, ch, USERINFO_ENCODE_SET); |
1751 | } |
1752 | args.GetReturnValue().Set( |
1753 | String::NewFromUtf8(env->isolate(), output.c_str()).ToLocalChecked()); |
1754 | } |
1755 | |
1756 | void DomainToASCII(const FunctionCallbackInfo<Value>& args) { |
1757 | Environment* env = Environment::GetCurrent(args); |
1758 | CHECK_GE(args.Length(), 1)do { if (__builtin_expect(!!(!((args.Length()) >= (1))), 0 )) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1758", "(args.Length()) >= (1)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1759 | CHECK(args[0]->IsString())do { if (__builtin_expect(!!(!(args[0]->IsString())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1759", "args[0]->IsString()", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1760 | Utf8Value value(env->isolate(), args[0]); |
1761 | |
1762 | URLHost host; |
1763 | // Assuming the host is used for a special scheme. |
1764 | host.ParseHost(*value, value.length(), true); |
1765 | if (host.ParsingFailed()) { |
1766 | args.GetReturnValue().Set(FIXED_ONE_BYTE_STRING(env->isolate(), "")); |
1767 | return; |
1768 | } |
1769 | std::string out = host.ToStringMove(); |
1770 | args.GetReturnValue().Set( |
1771 | String::NewFromUtf8(env->isolate(), out.c_str()).ToLocalChecked()); |
1772 | } |
1773 | |
1774 | void DomainToUnicode(const FunctionCallbackInfo<Value>& args) { |
1775 | Environment* env = Environment::GetCurrent(args); |
1776 | CHECK_GE(args.Length(), 1)do { if (__builtin_expect(!!(!((args.Length()) >= (1))), 0 )) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1776", "(args.Length()) >= (1)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1777 | CHECK(args[0]->IsString())do { if (__builtin_expect(!!(!(args[0]->IsString())), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1777", "args[0]->IsString()", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1778 | Utf8Value value(env->isolate(), args[0]); |
1779 | |
1780 | URLHost host; |
1781 | // Assuming the host is used for a special scheme. |
1782 | host.ParseHost(*value, value.length(), true, true); |
1783 | if (host.ParsingFailed()) { |
1784 | args.GetReturnValue().Set(FIXED_ONE_BYTE_STRING(env->isolate(), "")); |
1785 | return; |
1786 | } |
1787 | std::string out = host.ToStringMove(); |
1788 | args.GetReturnValue().Set( |
1789 | String::NewFromUtf8(env->isolate(), out.c_str()).ToLocalChecked()); |
1790 | } |
1791 | |
1792 | void SetURLConstructor(const FunctionCallbackInfo<Value>& args) { |
1793 | Environment* env = Environment::GetCurrent(args); |
1794 | CHECK_EQ(args.Length(), 1)do { if (__builtin_expect(!!(!((args.Length()) == (1))), 0)) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1794", "(args.Length()) == (1)", __PRETTY_FUNCTION__ }; node::Assert(args); } while (0); } } while (0); |
1795 | CHECK(args[0]->IsFunction())do { if (__builtin_expect(!!(!(args[0]->IsFunction())), 0) ) { do { static const node::AssertionInfo args = { "../src/node_url.cc" ":" "1795", "args[0]->IsFunction()", __PRETTY_FUNCTION__ } ; node::Assert(args); } while (0); } } while (0); |
1796 | env->set_url_constructor_function(args[0].As<Function>()); |
1797 | } |
1798 | |
1799 | void Initialize(Local<Object> target, |
1800 | Local<Value> unused, |
1801 | Local<Context> context, |
1802 | void* priv) { |
1803 | Environment* env = Environment::GetCurrent(context); |
1804 | env->SetMethod(target, "parse", Parse); |
1805 | env->SetMethodNoSideEffect(target, "encodeAuth", EncodeAuthSet); |
1806 | env->SetMethodNoSideEffect(target, "domainToASCII", DomainToASCII); |
1807 | env->SetMethodNoSideEffect(target, "domainToUnicode", DomainToUnicode); |
