| 1 | //===----------------------------------------------------------------------===// |
|---|---|
| 2 | // |
| 3 | // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. |
| 4 | // See https://llvm.org/LICENSE.txt for license information. |
| 5 | // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception |
| 6 | // |
| 7 | //===----------------------------------------------------------------------===// |
| 8 | |
| 9 | #include <__utility/no_destroy.h> |
| 10 | #include <algorithm> |
| 11 | #include <clocale> |
| 12 | #include <codecvt> |
| 13 | #include <cstddef> |
| 14 | #include <cstdio> |
| 15 | #include <cstdlib> |
| 16 | #include <cstring> |
| 17 | #include <locale> |
| 18 | #include <new> |
| 19 | #include <string> |
| 20 | #include <type_traits> |
| 21 | #include <typeinfo> |
| 22 | #include <utility> |
| 23 | #include <vector> |
| 24 | |
| 25 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 26 | # include <cwctype> |
| 27 | #endif |
| 28 | |
| 29 | #if defined(_AIX) |
| 30 | # include <sys/localedef.h> // for __lc_ctype_ptr |
| 31 | #endif |
| 32 | |
| 33 | #if defined(_LIBCPP_MSVCRT) |
| 34 | # define _CTYPE_DISABLE_MACROS |
| 35 | #endif |
| 36 | |
| 37 | #include "include/atomic_support.h" |
| 38 | #include "include/sso_allocator.h" |
| 39 | |
| 40 | // On Linux, wint_t and wchar_t have different signed-ness, and this causes |
| 41 | // lots of noise in the build log, but no bugs that I know of. |
| 42 | _LIBCPP_CLANG_DIAGNOSTIC_IGNORED("-Wsign-conversion") |
| 43 | |
| 44 | _LIBCPP_PUSH_MACROS |
| 45 | #include <__undef_macros> |
| 46 | |
| 47 | _LIBCPP_BEGIN_NAMESPACE_STD |
| 48 | _LIBCPP_BEGIN_EXPLICIT_ABI_ANNOTATIONS |
| 49 | |
| 50 | struct __libcpp_unique_locale { |
| 51 | __libcpp_unique_locale(const char* nm) : __loc_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: nm, base: 0)) {} |
| 52 | |
| 53 | ~__libcpp_unique_locale() { |
| 54 | if (__loc_) |
| 55 | __locale::__freelocale(loc: __loc_); |
| 56 | } |
| 57 | |
| 58 | explicit operator bool() const { return __loc_; } |
| 59 | |
| 60 | __locale::__locale_t& get() { return __loc_; } |
| 61 | |
| 62 | __locale::__locale_t __loc_; |
| 63 | |
| 64 | __libcpp_unique_locale(__libcpp_unique_locale const&) = delete; |
| 65 | __libcpp_unique_locale& operator=(__libcpp_unique_locale const&) = delete; |
| 66 | }; |
| 67 | |
| 68 | #ifdef __cloc_defined |
| 69 | __locale::__locale_t __cloc() { |
| 70 | // In theory this could create a race condition. In practice |
| 71 | // the race condition is non-fatal since it will just create |
| 72 | // a little resource leak. Better approach would be appreciated. |
| 73 | static __locale::__locale_t result = __locale::__newlocale(_LIBCPP_ALL_MASK, locale: "C", base: 0); |
| 74 | return result; |
| 75 | } |
| 76 | #endif // __cloc_defined |
| 77 | |
| 78 | namespace { |
| 79 | |
| 80 | struct releaser { |
| 81 | void operator()(locale::facet* p) { p->__release_shared(); } |
| 82 | }; |
| 83 | |
| 84 | template <class T, class... Args> |
| 85 | T& make(Args... args) { |
| 86 | alignas(T) static std::byte buf[sizeof(T)]; |
| 87 | auto* obj = ::new (&buf) T(args...); |
| 88 | return *obj; |
| 89 | } |
| 90 | |
| 91 | string build_name(const string& other, const string& one, locale::category c) { |
| 92 | if (other == "*"|| one == "*") |
| 93 | return "*"; |
| 94 | if (c == locale::none || other == one) |
| 95 | return other; |
| 96 | |
| 97 | // FIXME: Handle the more complicated cases, such as when the locale has |
| 98 | // different names for different categories. |
| 99 | return "*"; |
| 100 | } |
| 101 | |
| 102 | } // namespace |
| 103 | |
| 104 | const locale::category locale::none; |
| 105 | const locale::category locale::collate; |
| 106 | const locale::category locale::ctype; |
| 107 | const locale::category locale::monetary; |
| 108 | const locale::category locale::numeric; |
| 109 | const locale::category locale::time; |
| 110 | const locale::category locale::messages; |
| 111 | const locale::category locale::all; |
| 112 | |
| 113 | class _LIBCPP_HIDDEN locale::__imp : public facet { |
| 114 | enum { N = 30 }; |
| 115 | vector<facet*, __sso_allocator<facet*, N> > facets_; |
| 116 | string name_; |
| 117 | |
| 118 | public: |
| 119 | explicit __imp(size_t refs = 0); |
| 120 | explicit __imp(const string& name, size_t refs = 0); |
| 121 | __imp(const __imp&); |
| 122 | __imp(const __imp&, const string&, locale::category c); |
| 123 | __imp(const __imp& other, const __imp& one, locale::category c); |
| 124 | __imp(const __imp&, facet* f, long id); |
| 125 | ~__imp(); |
| 126 | |
| 127 | const string& name() const { return name_; } |
| 128 | bool has_facet(long id) const { return static_cast<size_t>(id) < facets_.size() && facets_[static_cast<size_t>(id)]; } |
| 129 | const locale::facet* use_facet(long id) const; |
| 130 | |
| 131 | void acquire(); |
| 132 | void release(); |
| 133 | static __no_destroy<__imp> classic_locale_imp_; |
| 134 | |
| 135 | private: |
| 136 | void install(facet* f, long id); |
| 137 | template <class F> |
| 138 | void install(F* f) { |
| 139 | install(f, f->id.__get()); |
| 140 | } |
| 141 | template <class F> |
| 142 | void install_from(const __imp& other); |
| 143 | }; |
| 144 | |
| 145 | locale::__imp::__imp(size_t refs) : facet(refs), facets_(N), name_("C") { |
| 146 | facets_.clear(); |
| 147 | install(f: &make<std::collate<char> >(args: 1u)); |
| 148 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 149 | install(f: &make<std::collate<wchar_t> >(args: 1u)); |
| 150 | #endif |
| 151 | install(f: &make<std::ctype<char> >(args: nullptr, args: false, args: 1u)); |
| 152 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 153 | install(f: &make<std::ctype<wchar_t> >(args: 1u)); |
| 154 | #endif |
| 155 | install(f: &make<codecvt<char, char, mbstate_t> >(args: 1u)); |
| 156 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 157 | install(f: &make<codecvt<wchar_t, char, mbstate_t> >(args: 1u)); |
| 158 | #endif |
| 159 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 160 | install(f: &make<codecvt<char16_t, char, mbstate_t> >(args: 1u)); |
| 161 | install(f: &make<codecvt<char32_t, char, mbstate_t> >(args: 1u)); |
| 162 | #if _LIBCPP_HAS_CHAR8_T |
| 163 | install(f: &make<codecvt<char16_t, char8_t, mbstate_t> >(args: 1u)); |
| 164 | install(f: &make<codecvt<char32_t, char8_t, mbstate_t> >(args: 1u)); |
| 165 | #endif |
| 166 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 167 | install(f: &make<numpunct<char> >(args: 1u)); |
| 168 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 169 | install(f: &make<numpunct<wchar_t> >(args: 1u)); |
| 170 | #endif |
| 171 | install(f: &make<num_get<char> >(args: 1u)); |
| 172 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 173 | install(f: &make<num_get<wchar_t> >(args: 1u)); |
| 174 | #endif |
| 175 | install(f: &make<num_put<char> >(args: 1u)); |
| 176 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 177 | install(f: &make<num_put<wchar_t> >(args: 1u)); |
| 178 | #endif |
| 179 | install(f: &make<moneypunct<char, false> >(args: 1u)); |
| 180 | install(f: &make<moneypunct<char, true> >(args: 1u)); |
| 181 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 182 | install(f: &make<moneypunct<wchar_t, false> >(args: 1u)); |
| 183 | install(f: &make<moneypunct<wchar_t, true> >(args: 1u)); |
| 184 | #endif |
| 185 | install(f: &make<money_get<char> >(args: 1u)); |
| 186 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 187 | install(f: &make<money_get<wchar_t> >(args: 1u)); |
| 188 | #endif |
| 189 | install(f: &make<money_put<char> >(args: 1u)); |
| 190 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 191 | install(f: &make<money_put<wchar_t> >(args: 1u)); |
| 192 | #endif |
| 193 | install(f: &make<time_get<char> >(args: 1u)); |
| 194 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 195 | install(f: &make<time_get<wchar_t> >(args: 1u)); |
| 196 | #endif |
| 197 | install(f: &make<time_put<char> >(args: 1u)); |
| 198 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 199 | install(f: &make<time_put<wchar_t> >(args: 1u)); |
| 200 | #endif |
| 201 | install(f: &make<std::messages<char> >(args: 1u)); |
| 202 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 203 | install(f: &make<std::messages<wchar_t> >(args: 1u)); |
| 204 | #endif |
| 205 | } |
| 206 | |
| 207 | locale::__imp::__imp(const string& name, size_t refs) : facet(refs), facets_(N), name_(name) { |
| 208 | __exception_guard guard([&] { |
| 209 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 210 | if (facets_[i]) |
| 211 | facets_[i]->__release_shared(); |
| 212 | }); |
| 213 | facets_ = locale::classic().__locale_->facets_; |
| 214 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 215 | if (facets_[i]) |
| 216 | facets_[i]->__add_shared(); |
| 217 | install(f: new collate_byname<char>(name_)); |
| 218 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 219 | install(f: new collate_byname<wchar_t>(name_)); |
| 220 | #endif |
| 221 | install(f: new ctype_byname<char>(name_)); |
| 222 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 223 | install(f: new ctype_byname<wchar_t>(name_)); |
| 224 | #endif |
| 225 | install(f: new codecvt_byname<char, char, mbstate_t>(name_)); |
| 226 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 227 | install(f: new codecvt_byname<wchar_t, char, mbstate_t>(name_)); |
| 228 | #endif |
| 229 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 230 | install(f: new codecvt_byname<char16_t, char, mbstate_t>(name_)); |
| 231 | install(f: new codecvt_byname<char32_t, char, mbstate_t>(name_)); |
| 232 | #if _LIBCPP_HAS_CHAR8_T |
| 233 | install(f: new codecvt_byname<char16_t, char8_t, mbstate_t>(name_)); |
| 234 | install(f: new codecvt_byname<char32_t, char8_t, mbstate_t>(name_)); |
| 235 | #endif |
| 236 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 237 | install(f: new numpunct_byname<char>(name_)); |
| 238 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 239 | install(f: new numpunct_byname<wchar_t>(name_)); |
| 240 | #endif |
| 241 | install(f: new moneypunct_byname<char, false>(name_)); |
| 242 | install(f: new moneypunct_byname<char, true>(name_)); |
| 243 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 244 | install(f: new moneypunct_byname<wchar_t, false>(name_)); |
| 245 | install(f: new moneypunct_byname<wchar_t, true>(name_)); |
| 246 | #endif |
| 247 | install(f: new time_get_byname<char>(name_)); |
| 248 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 249 | install(f: new time_get_byname<wchar_t>(name_)); |
| 250 | #endif |
| 251 | install(f: new time_put_byname<char>(name_)); |
| 252 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 253 | install(f: new time_put_byname<wchar_t>(name_)); |
| 254 | #endif |
| 255 | install(f: new messages_byname<char>(name_)); |
| 256 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 257 | install(f: new messages_byname<wchar_t>(name_)); |
| 258 | #endif |
| 259 | guard.__complete(); |
| 260 | } |
| 261 | |
| 262 | locale::__imp::__imp(const __imp& other) : facets_(max<size_t>(a: N, b: other.facets_.size())), name_(other.name_) { |
| 263 | facets_ = other.facets_; |
| 264 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 265 | if (facets_[i]) |
| 266 | facets_[i]->__add_shared(); |
| 267 | } |
| 268 | |
| 269 | locale::__imp::__imp(const __imp& other, const string& name, locale::category c) |
| 270 | : facets_(N), name_(build_name(other: other.name_, one: name, c)) { |
| 271 | facets_ = other.facets_; |
| 272 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 273 | if (facets_[i]) |
| 274 | facets_[i]->__add_shared(); |
| 275 | __exception_guard guard([&] { |
| 276 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 277 | if (facets_[i]) |
| 278 | facets_[i]->__release_shared(); |