1808 | env->SetMethod(target, "setURLConstructor", SetURLConstructor); |
1809 | |
1810 | #define XX(name, _) NODE_DEFINE_CONSTANT(target, name)do { v8::Isolate* isolate = target->GetIsolate(); v8::Local <v8::Context> context = isolate->GetCurrentContext() ; v8::Local<v8::String> constant_name = v8::String::NewFromUtf8 (isolate, "name", v8::NewStringType::kInternalized).ToLocalChecked (); v8::Local<v8::Number> constant_value = v8::Number:: New(isolate, static_cast<double>(name)); v8::PropertyAttribute constant_attributes = static_cast<v8::PropertyAttribute> (v8::ReadOnly | v8::DontDelete); (target)->DefineOwnProperty (context, constant_name, constant_value, constant_attributes) .Check(); } while (0); |
1811 | FLAGS(XX)XX(URL_FLAGS_NONE, 0) XX(URL_FLAGS_FAILED, 0x01) XX(URL_FLAGS_CANNOT_BE_BASE , 0x02) XX(URL_FLAGS_INVALID_PARSE_STATE, 0x04) XX(URL_FLAGS_TERMINATED , 0x08) XX(URL_FLAGS_SPECIAL, 0x10) XX(URL_FLAGS_HAS_USERNAME , 0x20) XX(URL_FLAGS_HAS_PASSWORD, 0x40) XX(URL_FLAGS_HAS_HOST , 0x80) XX(URL_FLAGS_HAS_PATH, 0x100) XX(URL_FLAGS_HAS_QUERY, 0x200) XX(URL_FLAGS_HAS_FRAGMENT, 0x400) XX(URL_FLAGS_IS_DEFAULT_SCHEME_PORT , 0x800) |
1812 | #undef XX |
1813 | |
1814 | #define XX(name) NODE_DEFINE_CONSTANT(target, name)do { v8::Isolate* isolate = target->GetIsolate(); v8::Local <v8::Context> context = isolate->GetCurrentContext() ; v8::Local<v8::String> constant_name = v8::String::NewFromUtf8 (isolate, "name", v8::NewStringType::kInternalized).ToLocalChecked (); v8::Local<v8::Number> constant_value = v8::Number:: New(isolate, static_cast<double>(name)); v8::PropertyAttribute constant_attributes = static_cast<v8::PropertyAttribute> (v8::ReadOnly | v8::DontDelete); (target)->DefineOwnProperty (context, constant_name, constant_value, constant_attributes) .Check(); } while (0); |
1815 | PARSESTATES(XX)XX(kSchemeStart) XX(kScheme) XX(kNoScheme) XX(kSpecialRelativeOrAuthority ) XX(kPathOrAuthority) XX(kRelative) XX(kRelativeSlash) XX(kSpecialAuthoritySlashes ) XX(kSpecialAuthorityIgnoreSlashes) XX(kAuthority) XX(kHost) XX(kHostname) XX(kPort) XX(kFile) XX(kFileSlash) XX(kFileHost ) XX(kPathStart) XX(kPath) XX(kCannotBeBase) XX(kQuery) XX(kFragment ) |
1816 | #undef XX |
1817 | } |
1818 | } // namespace |
1819 | |
1820 | void RegisterExternalReferences(ExternalReferenceRegistry* registry) { |
1821 | registry->Register(Parse); |
1822 | registry->Register(EncodeAuthSet); |
1823 | registry->Register(DomainToASCII); |
1824 | registry->Register(DomainToUnicode); |
1825 | registry->Register(SetURLConstructor); |
1826 | } |
1827 | |
1828 | std::string URL::ToFilePath() const { |
1829 | if (context_.scheme != "file:") { |
1830 | return ""; |
1831 | } |
1832 | |
1833 | #ifdef _WIN32 |
1834 | const char* slash = "\\"; |
1835 | auto is_slash = [] (char ch) { |
1836 | return ch == '/' || ch == '\\'; |
1837 | }; |
1838 | #else |
1839 | const char* slash = "/"; |
1840 | auto is_slash = [] (char ch) { |
1841 | return ch == '/'; |
1842 | }; |
1843 | if ((context_.flags & URL_FLAGS_HAS_HOST) && |
1844 | context_.host.length() > 0) { |
1845 | return ""; |
1846 | } |
1847 | #endif |
1848 | std::string decoded_path; |
1849 | for (const std::string& part : context_.path) { |
1850 | std::string decoded = PercentDecode(part.c_str(), part.length()); |
1851 | for (char& ch : decoded) { |
1852 | if (is_slash(ch)) { |
1853 | return ""; |
1854 | } |
1855 | } |
1856 | decoded_path += slash + decoded; |
1857 | } |
1858 | |
1859 | #ifdef _WIN32 |
1860 | // TODO(TimothyGu): Use "\\?\" long paths on Windows. |
1861 | |
1862 | // If hostname is set, then we have a UNC path. Pass the hostname through |