| 279 | }); |
| 280 | if (c & locale::collate) { |
| 281 | install(f: new collate_byname<char>(name)); |
| 282 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 283 | install(f: new collate_byname<wchar_t>(name)); |
| 284 | #endif |
| 285 | } |
| 286 | if (c & locale::ctype) { |
| 287 | install(f: new ctype_byname<char>(name)); |
| 288 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 289 | install(f: new ctype_byname<wchar_t>(name)); |
| 290 | #endif |
| 291 | install(f: new codecvt_byname<char, char, mbstate_t>(name)); |
| 292 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 293 | install(f: new codecvt_byname<wchar_t, char, mbstate_t>(name)); |
| 294 | #endif |
| 295 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 296 | install(f: new codecvt_byname<char16_t, char, mbstate_t>(name)); |
| 297 | install(f: new codecvt_byname<char32_t, char, mbstate_t>(name)); |
| 298 | #if _LIBCPP_HAS_CHAR8_T |
| 299 | install(f: new codecvt_byname<char16_t, char8_t, mbstate_t>(name)); |
| 300 | install(f: new codecvt_byname<char32_t, char8_t, mbstate_t>(name)); |
| 301 | #endif |
| 302 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 303 | } |
| 304 | if (c & locale::monetary) { |
| 305 | install(f: new moneypunct_byname<char, false>(name)); |
| 306 | install(f: new moneypunct_byname<char, true>(name)); |
| 307 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 308 | install(f: new moneypunct_byname<wchar_t, false>(name)); |
| 309 | install(f: new moneypunct_byname<wchar_t, true>(name)); |
| 310 | #endif |
| 311 | } |
| 312 | if (c & locale::numeric) { |
| 313 | install(f: new numpunct_byname<char>(name)); |
| 314 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 315 | install(f: new numpunct_byname<wchar_t>(name)); |
| 316 | #endif |
| 317 | } |
| 318 | if (c & locale::time) { |
| 319 | install(f: new time_get_byname<char>(name)); |
| 320 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 321 | install(f: new time_get_byname<wchar_t>(name)); |
| 322 | #endif |
| 323 | install(f: new time_put_byname<char>(name)); |
| 324 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 325 | install(f: new time_put_byname<wchar_t>(name)); |
| 326 | #endif |
| 327 | } |
| 328 | if (c & locale::messages) { |
| 329 | install(f: new messages_byname<char>(name)); |
| 330 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 331 | install(f: new messages_byname<wchar_t>(name)); |
| 332 | #endif |
| 333 | } |
| 334 | guard.__complete(); |
| 335 | } |
| 336 | |
| 337 | template <class F> |
| 338 | inline void locale::__imp::install_from(const locale::__imp& one) { |
| 339 | long id = F::id.__get(); |
| 340 | install(const_cast<F*>(static_cast<const F*>(one.use_facet(id))), id); |
| 341 | } |
| 342 | |
| 343 | locale::__imp::__imp(const __imp& other, const __imp& one, locale::category c) |
| 344 | : facets_(N), name_(build_name(other: other.name_, one: one.name_, c)) { |
| 345 | facets_ = other.facets_; |
| 346 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 347 | if (facets_[i]) |
| 348 | facets_[i]->__add_shared(); |
| 349 | __exception_guard guard([&] { |
| 350 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 351 | if (facets_[i]) |
| 352 | facets_[i]->__release_shared(); |
| 353 | }); |
| 354 | |
| 355 | if (c & locale::collate) { |
| 356 | install_from<std::collate<char> >(one); |
| 357 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 358 | install_from<std::collate<wchar_t> >(one); |
| 359 | #endif |
| 360 | } |
| 361 | if (c & locale::ctype) { |
| 362 | install_from<std::ctype<char> >(one); |
| 363 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 364 | install_from<std::ctype<wchar_t> >(one); |
| 365 | #endif |
| 366 | install_from<std::codecvt<char, char, mbstate_t> >(one); |
| 367 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 368 | install_from<std::codecvt<char16_t, char, mbstate_t> >(one); |
| 369 | install_from<std::codecvt<char32_t, char, mbstate_t> >(one); |
| 370 | #if _LIBCPP_HAS_CHAR8_T |
| 371 | install_from<std::codecvt<char16_t, char8_t, mbstate_t> >(one); |
| 372 | install_from<std::codecvt<char32_t, char8_t, mbstate_t> >(one); |
| 373 | #endif |
| 374 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 375 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 376 | install_from<std::codecvt<wchar_t, char, mbstate_t> >(one); |
| 377 | #endif |
| 378 | } |
| 379 | if (c & locale::monetary) { |
| 380 | install_from<moneypunct<char, false> >(one); |
| 381 | install_from<moneypunct<char, true> >(one); |
| 382 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 383 | install_from<moneypunct<wchar_t, false> >(one); |
| 384 | install_from<moneypunct<wchar_t, true> >(one); |
| 385 | #endif |
| 386 | install_from<money_get<char> >(one); |
| 387 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 388 | install_from<money_get<wchar_t> >(one); |
| 389 | #endif |
| 390 | install_from<money_put<char> >(one); |
| 391 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 392 | install_from<money_put<wchar_t> >(one); |
| 393 | #endif |
| 394 | } |
| 395 | if (c & locale::numeric) { |
| 396 | install_from<numpunct<char> >(one); |
| 397 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 398 | install_from<numpunct<wchar_t> >(one); |
| 399 | #endif |
| 400 | install_from<num_get<char> >(one); |
| 401 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 402 | install_from<num_get<wchar_t> >(one); |
| 403 | #endif |
| 404 | install_from<num_put<char> >(one); |
| 405 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 406 | install_from<num_put<wchar_t> >(one); |
| 407 | #endif |
| 408 | } |
| 409 | if (c & locale::time) { |
| 410 | install_from<time_get<char> >(one); |
| 411 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 412 | install_from<time_get<wchar_t> >(one); |
| 413 | #endif |
| 414 | install_from<time_put<char> >(one); |
| 415 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 416 | install_from<time_put<wchar_t> >(one); |
| 417 | #endif |
| 418 | } |
| 419 | if (c & locale::messages) { |
| 420 | install_from<std::messages<char> >(one); |
| 421 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 422 | install_from<std::messages<wchar_t> >(one); |
| 423 | #endif |
| 424 | } |
| 425 | guard.__complete(); |
| 426 | } |
| 427 | |
| 428 | locale::__imp::__imp(const __imp& other, facet* f, long id) |
| 429 | : facets_(max<size_t>(a: N, b: other.facets_.size() + 1)), name_("*") { |
| 430 | f->__add_shared(); |
| 431 | unique_ptr<facet, releaser> hold(f); |
| 432 | facets_ = other.facets_; |
| 433 | for (unsigned i = 0; i < other.facets_.size(); ++i) |
| 434 | if (facets_[i]) |
| 435 | facets_[i]->__add_shared(); |
| 436 | install(f: hold.get(), id); |
| 437 | } |
| 438 | |
| 439 | locale::__imp::~__imp() { |
| 440 | for (unsigned i = 0; i < facets_.size(); ++i) |
| 441 | if (facets_[i]) |
| 442 | facets_[i]->__release_shared(); |
| 443 | } |
| 444 | |
| 445 | void locale::__imp::install(facet* f, long id) { |
| 446 | f->__add_shared(); |
| 447 | unique_ptr<facet, releaser> hold(f); |
| 448 | if (static_cast<size_t>(id) >= facets_.size()) |
| 449 | facets_.resize(new_size: static_cast<size_t>(id + 1)); |
| 450 | if (facets_[static_cast<size_t>(id)]) |
| 451 | facets_[static_cast<size_t>(id)]->__release_shared(); |
| 452 | facets_[static_cast<size_t>(id)] = hold.release(); |
| 453 | } |
| 454 | |
| 455 | const locale::facet* locale::__imp::use_facet(long id) const { |
| 456 | if (!has_facet(id)) |
| 457 | std::__throw_bad_cast(); |
| 458 | return facets_[static_cast<size_t>(id)]; |
| 459 | } |
| 460 | |
| 461 | // locale |
| 462 | |
| 463 | // We don't do reference counting on the classic locale. |
| 464 | // It's never destroyed anyway, but atomic reference counting may be very |
| 465 | // expensive in parallel applications. The classic locale is used by default |
| 466 | // in all streams. Note: if a new global locale is installed, then we lose |
| 467 | // the benefit of no reference counting. |
| 468 | constinit __no_destroy<locale::__imp> |
| 469 | locale::__imp::classic_locale_imp_(__uninitialized_tag{}); // initialized below in classic() |
| 470 | |
| 471 | const locale& locale::classic() { |
| 472 | static const __no_destroy<locale> classic_locale(__private_constructor_tag{}, [] { |
| 473 | // executed exactly once on first initialization of `classic_locale` |
| 474 | locale::__imp::classic_locale_imp_.__emplace(args: 1u); |
| 475 | return &locale::__imp::classic_locale_imp_.__get(); |
| 476 | }()); |
| 477 | return classic_locale.__get(); |
| 478 | } |
| 479 | |
| 480 | locale& locale::__global() { |
| 481 | static __no_destroy<locale> g(locale::classic()); |
| 482 | return g.__get(); |
| 483 | } |
| 484 | |
| 485 | void locale::__imp::acquire() { |
| 486 | if (this != &locale::__imp::classic_locale_imp_.__get()) |
| 487 | __add_shared(); |
| 488 | } |
| 489 | |
| 490 | void locale::__imp::release() { |
| 491 | if (this != &locale::__imp::classic_locale_imp_.__get()) |
| 492 | __release_shared(); |
| 493 | } |
| 494 | |
| 495 | locale::locale() noexcept : __locale_(__global().__locale_) { __locale_->acquire(); } |
| 496 | |
| 497 | locale::locale(const locale& l) noexcept : __locale_(l.__locale_) { __locale_->acquire(); } |
| 498 | |
| 499 | locale::~locale() { __locale_->release(); } |
| 500 | |
| 501 | const locale& locale::operator=(const locale& other) noexcept { |
| 502 | other.__locale_->acquire(); |
| 503 | __locale_->release(); |
| 504 | __locale_ = other.__locale_; |
| 505 | return *this; |
| 506 | } |
| 507 | |
| 508 | locale::locale(const char* name) |
| 509 | : __locale_(name ? new __imp(name) : (__throw_runtime_error("locale constructed with null"), nullptr)) { |
| 510 | __locale_->acquire(); |
| 511 | } |
| 512 | |
| 513 | locale::locale(const string& name) : __locale_(new __imp(name)) { __locale_->acquire(); } |
| 514 | |
| 515 | locale::locale(const locale& other, const char* name, category c) |
| 516 | : __locale_(name ? new __imp(*other.__locale_, name, c) |
| 517 | : (__throw_runtime_error("locale constructed with null"), nullptr)) { |
| 518 | __locale_->acquire(); |
| 519 | } |
| 520 | |
| 521 | locale::locale(const locale& other, const string& name, category c) : __locale_(new __imp(*other.__locale_, name, c)) { |
| 522 | __locale_->acquire(); |
| 523 | } |
| 524 | |
| 525 | locale::locale(const locale& other, const locale& one, category c) |
| 526 | : __locale_(new __imp(*other.__locale_, *one.__locale_, c)) { |
| 527 | __locale_->acquire(); |
| 528 | } |
| 529 | |
| 530 | string locale::name() const { return __locale_->name(); } |
| 531 | |
| 532 | void locale::__install_ctor(const locale& other, facet* f, long facet_id) { |
| 533 | if (f) |
| 534 | __locale_ = new __imp(*other.__locale_, f, facet_id); |
| 535 | else |
| 536 | __locale_ = other.__locale_; |
| 537 | __locale_->acquire(); |
| 538 | } |
| 539 | |
| 540 | locale locale::global(const locale& loc) { |
| 541 | locale& g = __global(); |
| 542 | locale r = g; |
| 543 | g = loc; |
| 544 | if (g.name() != "*") |
| 545 | __locale::__setlocale(_LIBCPP_LC_ALL, locale: g.name().c_str()); |
| 546 | return r; |
| 547 | } |
| 548 | |
| 549 | bool locale::has_facet(id& x) const { return __locale_->has_facet(id: x.__get()); } |
| 550 | |
| 551 | const locale::facet* locale::use_facet(id& x) const { return __locale_->use_facet(id: x.__get()); } |
| 552 | |
| 553 | bool locale::operator==(const locale& y) const { |
| 554 | return (__locale_ == y.__locale_) || (__locale_->name() != "*"&& __locale_->name() == y.__locale_->name()); |
| 555 | } |
| 556 | |
| 557 | // locale::facet |
| 558 | |
| 559 | locale::facet::~facet() {} |
| 560 | |
| 561 | void locale::facet::__on_zero_shared() noexcept { delete this; } |
| 562 | |
| 563 | // locale::id |
| 564 | |
| 565 | long locale::id::__get() { |
| 566 | constexpr __id_type uninitialized = 0; |
| 567 | constexpr __id_type pending = 1; |
| 568 | constexpr __id_type first_id = 2; |
| 569 | |
| 570 | constinit static __id_type next_id = first_id; |
| 571 | |
| 572 | auto val = __libcpp_atomic_load(val: &__id_, order: _AO_Relaxed); |
| 573 | if (val >= first_id) |
| 574 | return val - first_id; |
| 575 | |
| 576 | if (val == uninitialized && __libcpp_atomic_compare_exchange_strong(val: &__id_, expected: &val, after: pending)) { |
| 577 | auto new_id = __libcpp_atomic_add(val: &next_id, a: 1) - 1; |
| 578 | __libcpp_atomic_store(dest: &__id_, val: new_id); |
| 579 | return new_id - first_id; |
| 580 | } |
| 581 | |
| 582 | // Another thread is already initializing __id_. Wait for that in a spin loop. The initialization should be incredibly |
| 583 | // fast, a spin loop should be efficient enough. |
| 584 | while (true) { |
| 585 | auto id = __libcpp_atomic_load(val: &__id_); |
| 586 | if (id >= first_id) |
| 587 | return id - first_id; |
| 588 | } |
| 589 | } |
| 590 | |
| 591 | // template <> class collate_byname<char> |
| 592 | |
| 593 | collate_byname<char>::collate_byname(const char* n, size_t refs) |
| 594 | : collate<char>(refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: n, base: 0)) { |
| 595 | if (__l_ == 0) |
| 596 | std::__throw_runtime_error( |
| 597 | ("collate_byname<char>::collate_byname" |
| 598 | " failed to construct for "+ |
| 599 | string(n)) |
| 600 | .c_str()); |
| 601 | } |
| 602 | |
| 603 | collate_byname<char>::collate_byname(const string& name, size_t refs) |
| 604 | : collate<char>(refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: name.c_str(), base: 0)) { |
| 605 | if (__l_ == 0) |
| 606 | std::__throw_runtime_error( |
| 607 | ("collate_byname<char>::collate_byname" |
| 608 | " failed to construct for "+ |
| 609 | name) |
| 610 | .c_str()); |
| 611 | } |
| 612 | |
| 613 | collate_byname<char>::~collate_byname() { __locale::__freelocale(loc: __l_); } |
| 614 | |
| 615 | int collate_byname<char>::do_compare( |
| 616 | const char_type* __lo1, const char_type* __hi1, const char_type* __lo2, const char_type* __hi2) const { |
| 617 | string_type lhs(__lo1, __hi1); |
| 618 | string_type rhs(__lo2, __hi2); |
| 619 | int r = __locale::__strcoll(s1: lhs.c_str(), s2: rhs.c_str(), loc: __l_); |
| 620 | if (r < 0) |
| 621 | return -1; |
| 622 | if (r > 0) |
| 623 | return 1; |
| 624 | return r; |
| 625 | } |
| 626 | |
| 627 | collate_byname<char>::string_type collate_byname<char>::do_transform(const char_type* lo, const char_type* hi) const { |
| 628 | const string_type in(lo, hi); |
| 629 | string_type out(__locale::__strxfrm(dest: 0, src: in.c_str(), n: 0, loc: __l_), char()); |
| 630 | __locale::__strxfrm(dest: const_cast<char*>(out.c_str()), src: in.c_str(), n: out.size() + 1, loc: __l_); |
| 631 | return out; |
| 632 | } |
| 633 | |
| 634 | // template <> class collate_byname<wchar_t> |
| 635 | |
| 636 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 637 | collate_byname<wchar_t>::collate_byname(const char* n, size_t refs) |
| 638 | : collate<wchar_t>(refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: n, base: 0)) { |
| 639 | if (__l_ == 0) |
| 640 | std::__throw_runtime_error( |
| 641 | ("collate_byname<wchar_t>::collate_byname(size_t refs)" |
| 642 | " failed to construct for "+ |
| 643 | string(n)) |
| 644 | .c_str()); |
| 645 | } |
| 646 | |
| 647 | collate_byname<wchar_t>::collate_byname(const string& name, size_t refs) |
| 648 | : collate<wchar_t>(refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: name.c_str(), base: 0)) { |
| 649 | if (__l_ == 0) |
| 650 | std::__throw_runtime_error( |
| 651 | ("collate_byname<wchar_t>::collate_byname(size_t refs)" |
| 652 | " failed to construct for "+ |
| 653 | name) |
| 654 | .c_str()); |
| 655 | } |
| 656 | |
| 657 | collate_byname<wchar_t>::~collate_byname() { __locale::__freelocale(loc: __l_); } |
| 658 | |
| 659 | int collate_byname<wchar_t>::do_compare( |
| 660 | const char_type* __lo1, const char_type* __hi1, const char_type* __lo2, const char_type* __hi2) const { |
| 661 | string_type lhs(__lo1, __hi1); |
| 662 | string_type rhs(__lo2, __hi2); |
| 663 | int r = __locale::__wcscoll(ws1: lhs.c_str(), ws2: rhs.c_str(), loc: __l_); |
| 664 | if (r < 0) |
| 665 | return -1; |
| 666 | if (r > 0) |
| 667 | return 1; |
| 668 | return r; |
| 669 | } |
| 670 | |
| 671 | collate_byname<wchar_t>::string_type |
| 672 | collate_byname<wchar_t>::do_transform(const char_type* lo, const char_type* hi) const { |
| 673 | const string_type in(lo, hi); |
| 674 | string_type out(__locale::__wcsxfrm(dest: 0, src: in.c_str(), n: 0, loc: __l_), wchar_t()); |
| 675 | __locale::__wcsxfrm(dest: const_cast<wchar_t*>(out.c_str()), src: in.c_str(), n: out.size() + 1, loc: __l_); |
| 676 | return out; |
| 677 | } |
| 678 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 679 | |
| 680 | const ctype_base::mask ctype_base::space; |
| 681 | const ctype_base::mask ctype_base::print; |
| 682 | const ctype_base::mask ctype_base::cntrl; |
| 683 | const ctype_base::mask ctype_base::upper; |
| 684 | const ctype_base::mask ctype_base::lower; |
| 685 | const ctype_base::mask ctype_base::alpha; |
| 686 | const ctype_base::mask ctype_base::digit; |
| 687 | const ctype_base::mask ctype_base::punct; |
| 688 | const ctype_base::mask ctype_base::xdigit; |
| 689 | const ctype_base::mask ctype_base::blank; |
| 690 | const ctype_base::mask ctype_base::alnum; |
| 691 | const ctype_base::mask ctype_base::graph; |
| 692 | |
| 693 | // template <> class ctype<wchar_t>; |
| 694 | |
| 695 | template <class CharT> |
| 696 | static CharT to_upper_impl(CharT c) { |
| 697 | if (c < 'a' || c > 'z') |
| 698 | return c; |
| 699 | return c & ~0x20; |
| 700 | } |
| 701 | |
| 702 | template <class CharT> |
| 703 | static CharT to_lower_impl(CharT c) { |
| 704 | if (c < 'A' || c > 'Z') |
| 705 | return c; |
| 706 | return c | 0x20; |
| 707 | } |
| 708 | |
| 709 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 710 | constinit locale::id ctype<wchar_t>::id; |
| 711 | |
| 712 | ctype<wchar_t>::~ctype() {} |
| 713 | |
| 714 | bool ctype<wchar_t>::do_is(mask m, char_type c) const { |
| 715 | return std::__libcpp_isascii(c: c) ? (ctype<char>::classic_table()[c] & m) != 0 : false; |
| 716 | } |
| 717 | |
| 718 | const wchar_t* ctype<wchar_t>::do_is(const char_type* low, const char_type* high, mask* vec) const { |
| 719 | for (; low != high; ++low, ++vec) |
| 720 | *vec = static_cast<mask>(std::__libcpp_isascii(c: *low) ? ctype<char>::classic_table()[*low] : 0); |
| 721 | return low; |
| 722 | } |
| 723 | |
| 724 | const wchar_t* ctype<wchar_t>::do_scan_is(mask m, const char_type* low, const char_type* high) const { |
| 725 | for (; low != high; ++low) |
| 726 | if (std::__libcpp_isascii(c: *low) && (ctype<char>::classic_table()[*low] & m)) |
| 727 | break; |
| 728 | return low; |
| 729 | } |
| 730 | |
| 731 | const wchar_t* ctype<wchar_t>::do_scan_not(mask m, const char_type* low, const char_type* high) const { |
| 732 | for (; low != high; ++low) |
| 733 | if (!(std::__libcpp_isascii(c: *low) && (ctype<char>::classic_table()[*low] & m))) |
| 734 | break; |
| 735 | return low; |
| 736 | } |
| 737 | |
| 738 | wchar_t ctype<wchar_t>::do_toupper(char_type c) const { return to_upper_impl(c); } |
| 739 | |
| 740 | const wchar_t* ctype<wchar_t>::do_toupper(char_type* low, const char_type* high) const { |
| 741 | for (; low != high; ++low) |
| 742 | *low = to_upper_impl(c: *low); |
| 743 | return low; |
| 744 | } |
| 745 | |
| 746 | wchar_t ctype<wchar_t>::do_tolower(char_type c) const { return to_lower_impl(c); } |
| 747 | |
| 748 | const wchar_t* ctype<wchar_t>::do_tolower(char_type* low, const char_type* high) const { |
| 749 | for (; low != high; ++low) |
| 750 | *low = to_lower_impl(c: *low); |
| 751 | return low; |
| 752 | } |
| 753 | |
| 754 | wchar_t ctype<wchar_t>::do_widen(char c) const { return c; } |
| 755 | |
| 756 | const char* ctype<wchar_t>::do_widen(const char* low, const char* high, char_type* dest) const { |
| 757 | for (; low != high; ++low, ++dest) |
| 758 | *dest = *low; |
| 759 | return low; |
| 760 | } |
| 761 | |
| 762 | char ctype<wchar_t>::do_narrow(char_type c, char dfault) const { |
| 763 | if (std::__libcpp_isascii(c: c)) |
| 764 | return static_cast<char>(c); |
| 765 | return dfault; |
| 766 | } |
| 767 | |
| 768 | const wchar_t* ctype<wchar_t>::do_narrow(const char_type* low, const char_type* high, char dfault, char* dest) const { |
| 769 | for (; low != high; ++low, ++dest) |
| 770 | if (std::__libcpp_isascii(c: *low)) |
| 771 | *dest = static_cast<char>(*low); |
| 772 | else |
| 773 | *dest = dfault; |
| 774 | return low; |
| 775 | } |
| 776 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 777 | |
| 778 | // template <> class ctype<char>; |
| 779 | |
| 780 | constinit locale::id ctype<char>::id; |
| 781 | |
| 782 | const size_t ctype<char>::table_size; |
| 783 | |
| 784 | ctype<char>::ctype(const mask* tab, bool del, size_t refs) : locale::facet(refs), __tab_(tab), __del_(del) { |
| 785 | if (__tab_ == 0) |
| 786 | __tab_ = classic_table(); |
| 787 | } |
| 788 | |
| 789 | ctype<char>::~ctype() { |
| 790 | if (__tab_ && __del_) |
| 791 | delete[] __tab_; |
| 792 | } |
| 793 | |
| 794 | char ctype<char>::do_toupper(char_type c) const { return to_upper_impl(c); } |
| 795 | |
| 796 | const char* ctype<char>::do_toupper(char_type* low, const char_type* high) const { |
| 797 | for (; low != high; ++low) |
| 798 | *low = to_upper_impl(c: *low); |
| 799 | return low; |
| 800 | } |
| 801 | |
| 802 | char ctype<char>::do_tolower(char_type c) const { return to_lower_impl(c); } |
| 803 | |
| 804 | const char* ctype<char>::do_tolower(char_type* low, const char_type* high) const { |
| 805 | for (; low != high; ++low) |
| 806 | *low = to_lower_impl(c: *low); |
| 807 | return low; |
| 808 | } |
| 809 | |
| 810 | char ctype<char>::do_widen(char c) const { return c; } |
| 811 | |
| 812 | const char* ctype<char>::do_widen(const char* low, const char* high, char_type* dest) const { |
| 813 | for (; low != high; ++low, ++dest) |
| 814 | *dest = *low; |
| 815 | return low; |
| 816 | } |
| 817 | |
| 818 | char ctype<char>::do_narrow(char_type c, char dfault) const { |
| 819 | if (std::__libcpp_isascii(c: c)) |
| 820 | return static_cast<char>(c); |
| 821 | return dfault; |
| 822 | } |
| 823 | |
| 824 | const char* ctype<char>::do_narrow(const char_type* low, const char_type* high, char dfault, char* dest) const { |
| 825 | for (; low != high; ++low, ++dest) |
| 826 | if (std::__libcpp_isascii(c: *low)) |
| 827 | *dest = *low; |
| 828 | else |
| 829 | *dest = dfault; |
| 830 | return low; |
| 831 | } |
| 832 | |
| 833 | #if defined(__EMSCRIPTEN__) |
| 834 | extern "C"const unsigned short** __ctype_b_loc(); |
| 835 | extern "C"const int** __ctype_tolower_loc(); |
| 836 | extern "C"const int** __ctype_toupper_loc(); |
| 837 | #endif |
| 838 | |
| 839 | #if _LIBCPP_PROVIDES_DEFAULT_RUNE_TABLE |
| 840 | const ctype<char>::mask* ctype<char>::classic_table() noexcept { |
| 841 | // clang-format off |
| 842 | static constexpr const ctype<char>::mask builtin_table[table_size] = { |
| 843 | cntrl, cntrl, |
| 844 | cntrl, cntrl, |
| 845 | cntrl, cntrl, |
| 846 | cntrl, cntrl, |
| 847 | cntrl, cntrl | space | blank, |
| 848 | cntrl | space, cntrl | space, |
| 849 | cntrl | space, cntrl | space, |
| 850 | cntrl, cntrl, |
| 851 | cntrl, cntrl, |
| 852 | cntrl, cntrl, |
| 853 | cntrl, cntrl, |
| 854 | cntrl, cntrl, |
| 855 | cntrl, cntrl, |
| 856 | cntrl, cntrl, |
| 857 | cntrl, cntrl, |
| 858 | cntrl, cntrl, |
| 859 | space | blank | print, punct | print, |
| 860 | punct | print, punct | print, |
| 861 | punct | print, punct | print, |
| 862 | punct | print, punct | print, |