1863 | // ToUnicode just in case it is an IDN using punycode encoding. We do not |
1864 | // need to worry about percent encoding because the URL parser will have |
1865 | // already taken care of that for us. Note that this only causes IDNs with an |
1866 | // appropriate `xn--` prefix to be decoded. |
1867 | if ((context_.flags & URL_FLAGS_HAS_HOST) && |
1868 | context_.host.length() > 0) { |
1869 | std::string unicode_host; |
1870 | if (!ToUnicode(context_.host, &unicode_host)) { |
1871 | return ""; |
1872 | } |
1873 | return "\\\\" + unicode_host + decoded_path; |
1874 | } |
1875 | // Otherwise, it's a local path that requires a drive letter. |
1876 | if (decoded_path.length() < 3) { |
1877 | return ""; |
1878 | } |
1879 | if (decoded_path[2] != ':' || |
1880 | !IsASCIIAlpha(decoded_path[1])) { |
1881 | return ""; |
1882 | } |
1883 | // Strip out the leading '\'. |
1884 | return decoded_path.substr(1); |
1885 | #else |
1886 | return decoded_path; |
1887 | #endif |
1888 | } |
1889 | |
1890 | URL URL::FromFilePath(const std::string& file_path) { |
1891 | URL url("file://"); |
1892 | std::string escaped_file_path; |
1893 | for (size_t i = 0; i < file_path.length(); ++i) { |
1894 | escaped_file_path += file_path[i]; |
1895 | if (file_path[i] == '%') |
1896 | escaped_file_path += "25"; |
1897 | } |
1898 | URL::Parse(escaped_file_path.c_str(), escaped_file_path.length(), kPathStart, |
1899 | &url.context_, true, nullptr, false); |
1900 | return url; |
1901 | } |
1902 | |
1903 | // This function works by calling out to a JS function that creates and |
1904 | // returns the JS URL object. Be mindful of the JS<->Native boundary |
1905 | // crossing that is required. |
1906 | MaybeLocal<Value> URL::ToObject(Environment* env) const { |
1907 | Isolate* isolate = env->isolate(); |
1908 | Local<Context> context = env->context(); |
1909 | Context::Scope context_scope(context); |
1910 | |
1911 | const Local<Value> undef = Undefined(isolate); |
1912 | const Local<Value> null = Null(isolate); |
1913 | |
1914 | if (context_.flags & URL_FLAGS_FAILED) |
1915 | return Local<Value>(); |
1916 | |
1917 | Local<Value> argv[] = { |
1918 | undef, |
1919 | undef, |
1920 | undef, |
1921 | undef, |
1922 | null, // host defaults to null |
1923 | null, // port defaults to null |
1924 | undef, |
1925 | null, // query defaults to null |
1926 | null, // fragment defaults to null |
1927 | }; |
1928 | SetArgs(env, argv, context_); |
1929 | |
1930 | MaybeLocal<Value> ret; |
1931 | { |
1932 | TryCatchScope try_catch(env, TryCatchScope::CatchMode::kFatal); |
1933 | |
1934 | // The SetURLConstructor method must have been called already to |
1935 | // set the constructor function used below. SetURLConstructor is |
1936 | // called automatically when the internal/url.js module is loaded |
1937 | // during the internal/bootstrap/node.js processing. |
1938 | ret = env->url_constructor_function() |
1939 | ->Call(env->context(), undef, arraysize(argv), argv); |
1940 | } |
1941 | |
1942 | return ret; |
1943 | } |
1944 | |
1945 | } // namespace url |
1946 | } // namespace node |
1947 | |
1948 | NODE_MODULE_CONTEXT_AWARE_INTERNAL(url, node::url::Initialize)static node::node_module _module = { 108, NM_F_INTERNAL, nullptr , "../src/node_url.cc", nullptr, (node::addon_context_register_func )(node::url::Initialize), "url", nullptr, nullptr}; void _register_url () { node_module_register(&_module); } |
1949 | NODE_MODULE_EXTERNAL_REFERENCE(url, node::url::RegisterExternalReferences)void _register_external_reference_url( node::ExternalReferenceRegistry * registry) { node::url::RegisterExternalReferences(registry) ; } |