| 863 | punct | print, punct | print, |
| 864 | punct | print, punct | print, |
| 865 | punct | print, punct | print, |
| 866 | punct | print, punct | print, |
| 867 | digit | print | xdigit, digit | print | xdigit, |
| 868 | digit | print | xdigit, digit | print | xdigit, |
| 869 | digit | print | xdigit, digit | print | xdigit, |
| 870 | digit | print | xdigit, digit | print | xdigit, |
| 871 | digit | print | xdigit, digit | print | xdigit, |
| 872 | punct | print, punct | print, |
| 873 | punct | print, punct | print, |
| 874 | punct | print, punct | print, |
| 875 | punct | print, upper | xdigit | print | alpha, |
| 876 | upper | xdigit | print | alpha, upper | xdigit | print | alpha, |
| 877 | upper | xdigit | print | alpha, upper | xdigit | print | alpha, |
| 878 | upper | xdigit | print | alpha, upper | print | alpha, |
| 879 | upper | print | alpha, upper | print | alpha, |
| 880 | upper | print | alpha, upper | print | alpha, |
| 881 | upper | print | alpha, upper | print | alpha, |
| 882 | upper | print | alpha, upper | print | alpha, |
| 883 | upper | print | alpha, upper | print | alpha, |
| 884 | upper | print | alpha, upper | print | alpha, |
| 885 | upper | print | alpha, upper | print | alpha, |
| 886 | upper | print | alpha, upper | print | alpha, |
| 887 | upper | print | alpha, upper | print | alpha, |
| 888 | upper | print | alpha, punct | print, |
| 889 | punct | print, punct | print, |
| 890 | punct | print, punct | print, |
| 891 | punct | print, lower | xdigit | print | alpha, |
| 892 | lower | xdigit | print | alpha, lower | xdigit | print | alpha, |
| 893 | lower | xdigit | print | alpha, lower | xdigit | print | alpha, |
| 894 | lower | xdigit | print | alpha, lower | print | alpha, |
| 895 | lower | print | alpha, lower | print | alpha, |
| 896 | lower | print | alpha, lower | print | alpha, |
| 897 | lower | print | alpha, lower | print | alpha, |
| 898 | lower | print | alpha, lower | print | alpha, |
| 899 | lower | print | alpha, lower | print | alpha, |
| 900 | lower | print | alpha, lower | print | alpha, |
| 901 | lower | print | alpha, lower | print | alpha, |
| 902 | lower | print | alpha, lower | print | alpha, |
| 903 | lower | print | alpha, lower | print | alpha, |
| 904 | lower | print | alpha, punct | print, |
| 905 | punct | print, punct | print, |
| 906 | punct | print, cntrl, |
| 907 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 908 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 909 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 910 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 911 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 912 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 913 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 914 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 |
| 915 | }; |
| 916 | // clang-format on |
| 917 | return builtin_table; |
| 918 | } |
| 919 | #else |
| 920 | const ctype<char>::mask* ctype<char>::classic_table() noexcept { |
| 921 | # if defined(__APPLE__) || defined(__FreeBSD__) |
| 922 | return _DefaultRuneLocale.__runetype; |
| 923 | # elif defined(__NetBSD__) |
| 924 | return _C_ctype_tab_ + 1; |
| 925 | # elif defined(__GLIBC__) |
| 926 | return __locale::__get_c_locale()->__ctype_b; |
| 927 | # elif defined(_WIN32) |
| 928 | return __pctype_func(); |
| 929 | # elif defined(__EMSCRIPTEN__) |
| 930 | return *__ctype_b_loc(); |
| 931 | # elif _LIBCPP_LIBC_NEWLIB |
| 932 | // Newlib has a 257-entry table in ctype_.c, where (char)0 starts at [1]. |
| 933 | return _ctype_ + 1; |
| 934 | # elif defined(_AIX) |
| 935 | return (const unsigned int*)__lc_ctype_ptr->obj->mask; |
| 936 | # elif defined(__MVS__) |
| 937 | # if defined(__NATIVE_ASCII_F) |
| 938 | return const_cast<const ctype<char>::mask*>(__OBJ_DATA(__lc_ctype_a)->mask); |
| 939 | # else |
| 940 | return const_cast<const ctype<char>::mask*>(__ctypec); |
| 941 | # endif |
| 942 | # else |
| 943 | // Platform not supported: abort so the person doing the port knows what to |
| 944 | // fix |
| 945 | # warning ctype<char>::classic_table() is not implemented |
| 946 | printf("ctype<char>::classic_table() is not implemented\n"); |
| 947 | abort(); |
| 948 | return nullptr; |
| 949 | # endif |
| 950 | } |
| 951 | #endif |
| 952 | |
| 953 | // template <> class ctype_byname<char> |
| 954 | |
| 955 | ctype_byname<char>::ctype_byname(const char* name, size_t refs) |
| 956 | : ctype<char>(0, false, refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: name, base: 0)) { |
| 957 | if (__l_ == 0) |
| 958 | std::__throw_runtime_error( |
| 959 | ("ctype_byname<char>::ctype_byname" |
| 960 | " failed to construct for "+ |
| 961 | string(name)) |
| 962 | .c_str()); |
| 963 | } |
| 964 | |
| 965 | ctype_byname<char>::ctype_byname(const string& name, size_t refs) |
| 966 | : ctype<char>(0, false, refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: name.c_str(), base: 0)) { |
| 967 | if (__l_ == 0) |
| 968 | std::__throw_runtime_error( |
| 969 | ("ctype_byname<char>::ctype_byname" |
| 970 | " failed to construct for "+ |
| 971 | name) |
| 972 | .c_str()); |
| 973 | } |
| 974 | |
| 975 | ctype_byname<char>::~ctype_byname() { __locale::__freelocale(loc: __l_); } |
| 976 | |
| 977 | char ctype_byname<char>::do_toupper(char_type c) const { |
| 978 | return static_cast<char>(__locale::__toupper(c: static_cast<unsigned char>(c), loc: __l_)); |
| 979 | } |
| 980 | |
| 981 | const char* ctype_byname<char>::do_toupper(char_type* low, const char_type* high) const { |
| 982 | for (; low != high; ++low) |
| 983 | *low = static_cast<char>(__locale::__toupper(c: static_cast<unsigned char>(*low), loc: __l_)); |
| 984 | return low; |
| 985 | } |
| 986 | |
| 987 | char ctype_byname<char>::do_tolower(char_type c) const { |
| 988 | return static_cast<char>(__locale::__tolower(c: static_cast<unsigned char>(c), loc: __l_)); |
| 989 | } |
| 990 | |
| 991 | const char* ctype_byname<char>::do_tolower(char_type* low, const char_type* high) const { |
| 992 | for (; low != high; ++low) |
| 993 | *low = static_cast<char>(__locale::__tolower(c: static_cast<unsigned char>(*low), loc: __l_)); |
| 994 | return low; |
| 995 | } |
| 996 | |
| 997 | // template <> class ctype_byname<wchar_t> |
| 998 | |
| 999 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 1000 | ctype_byname<wchar_t>::ctype_byname(const char* name, size_t refs) |
| 1001 | : ctype<wchar_t>(refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: name, base: 0)) { |
| 1002 | if (__l_ == 0) |
| 1003 | std::__throw_runtime_error( |
| 1004 | ("ctype_byname<wchar_t>::ctype_byname" |
| 1005 | " failed to construct for "+ |
| 1006 | string(name)) |
| 1007 | .c_str()); |
| 1008 | } |
| 1009 | |
| 1010 | ctype_byname<wchar_t>::ctype_byname(const string& name, size_t refs) |
| 1011 | : ctype<wchar_t>(refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: name.c_str(), base: 0)) { |
| 1012 | if (__l_ == 0) |
| 1013 | std::__throw_runtime_error( |
| 1014 | ("ctype_byname<wchar_t>::ctype_byname" |
| 1015 | " failed to construct for "+ |
| 1016 | name) |
| 1017 | .c_str()); |
| 1018 | } |
| 1019 | |
| 1020 | ctype_byname<wchar_t>::~ctype_byname() { __locale::__freelocale(loc: __l_); } |
| 1021 | |
| 1022 | bool ctype_byname<wchar_t>::do_is(mask m, char_type c) const { |
| 1023 | wint_t ch = static_cast<wint_t>(c); |
| 1024 | # ifdef _LIBCPP_WCTYPE_IS_MASK |
| 1025 | return static_cast<bool>(__locale::__iswctype(ch, m, __l_)); |
| 1026 | # else |
| 1027 | bool result = false; |
| 1028 | if ((m & space) == space) |
| 1029 | result |= (__locale::__iswspace(c: ch, loc: __l_) != 0); |
| 1030 | if ((m & print) == print) |
| 1031 | result |= (__locale::__iswprint(c: ch, loc: __l_) != 0); |
| 1032 | if ((m & cntrl) == cntrl) |
| 1033 | result |= (__locale::__iswcntrl(c: ch, loc: __l_) != 0); |
| 1034 | if ((m & upper) == upper) |
| 1035 | result |= (__locale::__iswupper(c: ch, loc: __l_) != 0); |
| 1036 | if ((m & lower) == lower) |
| 1037 | result |= (__locale::__iswlower(c: ch, loc: __l_) != 0); |
| 1038 | if ((m & alpha) == alpha) |
| 1039 | result |= (__locale::__iswalpha(c: ch, loc: __l_) != 0); |
| 1040 | if ((m & digit) == digit) |
| 1041 | result |= (__locale::__iswdigit(c: ch, loc: __l_) != 0); |
| 1042 | if ((m & punct) == punct) |
| 1043 | result |= (__locale::__iswpunct(c: ch, loc: __l_) != 0); |
| 1044 | if ((m & xdigit) == xdigit) |
| 1045 | result |= (__locale::__iswxdigit(c: ch, loc: __l_) != 0); |
| 1046 | if ((m & blank) == blank) |
| 1047 | result |= (__locale::__iswblank(c: ch, loc: __l_) != 0); |
| 1048 | return result; |
| 1049 | # endif |
| 1050 | } |
| 1051 | |
| 1052 | const wchar_t* ctype_byname<wchar_t>::do_is(const char_type* low, const char_type* high, mask* vec) const { |
| 1053 | for (; low != high; ++low, ++vec) { |
| 1054 | if (std::__libcpp_isascii(c: *low)) |
| 1055 | *vec = static_cast<mask>(ctype<char>::classic_table()[*low]); |
| 1056 | else { |
| 1057 | *vec = 0; |
| 1058 | wint_t ch = static_cast<wint_t>(*low); |
| 1059 | if (__locale::__iswspace(c: ch, loc: __l_)) |
| 1060 | *vec |= space; |
| 1061 | # ifndef _LIBCPP_CTYPE_MASK_IS_COMPOSITE_PRINT |
| 1062 | if (__locale::__iswprint(c: ch, loc: __l_)) |
| 1063 | *vec |= print; |
| 1064 | # endif |
| 1065 | if (__locale::__iswcntrl(c: ch, loc: __l_)) |
| 1066 | *vec |= cntrl; |
| 1067 | if (__locale::__iswupper(c: ch, loc: __l_)) |
| 1068 | *vec |= upper; |
| 1069 | if (__locale::__iswlower(c: ch, loc: __l_)) |
| 1070 | *vec |= lower; |
| 1071 | # ifndef _LIBCPP_CTYPE_MASK_IS_COMPOSITE_ALPHA |
| 1072 | if (__locale::__iswalpha(c: ch, loc: __l_)) |
| 1073 | *vec |= alpha; |
| 1074 | # endif |
| 1075 | if (__locale::__iswdigit(c: ch, loc: __l_)) |
| 1076 | *vec |= digit; |
| 1077 | if (__locale::__iswpunct(c: ch, loc: __l_)) |
| 1078 | *vec |= punct; |
| 1079 | # ifndef _LIBCPP_CTYPE_MASK_IS_COMPOSITE_XDIGIT |
| 1080 | if (__locale::__iswxdigit(c: ch, loc: __l_)) |
| 1081 | *vec |= xdigit; |
| 1082 | # endif |
| 1083 | if (__locale::__iswblank(c: ch, loc: __l_)) |
| 1084 | *vec |= blank; |
| 1085 | } |
| 1086 | } |
| 1087 | return low; |
| 1088 | } |
| 1089 | |
| 1090 | const wchar_t* ctype_byname<wchar_t>::do_scan_is(mask m, const char_type* low, const char_type* high) const { |
| 1091 | for (; low != high; ++low) { |
| 1092 | # ifdef _LIBCPP_WCTYPE_IS_MASK |
| 1093 | if (__locale::__iswctype(static_cast<wint_t>(*low), m, __l_)) |
| 1094 | break; |
| 1095 | # else |
| 1096 | wint_t ch = static_cast<wint_t>(*low); |
| 1097 | if ((m & space) == space && __locale::__iswspace(c: ch, loc: __l_)) |
| 1098 | break; |
| 1099 | if ((m & print) == print && __locale::__iswprint(c: ch, loc: __l_)) |
| 1100 | break; |
| 1101 | if ((m & cntrl) == cntrl && __locale::__iswcntrl(c: ch, loc: __l_)) |
| 1102 | break; |
| 1103 | if ((m & upper) == upper && __locale::__iswupper(c: ch, loc: __l_)) |
| 1104 | break; |
| 1105 | if ((m & lower) == lower && __locale::__iswlower(c: ch, loc: __l_)) |
| 1106 | break; |
| 1107 | if ((m & alpha) == alpha && __locale::__iswalpha(c: ch, loc: __l_)) |
| 1108 | break; |
| 1109 | if ((m & digit) == digit && __locale::__iswdigit(c: ch, loc: __l_)) |
| 1110 | break; |
| 1111 | if ((m & punct) == punct && __locale::__iswpunct(c: ch, loc: __l_)) |
| 1112 | break; |
| 1113 | if ((m & xdigit) == xdigit && __locale::__iswxdigit(c: ch, loc: __l_)) |
| 1114 | break; |
| 1115 | if ((m & blank) == blank && __locale::__iswblank(c: ch, loc: __l_)) |
| 1116 | break; |
| 1117 | # endif |
| 1118 | } |
| 1119 | return low; |
| 1120 | } |
| 1121 | |
| 1122 | const wchar_t* ctype_byname<wchar_t>::do_scan_not(mask m, const char_type* low, const char_type* high) const { |
| 1123 | for (; low != high; ++low) { |
| 1124 | wint_t ch = static_cast<wint_t>(*low); |
| 1125 | # ifdef _LIBCPP_WCTYPE_IS_MASK |
| 1126 | if (!__locale::__iswctype(ch, m, __l_)) |
| 1127 | break; |
| 1128 | # else |
| 1129 | if ((m & space) == space && __locale::__iswspace(c: ch, loc: __l_)) |
| 1130 | continue; |
| 1131 | if ((m & print) == print && __locale::__iswprint(c: ch, loc: __l_)) |
| 1132 | continue; |
| 1133 | if ((m & cntrl) == cntrl && __locale::__iswcntrl(c: ch, loc: __l_)) |
| 1134 | continue; |
| 1135 | if ((m & upper) == upper && __locale::__iswupper(c: ch, loc: __l_)) |
| 1136 | continue; |
| 1137 | if ((m & lower) == lower && __locale::__iswlower(c: ch, loc: __l_)) |
| 1138 | continue; |
| 1139 | if ((m & alpha) == alpha && __locale::__iswalpha(c: ch, loc: __l_)) |
| 1140 | continue; |
| 1141 | if ((m & digit) == digit && __locale::__iswdigit(c: ch, loc: __l_)) |
| 1142 | continue; |
| 1143 | if ((m & punct) == punct && __locale::__iswpunct(c: ch, loc: __l_)) |
| 1144 | continue; |
| 1145 | if ((m & xdigit) == xdigit && __locale::__iswxdigit(c: ch, loc: __l_)) |
| 1146 | continue; |
| 1147 | if ((m & blank) == blank && __locale::__iswblank(c: ch, loc: __l_)) |
| 1148 | continue; |
| 1149 | break; |
| 1150 | # endif |
| 1151 | } |
| 1152 | return low; |
| 1153 | } |
| 1154 | |
| 1155 | wchar_t ctype_byname<wchar_t>::do_toupper(char_type c) const { return __locale::__towupper(c: c, loc: __l_); } |
| 1156 | |
| 1157 | const wchar_t* ctype_byname<wchar_t>::do_toupper(char_type* low, const char_type* high) const { |
| 1158 | for (; low != high; ++low) |
| 1159 | *low = __locale::__towupper(c: *low, loc: __l_); |
| 1160 | return low; |
| 1161 | } |
| 1162 | |
| 1163 | wchar_t ctype_byname<wchar_t>::do_tolower(char_type c) const { return __locale::__towlower(c: c, loc: __l_); } |
| 1164 | |
| 1165 | const wchar_t* ctype_byname<wchar_t>::do_tolower(char_type* low, const char_type* high) const { |
| 1166 | for (; low != high; ++low) |
| 1167 | *low = __locale::__towlower(c: *low, loc: __l_); |
| 1168 | return low; |
| 1169 | } |
| 1170 | |
| 1171 | wchar_t ctype_byname<wchar_t>::do_widen(char c) const { return __locale::__btowc(c: c, loc: __l_); } |
| 1172 | |
| 1173 | const char* ctype_byname<wchar_t>::do_widen(const char* low, const char* high, char_type* dest) const { |
| 1174 | for (; low != high; ++low, ++dest) |
| 1175 | *dest = __locale::__btowc(c: *low, loc: __l_); |
| 1176 | return low; |
| 1177 | } |
| 1178 | |
| 1179 | char ctype_byname<wchar_t>::do_narrow(char_type c, char dfault) const { |
| 1180 | int r = __locale::__wctob(c: c, loc: __l_); |
| 1181 | return (r != EOF) ? static_cast<char>(r) : dfault; |
| 1182 | } |
| 1183 | |
| 1184 | const wchar_t* |
| 1185 | ctype_byname<wchar_t>::do_narrow(const char_type* low, const char_type* high, char dfault, char* dest) const { |
| 1186 | for (; low != high; ++low, ++dest) { |
| 1187 | int r = __locale::__wctob(c: *low, loc: __l_); |
| 1188 | *dest = (r != EOF) ? static_cast<char>(r) : dfault; |
| 1189 | } |
| 1190 | return low; |
| 1191 | } |
| 1192 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 1193 | |
| 1194 | // template <> class codecvt<char, char, mbstate_t> |
| 1195 | |
| 1196 | constinit locale::id codecvt<char, char, mbstate_t>::id; |
| 1197 | |
| 1198 | codecvt<char, char, mbstate_t>::~codecvt() {} |
| 1199 | |
| 1200 | codecvt<char, char, mbstate_t>::result codecvt<char, char, mbstate_t>::do_out( |
| 1201 | state_type&, |
| 1202 | const intern_type* frm, |
| 1203 | const intern_type*, |
| 1204 | const intern_type*& frm_nxt, |
| 1205 | extern_type* to, |
| 1206 | extern_type*, |
| 1207 | extern_type*& to_nxt) const { |
| 1208 | frm_nxt = frm; |
| 1209 | to_nxt = to; |
| 1210 | return noconv; |
| 1211 | } |
| 1212 | |
| 1213 | codecvt<char, char, mbstate_t>::result codecvt<char, char, mbstate_t>::do_in( |
| 1214 | state_type&, |
| 1215 | const extern_type* frm, |
| 1216 | const extern_type*, |
| 1217 | const extern_type*& frm_nxt, |
| 1218 | intern_type* to, |
| 1219 | intern_type*, |
| 1220 | intern_type*& to_nxt) const { |
| 1221 | frm_nxt = frm; |
| 1222 | to_nxt = to; |
| 1223 | return noconv; |
| 1224 | } |
| 1225 | |
| 1226 | codecvt<char, char, mbstate_t>::result |
| 1227 | codecvt<char, char, mbstate_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 1228 | to_nxt = to; |
| 1229 | return noconv; |
| 1230 | } |
| 1231 | |
| 1232 | int codecvt<char, char, mbstate_t>::do_encoding() const noexcept { return 1; } |
| 1233 | |
| 1234 | bool codecvt<char, char, mbstate_t>::do_always_noconv() const noexcept { return true; } |
| 1235 | |
| 1236 | int codecvt<char, char, mbstate_t>::do_length( |
| 1237 | state_type&, const extern_type* frm, const extern_type* end, size_t mx) const { |
| 1238 | return static_cast<int>(min<size_t>(a: mx, b: static_cast<size_t>(end - frm))); |
| 1239 | } |
| 1240 | |
| 1241 | int codecvt<char, char, mbstate_t>::do_max_length() const noexcept { return 1; } |
| 1242 | |
| 1243 | // template <> class codecvt<wchar_t, char, mbstate_t> |
| 1244 | |
| 1245 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 1246 | constinit locale::id codecvt<wchar_t, char, mbstate_t>::id; |
| 1247 | |
| 1248 | codecvt<wchar_t, char, mbstate_t>::codecvt(size_t refs) : locale::facet(refs), __l_(__locale::__get_c_locale()) {} |
| 1249 | |
| 1250 | codecvt<wchar_t, char, mbstate_t>::codecvt(const char* nm, size_t refs) |
| 1251 | : locale::facet(refs), __l_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: nm, base: 0)) { |
| 1252 | if (__l_ == 0) |
| 1253 | std::__throw_runtime_error( |
| 1254 | ("codecvt_byname<wchar_t, char, mbstate_t>::codecvt_byname" |
| 1255 | " failed to construct for "+ |
| 1256 | string(nm)) |
| 1257 | .c_str()); |
| 1258 | } |
| 1259 | |
| 1260 | codecvt<wchar_t, char, mbstate_t>::~codecvt() { |
| 1261 | if (__l_ != __locale::__get_c_locale()) |
| 1262 | __locale::__freelocale(loc: __l_); |
| 1263 | } |
| 1264 | |
| 1265 | codecvt<wchar_t, char, mbstate_t>::result codecvt<wchar_t, char, mbstate_t>::do_out( |
| 1266 | state_type& st, |
| 1267 | const intern_type* frm, |
| 1268 | const intern_type* frm_end, |
| 1269 | const intern_type*& frm_nxt, |
| 1270 | extern_type* to, |
| 1271 | extern_type* to_end, |
| 1272 | extern_type*& to_nxt) const { |
| 1273 | // look for first internal null in frm |
| 1274 | const intern_type* fend = frm; |
| 1275 | for (; fend != frm_end; ++fend) |
| 1276 | if (*fend == 0) |
| 1277 | break; |
| 1278 | // loop over all null-terminated sequences in frm |
| 1279 | to_nxt = to; |
| 1280 | for (frm_nxt = frm; frm != frm_end && to != to_end; frm = frm_nxt, to = to_nxt) { |
| 1281 | // save state in case it is needed to recover to_nxt on error |
| 1282 | mbstate_t save_state = st; |
| 1283 | size_t n = __locale::__wcsnrtombs( |
| 1284 | dest: to, src: &frm_nxt, nwc: static_cast<size_t>(fend - frm), len: static_cast<size_t>(to_end - to), ps: &st, loc: __l_); |
| 1285 | if (n == size_t(-1)) { |
| 1286 | // need to recover to_nxt |
| 1287 | for (to_nxt = to; frm != frm_nxt; ++frm) { |
| 1288 | n = __locale::__wcrtomb(s: to_nxt, wc: *frm, ps: &save_state, loc: __l_); |
| 1289 | if (n == size_t(-1)) |
| 1290 | break; |
| 1291 | to_nxt += n; |
| 1292 | } |
| 1293 | frm_nxt = frm; |
| 1294 | return error; |
| 1295 | } |
| 1296 | if (n == 0) |
| 1297 | return partial; |
| 1298 | to_nxt += n; |
| 1299 | if (to_nxt == to_end) |
| 1300 | break; |
| 1301 | if (fend != frm_end) // set up next null terminated sequence |
| 1302 | { |
| 1303 | // Try to write the terminating null |
| 1304 | extern_type tmp[MB_LEN_MAX]; |
| 1305 | n = __locale::__wcrtomb(s: tmp, wc: intern_type(), ps: &st, loc: __l_); |
| 1306 | if (n == size_t(-1)) // on error |
| 1307 | return error; |
| 1308 | if (n > static_cast<size_t>(to_end - to_nxt)) // is there room? |
| 1309 | return partial; |
| 1310 | for (extern_type* p = tmp; n; --n) // write it |
| 1311 | *to_nxt++ = *p++; |
| 1312 | ++frm_nxt; |
| 1313 | // look for next null in frm |
| 1314 | for (fend = frm_nxt; fend != frm_end; ++fend) |
| 1315 | if (*fend == 0) |
| 1316 | break; |
| 1317 | } |
| 1318 | } |
| 1319 | return frm_nxt == frm_end ? ok : partial; |
| 1320 | } |
| 1321 | |
| 1322 | codecvt<wchar_t, char, mbstate_t>::result codecvt<wchar_t, char, mbstate_t>::do_in( |
| 1323 | state_type& st, |
| 1324 | const extern_type* frm, |
| 1325 | const extern_type* frm_end, |
| 1326 | const extern_type*& frm_nxt, |
| 1327 | intern_type* to, |
| 1328 | intern_type* to_end, |
| 1329 | intern_type*& to_nxt) const { |
| 1330 | // look for first internal null in frm |
| 1331 | const extern_type* fend = frm; |
| 1332 | for (; fend != frm_end; ++fend) |
| 1333 | if (*fend == 0) |
| 1334 | break; |
| 1335 | // loop over all null-terminated sequences in frm |
| 1336 | to_nxt = to; |
| 1337 | for (frm_nxt = frm; frm != frm_end && to != to_end; frm = frm_nxt, to = to_nxt) { |
| 1338 | // save state in case it is needed to recover to_nxt on error |
| 1339 | mbstate_t save_state = st; |
| 1340 | size_t n = __locale::__mbsnrtowcs( |
| 1341 | dest: to, src: &frm_nxt, nms: static_cast<size_t>(fend - frm), len: static_cast<size_t>(to_end - to), ps: &st, loc: __l_); |
| 1342 | if (n == size_t(-1)) { |
| 1343 | // need to recover to_nxt |
| 1344 | for (to_nxt = to; frm != frm_nxt; ++to_nxt) { |
| 1345 | n = __locale::__mbrtowc(pwc: to_nxt, s: frm, n: static_cast<size_t>(fend - frm), ps: &save_state, loc: __l_); |
| 1346 | switch (n) { |
| 1347 | case 0: |
| 1348 | ++frm; |
| 1349 | break; |
| 1350 | case size_t(-1): |
| 1351 | frm_nxt = frm; |
| 1352 | return error; |
| 1353 | case size_t(-2): |
| 1354 | frm_nxt = frm; |
| 1355 | return partial; |
| 1356 | default: |
| 1357 | frm += n; |
| 1358 | break; |
| 1359 | } |
| 1360 | } |
| 1361 | frm_nxt = frm; |
| 1362 | return frm_nxt == frm_end ? ok : partial; |
| 1363 | } |
| 1364 | if (n == size_t(-1)) |
| 1365 | return error; |
| 1366 | to_nxt += n; |
| 1367 | if (to_nxt == to_end) |
| 1368 | break; |
| 1369 | if (fend != frm_end) // set up next null terminated sequence |
| 1370 | { |
| 1371 | // Try to write the terminating null |
| 1372 | n = __locale::__mbrtowc(pwc: to_nxt, s: frm_nxt, n: 1, ps: &st, loc: __l_); |
| 1373 | if (n != 0) // on error |
| 1374 | return error; |
| 1375 | ++to_nxt; |
| 1376 | ++frm_nxt; |
| 1377 | // look for next null in frm |
| 1378 | for (fend = frm_nxt; fend != frm_end; ++fend) |
| 1379 | if (*fend == 0) |
| 1380 | break; |
| 1381 | } |
| 1382 | } |
| 1383 | return frm_nxt == frm_end ? ok : partial; |
| 1384 | } |
| 1385 | |
| 1386 | codecvt<wchar_t, char, mbstate_t>::result codecvt<wchar_t, char, mbstate_t>::do_unshift( |
| 1387 | state_type& st, extern_type* to, extern_type* to_end, extern_type*& to_nxt) const { |
| 1388 | to_nxt = to; |
| 1389 | extern_type tmp[MB_LEN_MAX]; |
| 1390 | size_t n = __locale::__wcrtomb(s: tmp, wc: intern_type(), ps: &st, loc: __l_); |
| 1391 | if (n == size_t(-1) || n == 0) // on error |
| 1392 | return error; |
| 1393 | --n; |
| 1394 | if (n > static_cast<size_t>(to_end - to_nxt)) // is there room? |
| 1395 | return partial; |
| 1396 | for (extern_type* p = tmp; n; --n) // write it |
| 1397 | *to_nxt++ = *p++; |
| 1398 | return ok; |
| 1399 | } |
| 1400 | |
| 1401 | int codecvt<wchar_t, char, mbstate_t>::do_encoding() const noexcept { |
| 1402 | if (__locale::__mbtowc(pwc: nullptr, pmb: nullptr, MB_LEN_MAX, loc: __l_) != 0) |
| 1403 | return -1; |
| 1404 | |
| 1405 | // stateless encoding |
| 1406 | if (__l_ == 0 || __locale::__mb_len_max(loc: __l_) == 1) // there are no known constant length encodings |
| 1407 | return 1; // which take more than 1 char to form a wchar_t |
| 1408 | return 0; |
| 1409 | } |
| 1410 | |
| 1411 | bool codecvt<wchar_t, char, mbstate_t>::do_always_noconv() const noexcept { return false; } |
| 1412 | |
| 1413 | int codecvt<wchar_t, char, mbstate_t>::do_length( |
| 1414 | state_type& st, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 1415 | int nbytes = 0; |
| 1416 | for (size_t nwchar_t = 0; nwchar_t < mx && frm != frm_end; ++nwchar_t) { |
| 1417 | size_t n = __locale::__mbrlen(s: frm, n: static_cast<size_t>(frm_end - frm), ps: &st, loc: __l_); |
| 1418 | switch (n) { |
| 1419 | case 0: |
| 1420 | ++nbytes; |
| 1421 | ++frm; |
| 1422 | break; |
| 1423 | case size_t(-1): |
| 1424 | case size_t(-2): |
| 1425 | return nbytes; |
| 1426 | default: |
| 1427 | nbytes += n; |
| 1428 | frm += n; |
| 1429 | break; |
| 1430 | } |
| 1431 | } |
| 1432 | return nbytes; |
| 1433 | } |
| 1434 | |
| 1435 | int codecvt<wchar_t, char, mbstate_t>::do_max_length() const noexcept { |
| 1436 | return __l_ == 0 ? 1 : static_cast<int>(__locale::__mb_len_max(loc: __l_)); |
| 1437 | } |
| 1438 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 1439 | |
| 1440 | // Valid UTF ranges |
| 1441 | // UTF-32 UTF-16 UTF-8 # of code points |
| 1442 | // first second first second third fourth |
| 1443 | // 000000 - 00007F 0000 - 007F 00 - 7F 127 |
| 1444 | // 000080 - 0007FF 0080 - 07FF C2 - DF, 80 - BF 1920 |
| 1445 | // 000800 - 000FFF 0800 - 0FFF E0 - E0, A0 - BF, 80 - BF 2048 |
| 1446 | // 001000 - 00CFFF 1000 - CFFF E1 - EC, 80 - BF, 80 - BF 49152 |
| 1447 | // 00D000 - 00D7FF D000 - D7FF ED - ED, 80 - 9F, 80 - BF 2048 |
| 1448 | // 00D800 - 00DFFF invalid |
| 1449 | // 00E000 - 00FFFF E000 - FFFF EE - EF, 80 - BF, 80 - BF 8192 |
| 1450 | // 010000 - 03FFFF D800 - D8BF, DC00 - DFFF F0 - F0, 90 - BF, 80 - BF, 80 - BF 196608 |
| 1451 | // 040000 - 0FFFFF D8C0 - DBBF, DC00 - DFFF F1 - F3, 80 - BF, 80 - BF, 80 - BF 786432 |
| 1452 | // 100000 - 10FFFF DBC0 - DBFF, DC00 - DFFF F4 - F4, 80 - 8F, 80 - BF, 80 - BF 65536 |
| 1453 | |
| 1454 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 1455 | static codecvt_base::result utf16_to_utf8( |
| 1456 | const uint16_t* frm, |
| 1457 | const uint16_t* frm_end, |
| 1458 | const uint16_t*& frm_nxt, |
| 1459 | uint8_t* to, |
| 1460 | uint8_t* to_end, |
| 1461 | uint8_t*& to_nxt, |
| 1462 | unsigned long Maxcode = 0x10FFFF, |
| 1463 | codecvt_mode mode = codecvt_mode(0)) { |
| 1464 | frm_nxt = frm; |
| 1465 | to_nxt = to; |
| 1466 | if (mode & generate_header) { |
| 1467 | if (to_end - to_nxt < 3) |
| 1468 | return codecvt_base::partial; |
| 1469 | *to_nxt++ = static_cast<uint8_t>(0xEF); |
| 1470 | *to_nxt++ = static_cast<uint8_t>(0xBB); |
| 1471 | *to_nxt++ = static_cast<uint8_t>(0xBF); |
| 1472 | } |
| 1473 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 1474 | uint16_t wc1 = *frm_nxt; |
| 1475 | if (wc1 > Maxcode) |
| 1476 | return codecvt_base::error; |
| 1477 | if (wc1 < 0x0080) { |
| 1478 | if (to_end - to_nxt < 1) |
| 1479 | return codecvt_base::partial; |
| 1480 | *to_nxt++ = static_cast<uint8_t>(wc1); |
| 1481 | } else if (wc1 < 0x0800) { |
| 1482 | if (to_end - to_nxt < 2) |
| 1483 | return codecvt_base::partial; |
| 1484 | *to_nxt++ = static_cast<uint8_t>(0xC0 | (wc1 >> 6)); |
| 1485 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc1 & 0x03F)); |
| 1486 | } else if (wc1 < 0xD800) { |
| 1487 | if (to_end - to_nxt < 3) |
| 1488 | return codecvt_base::partial; |
| 1489 | *to_nxt++ = static_cast<uint8_t>(0xE0 | (wc1 >> 12)); |
| 1490 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc1 & 0x0FC0) >> 6)); |
| 1491 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc1 & 0x003F)); |
| 1492 | } else if (wc1 < 0xDC00) { |
| 1493 | if (frm_end - frm_nxt < 2) |
| 1494 | return codecvt_base::partial; |
| 1495 | uint16_t wc2 = frm_nxt[1]; |
| 1496 | if ((wc2 & 0xFC00) != 0xDC00) |
| 1497 | return codecvt_base::error; |
| 1498 | if (to_end - to_nxt < 4) |
| 1499 | return codecvt_base::partial; |
| 1500 | if (((((wc1 & 0x03C0UL) >> 6) + 1) << 16) + ((wc1 & 0x003FUL) << 10) + (wc2 & 0x03FF) > Maxcode) |
| 1501 | return codecvt_base::error; |
| 1502 | ++frm_nxt; |
| 1503 | uint8_t z = ((wc1 & 0x03C0) >> 6) + 1; |
| 1504 | *to_nxt++ = static_cast<uint8_t>(0xF0 | (z >> 2)); |
| 1505 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((z & 0x03) << 4) | ((wc1 & 0x003C) >> 2)); |
| 1506 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc1 & 0x0003) << 4) | ((wc2 & 0x03C0) >> 6)); |
| 1507 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc2 & 0x003F)); |
| 1508 | } else if (wc1 < 0xE000) { |
| 1509 | return codecvt_base::error; |
| 1510 | } else { |
| 1511 | if (to_end - to_nxt < 3) |
| 1512 | return codecvt_base::partial; |
| 1513 | *to_nxt++ = static_cast<uint8_t>(0xE0 | (wc1 >> 12)); |
| 1514 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc1 & 0x0FC0) >> 6)); |
| 1515 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc1 & 0x003F)); |
| 1516 | } |
| 1517 | } |
| 1518 | return codecvt_base::ok; |
| 1519 | } |
| 1520 | |
| 1521 | static codecvt_base::result utf16_to_utf8( |
| 1522 | const uint32_t* frm, |
| 1523 | const uint32_t* frm_end, |
| 1524 | const uint32_t*& frm_nxt, |
| 1525 | uint8_t* to, |
| 1526 | uint8_t* to_end, |
| 1527 | uint8_t*& to_nxt, |
| 1528 | unsigned long Maxcode = 0x10FFFF, |
| 1529 | codecvt_mode mode = codecvt_mode(0)) { |
| 1530 | frm_nxt = frm; |
| 1531 | to_nxt = to; |
| 1532 | if (mode & generate_header) { |
| 1533 | if (to_end - to_nxt < 3) |
| 1534 | return codecvt_base::partial; |
| 1535 | *to_nxt++ = static_cast<uint8_t>(0xEF); |
| 1536 | *to_nxt++ = static_cast<uint8_t>(0xBB); |
| 1537 | *to_nxt++ = static_cast<uint8_t>(0xBF); |
| 1538 | } |
| 1539 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 1540 | uint16_t wc1 = static_cast<uint16_t>(*frm_nxt); |
| 1541 | if (wc1 > Maxcode) |
| 1542 | return codecvt_base::error; |
| 1543 | if (wc1 < 0x0080) { |
| 1544 | if (to_end - to_nxt < 1) |
| 1545 | return codecvt_base::partial; |
| 1546 | *to_nxt++ = static_cast<uint8_t>(wc1); |
| 1547 | } else if (wc1 < 0x0800) { |
| 1548 | if (to_end - to_nxt < 2) |
| 1549 | return codecvt_base::partial; |
| 1550 | *to_nxt++ = static_cast<uint8_t>(0xC0 | (wc1 >> 6)); |
| 1551 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc1 & 0x03F)); |
| 1552 | } else if (wc1 < 0xD800) { |
| 1553 | if (to_end - to_nxt < 3) |
| 1554 | return codecvt_base::partial; |
| 1555 | *to_nxt++ = static_cast<uint8_t>(0xE0 | (wc1 >> 12)); |
| 1556 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc1 & 0x0FC0) >> 6)); |
| 1557 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc1 & 0x003F)); |
| 1558 | } else if (wc1 < 0xDC00) { |
| 1559 | if (frm_end - frm_nxt < 2) |
| 1560 | return codecvt_base::partial; |
| 1561 | uint16_t wc2 = static_cast<uint16_t>(frm_nxt[1]); |
| 1562 | if ((wc2 & 0xFC00) != 0xDC00) |
| 1563 | return codecvt_base::error; |
| 1564 | if (to_end - to_nxt < 4) |
| 1565 | return codecvt_base::partial; |
| 1566 | if (((((wc1 & 0x03C0UL) >> 6) + 1) << 16) + ((wc1 & 0x003FUL) << 10) + (wc2 & 0x03FF) > Maxcode) |
| 1567 | return codecvt_base::error; |
| 1568 | ++frm_nxt; |
| 1569 | uint8_t z = ((wc1 & 0x03C0) >> 6) + 1; |
| 1570 | *to_nxt++ = static_cast<uint8_t>(0xF0 | (z >> 2)); |
| 1571 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((z & 0x03) << 4) | ((wc1 & 0x003C) >> 2)); |
| 1572 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc1 & 0x0003) << 4) | ((wc2 & 0x03C0) >> 6)); |
| 1573 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc2 & 0x003F)); |
| 1574 | } else if (wc1 < 0xE000) { |
| 1575 | return codecvt_base::error; |
| 1576 | } else { |
| 1577 | if (to_end - to_nxt < 3) |
| 1578 | return codecvt_base::partial; |
| 1579 | *to_nxt++ = static_cast<uint8_t>(0xE0 | (wc1 >> 12)); |
| 1580 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc1 & 0x0FC0) >> 6)); |
| 1581 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc1 & 0x003F)); |
| 1582 | } |
| 1583 | } |
| 1584 | return codecvt_base::ok; |
| 1585 | } |
| 1586 | |
| 1587 | static codecvt_base::result utf8_to_utf16( |
| 1588 | const uint8_t* frm, |
| 1589 | const uint8_t* frm_end, |
| 1590 | const uint8_t*& frm_nxt, |
| 1591 | uint16_t* to, |
| 1592 | uint16_t* to_end, |
| 1593 | uint16_t*& to_nxt, |
| 1594 | unsigned long Maxcode = 0x10FFFF, |
| 1595 | codecvt_mode mode = codecvt_mode(0)) { |
| 1596 | frm_nxt = frm; |
| 1597 | to_nxt = to; |
| 1598 | if (mode & consume_header) { |
| 1599 | if (frm_end - frm_nxt >= 3 && frm_nxt[0] == 0xEF && frm_nxt[1] == 0xBB && frm_nxt[2] == 0xBF) |
| 1600 | frm_nxt += 3; |
| 1601 | } |
| 1602 | for (; frm_nxt < frm_end && to_nxt < to_end; ++to_nxt) { |
| 1603 | uint8_t c1 = *frm_nxt; |
| 1604 | if (c1 > Maxcode) |
| 1605 | return codecvt_base::error; |
| 1606 | if (c1 < 0x80) { |
| 1607 | *to_nxt = static_cast<uint16_t>(c1); |
| 1608 | ++frm_nxt; |
| 1609 | } else if (c1 < 0xC2) { |
| 1610 | return codecvt_base::error; |
| 1611 | } else if (c1 < 0xE0) { |
| 1612 | if (frm_end - frm_nxt < 2) |
| 1613 | return codecvt_base::partial; |
| 1614 | uint8_t c2 = frm_nxt[1]; |
| 1615 | if ((c2 & 0xC0) != 0x80) |
| 1616 | return codecvt_base::error; |
| 1617 | uint16_t t = static_cast<uint16_t>(((c1 & 0x1F) << 6) | (c2 & 0x3F)); |
| 1618 | if (t > Maxcode) |
| 1619 | return codecvt_base::error; |
| 1620 | *to_nxt = t; |
| 1621 | frm_nxt += 2; |
| 1622 | } else if (c1 < 0xF0) { |
| 1623 | if (frm_end - frm_nxt < 2) |
| 1624 | return codecvt_base::partial; |
| 1625 | uint8_t c2 = frm_nxt[1]; |
| 1626 | switch (c1) { |
| 1627 | case 0xE0: |
| 1628 | if ((c2 & 0xE0) != 0xA0) |
| 1629 | return codecvt_base::error; |
| 1630 | break; |
| 1631 | case 0xED: |
| 1632 | if ((c2 & 0xE0) != 0x80) |
| 1633 | return codecvt_base::error; |
| 1634 | break; |
| 1635 | default: |
| 1636 | if ((c2 & 0xC0) != 0x80) |
| 1637 | return codecvt_base::error; |
| 1638 | break; |
| 1639 | } |
| 1640 | if (frm_end - frm_nxt < 3) |
| 1641 | return codecvt_base::partial; |
| 1642 | uint8_t c3 = frm_nxt[2]; |
| 1643 | if ((c3 & 0xC0) != 0x80) |
| 1644 | return codecvt_base::error; |
| 1645 | uint16_t t = static_cast<uint16_t>(((c1 & 0x0F) << 12) | ((c2 & 0x3F) << 6) | (c3 & 0x3F)); |
| 1646 | if (t > Maxcode) |
| 1647 | return codecvt_base::error; |
| 1648 | *to_nxt = t; |
| 1649 | frm_nxt += 3; |
| 1650 | } else if (c1 < 0xF5) { |
| 1651 | if (frm_end - frm_nxt < 2) |
| 1652 | return codecvt_base::partial; |
| 1653 | uint8_t c2 = frm_nxt[1]; |
| 1654 | switch (c1) { |
| 1655 | case 0xF0: |
| 1656 | if (!(0x90 <= c2 && c2 <= 0xBF)) |
| 1657 | return codecvt_base::error; |
| 1658 | break; |
| 1659 | case 0xF4: |
| 1660 | if ((c2 & 0xF0) != 0x80) |
| 1661 | return codecvt_base::error; |
| 1662 | break; |
| 1663 | default: |
| 1664 | if ((c2 & 0xC0) != 0x80) |
| 1665 | return codecvt_base::error; |
| 1666 | break; |
| 1667 | } |
| 1668 | if (frm_end - frm_nxt < 3) |
| 1669 | return codecvt_base::partial; |
| 1670 | uint8_t c3 = frm_nxt[2]; |
| 1671 | if ((c3 & 0xC0) != 0x80) |
| 1672 | return codecvt_base::error; |
| 1673 | if (frm_end - frm_nxt < 4) |
| 1674 | return codecvt_base::partial; |
| 1675 | uint8_t c4 = frm_nxt[3]; |
| 1676 | if ((c4 & 0xC0) != 0x80) |
| 1677 | return codecvt_base::error; |
| 1678 | if (to_end - to_nxt < 2) |
| 1679 | return codecvt_base::partial; |
| 1680 | if ((((c1 & 7UL) << 18) + ((c2 & 0x3FUL) << 12) + ((c3 & 0x3FUL) << 6) + (c4 & 0x3F)) > Maxcode) |
| 1681 | return codecvt_base::error; |
| 1682 | *to_nxt = static_cast<uint16_t>( |
| 1683 | 0xD800 | (((((c1 & 0x07) << 2) | ((c2 & 0x30) >> 4)) - 1) << 6) | ((c2 & 0x0F) << 2) | ((c3 & 0x30) >> 4)); |
| 1684 | *++to_nxt = static_cast<uint16_t>(0xDC00 | ((c3 & 0x0F) << 6) | (c4 & 0x3F)); |
| 1685 | frm_nxt += 4; |
| 1686 | } else { |
| 1687 | return codecvt_base::error; |
| 1688 | } |
| 1689 | } |
| 1690 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 1691 | } |
| 1692 | |
| 1693 | static codecvt_base::result utf8_to_utf16( |
| 1694 | const uint8_t* frm, |
| 1695 | const uint8_t* frm_end, |
| 1696 | const uint8_t*& frm_nxt, |
| 1697 | uint32_t* to, |
| 1698 | uint32_t* to_end, |
| 1699 | uint32_t*& to_nxt, |
| 1700 | unsigned long Maxcode = 0x10FFFF, |
| 1701 | codecvt_mode mode = codecvt_mode(0)) { |
| 1702 | frm_nxt = frm; |
| 1703 | to_nxt = to; |
| 1704 | if (mode & consume_header) { |
| 1705 | if (frm_end - frm_nxt >= 3 && frm_nxt[0] == 0xEF && frm_nxt[1] == 0xBB && frm_nxt[2] == 0xBF) |
| 1706 | frm_nxt += 3; |
| 1707 | } |
| 1708 | for (; frm_nxt < frm_end && to_nxt < to_end; ++to_nxt) { |
| 1709 | uint8_t c1 = *frm_nxt; |
| 1710 | if (c1 > Maxcode) |
| 1711 | return codecvt_base::error; |
| 1712 | if (c1 < 0x80) { |
| 1713 | *to_nxt = static_cast<uint32_t>(c1); |
| 1714 | ++frm_nxt; |
| 1715 | } else if (c1 < 0xC2) { |
| 1716 | return codecvt_base::error; |
| 1717 | } else if (c1 < 0xE0) { |
| 1718 | if (frm_end - frm_nxt < 2) |
| 1719 | return codecvt_base::partial; |
| 1720 | uint8_t c2 = frm_nxt[1]; |
| 1721 | if ((c2 & 0xC0) != 0x80) |
| 1722 | return codecvt_base::error; |
| 1723 | uint16_t t = static_cast<uint16_t>(((c1 & 0x1F) << 6) | (c2 & 0x3F)); |
| 1724 | if (t > Maxcode) |
| 1725 | return codecvt_base::error; |
| 1726 | *to_nxt = static_cast<uint32_t>(t); |
| 1727 | frm_nxt += 2; |
| 1728 | } else if (c1 < 0xF0) { |
| 1729 | if (frm_end - frm_nxt < 2) |
| 1730 | return codecvt_base::partial; |
| 1731 | uint8_t c2 = frm_nxt[1]; |
| 1732 | switch (c1) { |
| 1733 | case 0xE0: |
| 1734 | if ((c2 & 0xE0) != 0xA0) |
| 1735 | return codecvt_base::error; |
| 1736 | break; |
| 1737 | case 0xED: |
| 1738 | if ((c2 & 0xE0) != 0x80) |
| 1739 | return codecvt_base::error; |
| 1740 | break; |
| 1741 | default: |
| 1742 | if ((c2 & 0xC0) != 0x80) |
| 1743 | return codecvt_base::error; |
| 1744 | break; |
| 1745 | } |
| 1746 | if (frm_end - frm_nxt < 3) |
| 1747 | return codecvt_base::partial; |
| 1748 | uint8_t c3 = frm_nxt[2]; |
| 1749 | if ((c3 & 0xC0) != 0x80) |
| 1750 | return codecvt_base::error; |
| 1751 | uint16_t t = static_cast<uint16_t>(((c1 & 0x0F) << 12) | ((c2 & 0x3F) << 6) | (c3 & 0x3F)); |
| 1752 | if (t > Maxcode) |
| 1753 | return codecvt_base::error; |
| 1754 | *to_nxt = static_cast<uint32_t>(t); |
| 1755 | frm_nxt += 3; |
| 1756 | } else if (c1 < 0xF5) { |
| 1757 | if (frm_end - frm_nxt < 2) |
| 1758 | return codecvt_base::partial; |
| 1759 | uint8_t c2 = frm_nxt[1]; |
| 1760 | switch (c1) { |
| 1761 | case 0xF0: |
| 1762 | if (!(0x90 <= c2 && c2 <= 0xBF)) |
| 1763 | return codecvt_base::error; |
| 1764 | break; |
| 1765 | case 0xF4: |
| 1766 | if ((c2 & 0xF0) != 0x80) |
| 1767 | return codecvt_base::error; |
| 1768 | break; |
| 1769 | default: |
| 1770 | if ((c2 & 0xC0) != 0x80) |
| 1771 | return codecvt_base::error; |
| 1772 | break; |
| 1773 | } |
| 1774 | if (frm_end - frm_nxt < 3) |
| 1775 | return codecvt_base::partial; |
| 1776 | uint8_t c3 = frm_nxt[2]; |
| 1777 | if ((c3 & 0xC0) != 0x80) |
| 1778 | return codecvt_base::error; |
| 1779 | if (frm_end - frm_nxt < 4) |
| 1780 | return codecvt_base::partial; |
| 1781 | uint8_t c4 = frm_nxt[3]; |
| 1782 | if ((c4 & 0xC0) != 0x80) |
| 1783 | return codecvt_base::error; |
| 1784 | if (to_end - to_nxt < 2) |
| 1785 | return codecvt_base::partial; |
| 1786 | if ((((c1 & 7UL) << 18) + ((c2 & 0x3FUL) << 12) + ((c3 & 0x3FUL) << 6) + (c4 & 0x3F)) > Maxcode) |
| 1787 | return codecvt_base::error; |
| 1788 | *to_nxt = static_cast<uint32_t>( |
| 1789 | 0xD800 | (((((c1 & 0x07) << 2) | ((c2 & 0x30) >> 4)) - 1) << 6) | ((c2 & 0x0F) << 2) | ((c3 & 0x30) >> 4)); |
| 1790 | *++to_nxt = static_cast<uint32_t>(0xDC00 | ((c3 & 0x0F) << 6) | (c4 & 0x3F)); |
| 1791 | frm_nxt += 4; |
| 1792 | } else { |
| 1793 | return codecvt_base::error; |
| 1794 | } |
| 1795 | } |
| 1796 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 1797 | } |
| 1798 | |
| 1799 | static int utf8_to_utf16_length( |
| 1800 | const uint8_t* frm, |
| 1801 | const uint8_t* frm_end, |
| 1802 | size_t mx, |
| 1803 | unsigned long Maxcode = 0x10FFFF, |
| 1804 | codecvt_mode mode = codecvt_mode(0)) { |
| 1805 | const uint8_t* frm_nxt = frm; |
| 1806 | if (mode & consume_header) { |
| 1807 | if (frm_end - frm_nxt >= 3 && frm_nxt[0] == 0xEF && frm_nxt[1] == 0xBB && frm_nxt[2] == 0xBF) |
| 1808 | frm_nxt += 3; |
| 1809 | } |
| 1810 | for (size_t nchar16_t = 0; frm_nxt < frm_end && nchar16_t < mx; ++nchar16_t) { |
| 1811 | uint8_t c1 = *frm_nxt; |
| 1812 | if (c1 > Maxcode) |
| 1813 | break; |
| 1814 | if (c1 < 0x80) { |
| 1815 | ++frm_nxt; |
| 1816 | } else if (c1 < 0xC2) { |
| 1817 | break; |
| 1818 | } else if (c1 < 0xE0) { |
| 1819 | if ((frm_end - frm_nxt < 2) || (frm_nxt[1] & 0xC0) != 0x80) |
| 1820 | break; |
| 1821 | uint16_t t = static_cast<uint16_t>(((c1 & 0x1F) << 6) | (frm_nxt[1] & 0x3F)); |
| 1822 | if (t > Maxcode) |
| 1823 | break; |
| 1824 | frm_nxt += 2; |
| 1825 | } else if (c1 < 0xF0) { |
| 1826 | if (frm_end - frm_nxt < 3) |
| 1827 | break; |
| 1828 | uint8_t c2 = frm_nxt[1]; |
| 1829 | uint8_t c3 = frm_nxt[2]; |
| 1830 | switch (c1) { |
| 1831 | case 0xE0: |
| 1832 | if ((c2 & 0xE0) != 0xA0) |
| 1833 | return static_cast<int>(frm_nxt - frm); |
| 1834 | break; |
| 1835 | case 0xED: |
| 1836 | if ((c2 & 0xE0) != 0x80) |
| 1837 | return static_cast<int>(frm_nxt - frm); |
| 1838 | break; |
| 1839 | default: |
| 1840 | if ((c2 & 0xC0) != 0x80) |
| 1841 | return static_cast<int>(frm_nxt - frm); |
| 1842 | break; |
| 1843 | } |
| 1844 | if ((c3 & 0xC0) != 0x80) |
| 1845 | break; |
| 1846 | if ((((c1 & 0x0Fu) << 12) | ((c2 & 0x3Fu) << 6) | (c3 & 0x3Fu)) > Maxcode) |
| 1847 | break; |
| 1848 | frm_nxt += 3; |
| 1849 | } else if (c1 < 0xF5) { |
| 1850 | if (frm_end - frm_nxt < 4 || mx - nchar16_t < 2) |
| 1851 | break; |
| 1852 | uint8_t c2 = frm_nxt[1]; |
| 1853 | uint8_t c3 = frm_nxt[2]; |
| 1854 | uint8_t c4 = frm_nxt[3]; |
| 1855 | switch (c1) { |
| 1856 | case 0xF0: |
| 1857 | if (!(0x90 <= c2 && c2 <= 0xBF)) |
| 1858 | return static_cast<int>(frm_nxt - frm); |
| 1859 | break; |
| 1860 | case 0xF4: |
| 1861 | if ((c2 & 0xF0) != 0x80) |
| 1862 | return static_cast<int>(frm_nxt - frm); |
| 1863 | break; |
| 1864 | default: |
| 1865 | if ((c2 & 0xC0) != 0x80) |
| 1866 | return static_cast<int>(frm_nxt - frm); |
| 1867 | break; |
| 1868 | } |
| 1869 | if ((c3 & 0xC0) != 0x80 || (c4 & 0xC0) != 0x80) |
| 1870 | break; |
| 1871 | if ((((c1 & 7UL) << 18) + ((c2 & 0x3FUL) << 12) + ((c3 & 0x3FUL) << 6) + (c4 & 0x3F)) > Maxcode) |
| 1872 | break; |
| 1873 | ++nchar16_t; |
| 1874 | frm_nxt += 4; |
| 1875 | } else { |
| 1876 | break; |
| 1877 | } |
| 1878 | } |
| 1879 | return static_cast<int>(frm_nxt - frm); |
| 1880 | } |
| 1881 | |
| 1882 | static codecvt_base::result ucs4_to_utf8( |
| 1883 | const uint32_t* frm, |
| 1884 | const uint32_t* frm_end, |
| 1885 | const uint32_t*& frm_nxt, |
| 1886 | uint8_t* to, |
| 1887 | uint8_t* to_end, |
| 1888 | uint8_t*& to_nxt, |
| 1889 | unsigned long Maxcode = 0x10FFFF, |
| 1890 | codecvt_mode mode = codecvt_mode(0)) { |
| 1891 | frm_nxt = frm; |
| 1892 | to_nxt = to; |
| 1893 | if (mode & generate_header) { |
| 1894 | if (to_end - to_nxt < 3) |
| 1895 | return codecvt_base::partial; |
| 1896 | *to_nxt++ = static_cast<uint8_t>(0xEF); |
| 1897 | *to_nxt++ = static_cast<uint8_t>(0xBB); |
| 1898 | *to_nxt++ = static_cast<uint8_t>(0xBF); |
| 1899 | } |
| 1900 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 1901 | uint32_t wc = *frm_nxt; |
| 1902 | if ((wc & 0xFFFFF800) == 0x00D800 || wc > Maxcode) |
| 1903 | return codecvt_base::error; |
| 1904 | if (wc < 0x000080) { |
| 1905 | if (to_end - to_nxt < 1) |
| 1906 | return codecvt_base::partial; |
| 1907 | *to_nxt++ = static_cast<uint8_t>(wc); |
| 1908 | } else if (wc < 0x000800) { |
| 1909 | if (to_end - to_nxt < 2) |
| 1910 | return codecvt_base::partial; |
| 1911 | *to_nxt++ = static_cast<uint8_t>(0xC0 | (wc >> 6)); |
| 1912 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc & 0x03F)); |
| 1913 | } else if (wc < 0x010000) { |
| 1914 | if (to_end - to_nxt < 3) |
| 1915 | return codecvt_base::partial; |
| 1916 | *to_nxt++ = static_cast<uint8_t>(0xE0 | (wc >> 12)); |
| 1917 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc & 0x0FC0) >> 6)); |
| 1918 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc & 0x003F)); |
| 1919 | } else // if (wc < 0x110000) |
| 1920 | { |
| 1921 | if (to_end - to_nxt < 4) |
| 1922 | return codecvt_base::partial; |
| 1923 | *to_nxt++ = static_cast<uint8_t>(0xF0 | (wc >> 18)); |
| 1924 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc & 0x03F000) >> 12)); |
| 1925 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc & 0x000FC0) >> 6)); |
| 1926 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc & 0x00003F)); |
| 1927 | } |
| 1928 | } |
| 1929 | return codecvt_base::ok; |
| 1930 | } |
| 1931 | |
| 1932 | static codecvt_base::result utf8_to_ucs4( |
| 1933 | const uint8_t* frm, |
| 1934 | const uint8_t* frm_end, |
| 1935 | const uint8_t*& frm_nxt, |
| 1936 | uint32_t* to, |
| 1937 | uint32_t* to_end, |
| 1938 | uint32_t*& to_nxt, |
| 1939 | unsigned long Maxcode = 0x10FFFF, |
| 1940 | codecvt_mode mode = codecvt_mode(0)) { |
| 1941 | frm_nxt = frm; |
| 1942 | to_nxt = to; |
| 1943 | if (mode & consume_header) { |
| 1944 | if (frm_end - frm_nxt >= 3 && frm_nxt[0] == 0xEF && frm_nxt[1] == 0xBB && frm_nxt[2] == 0xBF) |
| 1945 | frm_nxt += 3; |
| 1946 | } |
| 1947 | for (; frm_nxt < frm_end && to_nxt < to_end; ++to_nxt) { |
| 1948 | uint8_t c1 = static_cast<uint8_t>(*frm_nxt); |
| 1949 | if (c1 < 0x80) { |
| 1950 | if (c1 > Maxcode) |
| 1951 | return codecvt_base::error; |
| 1952 | *to_nxt = static_cast<uint32_t>(c1); |
| 1953 | ++frm_nxt; |
| 1954 | } else if (c1 < 0xC2) { |
| 1955 | return codecvt_base::error; |
| 1956 | } else if (c1 < 0xE0) { |
| 1957 | if (frm_end - frm_nxt < 2) |
| 1958 | return codecvt_base::partial; |
| 1959 | uint8_t c2 = frm_nxt[1]; |
| 1960 | if ((c2 & 0xC0) != 0x80) |
| 1961 | return codecvt_base::error; |
| 1962 | uint32_t t = static_cast<uint32_t>(((c1 & 0x1F) << 6) | (c2 & 0x3F)); |
| 1963 | if (t > Maxcode) |
| 1964 | return codecvt_base::error; |
| 1965 | *to_nxt = t; |
| 1966 | frm_nxt += 2; |
| 1967 | } else if (c1 < 0xF0) { |
| 1968 | if (frm_end - frm_nxt < 2) |
| 1969 | return codecvt_base::partial; |
| 1970 | uint8_t c2 = frm_nxt[1]; |
| 1971 | switch (c1) { |
| 1972 | case 0xE0: |
| 1973 | if ((c2 & 0xE0) != 0xA0) |
| 1974 | return codecvt_base::error; |
| 1975 | break; |
| 1976 | case 0xED: |
| 1977 | if ((c2 & 0xE0) != 0x80) |
| 1978 | return codecvt_base::error; |
| 1979 | break; |
| 1980 | default: |
| 1981 | if ((c2 & 0xC0) != 0x80) |
| 1982 | return codecvt_base::error; |
| 1983 | break; |
| 1984 | } |
| 1985 | if (frm_end - frm_nxt < 3) |
| 1986 | return codecvt_base::partial; |
| 1987 | uint8_t c3 = frm_nxt[2]; |
| 1988 | if ((c3 & 0xC0) != 0x80) |
| 1989 | return codecvt_base::error; |
| 1990 | uint32_t t = static_cast<uint32_t>(((c1 & 0x0F) << 12) | ((c2 & 0x3F) << 6) | (c3 & 0x3F)); |
| 1991 | if (t > Maxcode) |
| 1992 | return codecvt_base::error; |
| 1993 | *to_nxt = t; |
| 1994 | frm_nxt += 3; |
| 1995 | } else if (c1 < 0xF5) { |
| 1996 | if (frm_end - frm_nxt < 2) |
| 1997 | return codecvt_base::partial; |
| 1998 | uint8_t c2 = frm_nxt[1]; |
| 1999 | switch (c1) { |
| 2000 | case 0xF0: |
| 2001 | if (!(0x90 <= c2 && c2 <= 0xBF)) |
| 2002 | return codecvt_base::error; |
| 2003 | break; |
| 2004 | case 0xF4: |
| 2005 | if ((c2 & 0xF0) != 0x80) |
| 2006 | return codecvt_base::error; |
| 2007 | break; |
| 2008 | default: |
| 2009 | if ((c2 & 0xC0) != 0x80) |
| 2010 | return codecvt_base::error; |
| 2011 | break; |
| 2012 | } |
| 2013 | if (frm_end - frm_nxt < 3) |
| 2014 | return codecvt_base::partial; |
| 2015 | uint8_t c3 = frm_nxt[2]; |
| 2016 | if ((c3 & 0xC0) != 0x80) |
| 2017 | return codecvt_base::error; |
| 2018 | if (frm_end - frm_nxt < 4) |
| 2019 | return codecvt_base::partial; |
| 2020 | uint8_t c4 = frm_nxt[3]; |
| 2021 | if ((c4 & 0xC0) != 0x80) |
| 2022 | return codecvt_base::error; |
| 2023 | uint32_t t = static_cast<uint32_t>(((c1 & 0x07) << 18) | ((c2 & 0x3F) << 12) | ((c3 & 0x3F) << 6) | (c4 & 0x3F)); |
| 2024 | if (t > Maxcode) |
| 2025 | return codecvt_base::error; |
| 2026 | *to_nxt = t; |
| 2027 | frm_nxt += 4; |
| 2028 | } else { |
| 2029 | return codecvt_base::error; |
| 2030 | } |
| 2031 | } |
| 2032 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 2033 | } |
| 2034 | |
| 2035 | static int utf8_to_ucs4_length( |
| 2036 | const uint8_t* frm, |
| 2037 | const uint8_t* frm_end, |
| 2038 | size_t mx, |
| 2039 | unsigned long Maxcode = 0x10FFFF, |
| 2040 | codecvt_mode mode = codecvt_mode(0)) { |
| 2041 | const uint8_t* frm_nxt = frm; |
| 2042 | if (mode & consume_header) { |
| 2043 | if (frm_end - frm_nxt >= 3 && frm_nxt[0] == 0xEF && frm_nxt[1] == 0xBB && frm_nxt[2] == 0xBF) |
| 2044 | frm_nxt += 3; |
| 2045 | } |
| 2046 | for (size_t nchar32_t = 0; frm_nxt < frm_end && nchar32_t < mx; ++nchar32_t) { |
| 2047 | uint8_t c1 = static_cast<uint8_t>(*frm_nxt); |
| 2048 | if (c1 < 0x80) { |
| 2049 | if (c1 > Maxcode) |
| 2050 | break; |
| 2051 | ++frm_nxt; |
| 2052 | } else if (c1 < 0xC2) { |
| 2053 | break; |
| 2054 | } else if (c1 < 0xE0) { |
| 2055 | if ((frm_end - frm_nxt < 2) || ((frm_nxt[1] & 0xC0) != 0x80)) |
| 2056 | break; |
| 2057 | if ((((c1 & 0x1Fu) << 6) | (frm_nxt[1] & 0x3Fu)) > Maxcode) |
| 2058 | break; |
| 2059 | frm_nxt += 2; |
| 2060 | } else if (c1 < 0xF0) { |
| 2061 | if (frm_end - frm_nxt < 3) |
| 2062 | break; |
| 2063 | uint8_t c2 = frm_nxt[1]; |
| 2064 | uint8_t c3 = frm_nxt[2]; |
| 2065 | switch (c1) { |
| 2066 | case 0xE0: |
| 2067 | if ((c2 & 0xE0) != 0xA0) |
| 2068 | return static_cast<int>(frm_nxt - frm); |
| 2069 | break; |
| 2070 | case 0xED: |
| 2071 | if ((c2 & 0xE0) != 0x80) |
| 2072 | return static_cast<int>(frm_nxt - frm); |
| 2073 | break; |
| 2074 | default: |
| 2075 | if ((c2 & 0xC0) != 0x80) |
| 2076 | return static_cast<int>(frm_nxt - frm); |
| 2077 | break; |
| 2078 | } |
| 2079 | if ((c3 & 0xC0) != 0x80) |
| 2080 | break; |
| 2081 | if ((((c1 & 0x0Fu) << 12) | ((c2 & 0x3Fu) << 6) | (c3 & 0x3Fu)) > Maxcode) |
| 2082 | break; |
| 2083 | frm_nxt += 3; |
| 2084 | } else if (c1 < 0xF5) { |
| 2085 | if (frm_end - frm_nxt < 4) |
| 2086 | break; |
| 2087 | uint8_t c2 = frm_nxt[1]; |
| 2088 | uint8_t c3 = frm_nxt[2]; |
| 2089 | uint8_t c4 = frm_nxt[3]; |
| 2090 | switch (c1) { |
| 2091 | case 0xF0: |
| 2092 | if (!(0x90 <= c2 && c2 <= 0xBF)) |
| 2093 | return static_cast<int>(frm_nxt - frm); |
| 2094 | break; |
| 2095 | case 0xF4: |
| 2096 | if ((c2 & 0xF0) != 0x80) |
| 2097 | return static_cast<int>(frm_nxt - frm); |
| 2098 | break; |
| 2099 | default: |
| 2100 | if ((c2 & 0xC0) != 0x80) |
| 2101 | return static_cast<int>(frm_nxt - frm); |
| 2102 | break; |
| 2103 | } |
| 2104 | if ((c3 & 0xC0) != 0x80 || (c4 & 0xC0) != 0x80) |
| 2105 | break; |
| 2106 | if ((((c1 & 0x07u) << 18) | ((c2 & 0x3Fu) << 12) | ((c3 & 0x3Fu) << 6) | (c4 & 0x3Fu)) > Maxcode) |
| 2107 | break; |
| 2108 | frm_nxt += 4; |
| 2109 | } else { |
| 2110 | break; |
| 2111 | } |
| 2112 | } |
| 2113 | return static_cast<int>(frm_nxt - frm); |
| 2114 | } |
| 2115 | |
| 2116 | static codecvt_base::result ucs2_to_utf8( |
| 2117 | const uint16_t* frm, |
| 2118 | const uint16_t* frm_end, |
| 2119 | const uint16_t*& frm_nxt, |
| 2120 | uint8_t* to, |
| 2121 | uint8_t* to_end, |
| 2122 | uint8_t*& to_nxt, |
| 2123 | unsigned long Maxcode = 0x10FFFF, |
| 2124 | codecvt_mode mode = codecvt_mode(0)) { |
| 2125 | frm_nxt = frm; |
| 2126 | to_nxt = to; |
| 2127 | if (mode & generate_header) { |
| 2128 | if (to_end - to_nxt < 3) |
| 2129 | return codecvt_base::partial; |
| 2130 | *to_nxt++ = static_cast<uint8_t>(0xEF); |
| 2131 | *to_nxt++ = static_cast<uint8_t>(0xBB); |
| 2132 | *to_nxt++ = static_cast<uint8_t>(0xBF); |
| 2133 | } |
| 2134 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 2135 | uint16_t wc = *frm_nxt; |
| 2136 | if ((wc & 0xF800) == 0xD800 || wc > Maxcode) |
| 2137 | return codecvt_base::error; |
| 2138 | if (wc < 0x0080) { |
| 2139 | if (to_end - to_nxt < 1) |
| 2140 | return codecvt_base::partial; |
| 2141 | *to_nxt++ = static_cast<uint8_t>(wc); |
| 2142 | } else if (wc < 0x0800) { |
| 2143 | if (to_end - to_nxt < 2) |
| 2144 | return codecvt_base::partial; |
| 2145 | *to_nxt++ = static_cast<uint8_t>(0xC0 | (wc >> 6)); |
| 2146 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc & 0x03F)); |
| 2147 | } else // if (wc <= 0xFFFF) |
| 2148 | { |
| 2149 | if (to_end - to_nxt < 3) |
| 2150 | return codecvt_base::partial; |
| 2151 | *to_nxt++ = static_cast<uint8_t>(0xE0 | (wc >> 12)); |
| 2152 | *to_nxt++ = static_cast<uint8_t>(0x80 | ((wc & 0x0FC0) >> 6)); |
| 2153 | *to_nxt++ = static_cast<uint8_t>(0x80 | (wc & 0x003F)); |
| 2154 | } |
| 2155 | } |
| 2156 | return codecvt_base::ok; |
| 2157 | } |
| 2158 | |
| 2159 | static codecvt_base::result utf8_to_ucs2( |
| 2160 | const uint8_t* frm, |
| 2161 | const uint8_t* frm_end, |
| 2162 | const uint8_t*& frm_nxt, |
| 2163 | uint16_t* to, |
| 2164 | uint16_t* to_end, |
| 2165 | uint16_t*& to_nxt, |
| 2166 | unsigned long Maxcode = 0x10FFFF, |
| 2167 | codecvt_mode mode = codecvt_mode(0)) { |
| 2168 | frm_nxt = frm; |
| 2169 | to_nxt = to; |
| 2170 | if (mode & consume_header) { |
| 2171 | if (frm_end - frm_nxt >= 3 && frm_nxt[0] == 0xEF && frm_nxt[1] == 0xBB && frm_nxt[2] == 0xBF) |
| 2172 | frm_nxt += 3; |
| 2173 | } |
| 2174 | for (; frm_nxt < frm_end && to_nxt < to_end; ++to_nxt) { |
| 2175 | uint8_t c1 = static_cast<uint8_t>(*frm_nxt); |
| 2176 | if (c1 < 0x80) { |
| 2177 | if (c1 > Maxcode) |
| 2178 | return codecvt_base::error; |
| 2179 | *to_nxt = static_cast<uint16_t>(c1); |
| 2180 | ++frm_nxt; |
| 2181 | } else if (c1 < 0xC2) { |
| 2182 | return codecvt_base::error; |
| 2183 | } else if (c1 < 0xE0) { |
| 2184 | if (frm_end - frm_nxt < 2) |
| 2185 | return codecvt_base::partial; |
| 2186 | uint8_t c2 = frm_nxt[1]; |
| 2187 | if ((c2 & 0xC0) != 0x80) |
| 2188 | return codecvt_base::error; |
| 2189 | uint16_t t = static_cast<uint16_t>(((c1 & 0x1F) << 6) | (c2 & 0x3F)); |
| 2190 | if (t > Maxcode) |
| 2191 | return codecvt_base::error; |
| 2192 | *to_nxt = t; |
| 2193 | frm_nxt += 2; |
| 2194 | } else if (c1 < 0xF0) { |
| 2195 | if (frm_end - frm_nxt < 2) |
| 2196 | return codecvt_base::partial; |
| 2197 | uint8_t c2 = frm_nxt[1]; |
| 2198 | switch (c1) { |
| 2199 | case 0xE0: |
| 2200 | if ((c2 & 0xE0) != 0xA0) |
| 2201 | return codecvt_base::error; |
| 2202 | break; |
| 2203 | case 0xED: |
| 2204 | if ((c2 & 0xE0) != 0x80) |
| 2205 | return codecvt_base::error; |
| 2206 | break; |
| 2207 | default: |
| 2208 | if ((c2 & 0xC0) != 0x80) |
| 2209 | return codecvt_base::error; |
| 2210 | break; |
| 2211 | } |
| 2212 | if (frm_end - frm_nxt < 3) |
| 2213 | return codecvt_base::partial; |
| 2214 | uint8_t c3 = frm_nxt[2]; |
| 2215 | if ((c3 & 0xC0) != 0x80) |
| 2216 | return codecvt_base::error; |
| 2217 | uint16_t t = static_cast<uint16_t>(((c1 & 0x0F) << 12) | ((c2 & 0x3F) << 6) | (c3 & 0x3F)); |
| 2218 | if (t > Maxcode) |
| 2219 | return codecvt_base::error; |
| 2220 | *to_nxt = t; |
| 2221 | frm_nxt += 3; |
| 2222 | } else { |
| 2223 | return codecvt_base::error; |
| 2224 | } |
| 2225 | } |
| 2226 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 2227 | } |
| 2228 | |
| 2229 | static int utf8_to_ucs2_length( |
| 2230 | const uint8_t* frm, |
| 2231 | const uint8_t* frm_end, |
| 2232 | size_t mx, |
| 2233 | unsigned long Maxcode = 0x10FFFF, |
| 2234 | codecvt_mode mode = codecvt_mode(0)) { |
| 2235 | const uint8_t* frm_nxt = frm; |
| 2236 | if (mode & consume_header) { |
| 2237 | if (frm_end - frm_nxt >= 3 && frm_nxt[0] == 0xEF && frm_nxt[1] == 0xBB && frm_nxt[2] == 0xBF) |
| 2238 | frm_nxt += 3; |
| 2239 | } |
| 2240 | for (size_t nchar32_t = 0; frm_nxt < frm_end && nchar32_t < mx; ++nchar32_t) { |
| 2241 | uint8_t c1 = static_cast<uint8_t>(*frm_nxt); |
| 2242 | if (c1 < 0x80) { |
| 2243 | if (c1 > Maxcode) |
| 2244 | break; |
| 2245 | ++frm_nxt; |
| 2246 | } else if (c1 < 0xC2) { |
| 2247 | break; |
| 2248 | } else if (c1 < 0xE0) { |
| 2249 | if ((frm_end - frm_nxt < 2) || ((frm_nxt[1] & 0xC0) != 0x80)) |
| 2250 | break; |
| 2251 | if ((((c1 & 0x1Fu) << 6) | (frm_nxt[1] & 0x3Fu)) > Maxcode) |
| 2252 | break; |
| 2253 | frm_nxt += 2; |
| 2254 | } else if (c1 < 0xF0) { |
| 2255 | if (frm_end - frm_nxt < 3) |
| 2256 | break; |
| 2257 | uint8_t c2 = frm_nxt[1]; |
| 2258 | uint8_t c3 = frm_nxt[2]; |
| 2259 | switch (c1) { |
| 2260 | case 0xE0: |
| 2261 | if ((c2 & 0xE0) != 0xA0) |
| 2262 | return static_cast<int>(frm_nxt - frm); |
| 2263 | break; |
| 2264 | case 0xED: |
| 2265 | if ((c2 & 0xE0) != 0x80) |
| 2266 | return static_cast<int>(frm_nxt - frm); |
| 2267 | break; |
| 2268 | default: |
| 2269 | if ((c2 & 0xC0) != 0x80) |
| 2270 | return static_cast<int>(frm_nxt - frm); |
| 2271 | break; |
| 2272 | } |
| 2273 | if ((c3 & 0xC0) != 0x80) |
| 2274 | break; |
| 2275 | if ((((c1 & 0x0Fu) << 12) | ((c2 & 0x3Fu) << 6) | (c3 & 0x3Fu)) > Maxcode) |
| 2276 | break; |
| 2277 | frm_nxt += 3; |
| 2278 | } else { |
| 2279 | break; |
| 2280 | } |
| 2281 | } |
| 2282 | return static_cast<int>(frm_nxt - frm); |
| 2283 | } |
| 2284 | |
| 2285 | static codecvt_base::result ucs4_to_utf16be( |
| 2286 | const uint32_t* frm, |
| 2287 | const uint32_t* frm_end, |
| 2288 | const uint32_t*& frm_nxt, |
| 2289 | uint8_t* to, |
| 2290 | uint8_t* to_end, |
| 2291 | uint8_t*& to_nxt, |
| 2292 | unsigned long Maxcode = 0x10FFFF, |
| 2293 | codecvt_mode mode = codecvt_mode(0)) { |
| 2294 | frm_nxt = frm; |
| 2295 | to_nxt = to; |
| 2296 | if (mode & generate_header) { |
| 2297 | if (to_end - to_nxt < 2) |
| 2298 | return codecvt_base::partial; |
| 2299 | *to_nxt++ = static_cast<uint8_t>(0xFE); |
| 2300 | *to_nxt++ = static_cast<uint8_t>(0xFF); |
| 2301 | } |
| 2302 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 2303 | uint32_t wc = *frm_nxt; |
| 2304 | if ((wc & 0xFFFFF800) == 0x00D800 || wc > Maxcode) |
| 2305 | return codecvt_base::error; |
| 2306 | if (wc < 0x010000) { |
| 2307 | if (to_end - to_nxt < 2) |
| 2308 | return codecvt_base::partial; |
| 2309 | *to_nxt++ = static_cast<uint8_t>(wc >> 8); |
| 2310 | *to_nxt++ = static_cast<uint8_t>(wc); |
| 2311 | } else { |
| 2312 | if (to_end - to_nxt < 4) |
| 2313 | return codecvt_base::partial; |
| 2314 | uint16_t t = static_cast<uint16_t>(0xD800 | ((((wc & 0x1F0000) >> 16) - 1) << 6) | ((wc & 0x00FC00) >> 10)); |
| 2315 | *to_nxt++ = static_cast<uint8_t>(t >> 8); |
| 2316 | *to_nxt++ = static_cast<uint8_t>(t); |
| 2317 | t = static_cast<uint16_t>(0xDC00 | (wc & 0x03FF)); |
| 2318 | *to_nxt++ = static_cast<uint8_t>(t >> 8); |
| 2319 | *to_nxt++ = static_cast<uint8_t>(t); |
| 2320 | } |
| 2321 | } |
| 2322 | return codecvt_base::ok; |
| 2323 | } |
| 2324 | |
| 2325 | static codecvt_base::result utf16be_to_ucs4( |
| 2326 | const uint8_t* frm, |
| 2327 | const uint8_t* frm_end, |
| 2328 | const uint8_t*& frm_nxt, |
| 2329 | uint32_t* to, |
| 2330 | uint32_t* to_end, |
| 2331 | uint32_t*& to_nxt, |
| 2332 | unsigned long Maxcode = 0x10FFFF, |
| 2333 | codecvt_mode mode = codecvt_mode(0)) { |
| 2334 | frm_nxt = frm; |
| 2335 | to_nxt = to; |
| 2336 | if (mode & consume_header) { |
| 2337 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFE && frm_nxt[1] == 0xFF) |
| 2338 | frm_nxt += 2; |
| 2339 | } |
| 2340 | for (; frm_nxt < frm_end - 1 && to_nxt < to_end; ++to_nxt) { |
| 2341 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[0] << 8) | frm_nxt[1]); |
| 2342 | if ((c1 & 0xFC00) == 0xDC00) |
| 2343 | return codecvt_base::error; |
| 2344 | if ((c1 & 0xFC00) != 0xD800) { |
| 2345 | if (c1 > Maxcode) |
| 2346 | return codecvt_base::error; |
| 2347 | *to_nxt = static_cast<uint32_t>(c1); |
| 2348 | frm_nxt += 2; |
| 2349 | } else { |
| 2350 | if (frm_end - frm_nxt < 4) |
| 2351 | return codecvt_base::partial; |
| 2352 | uint16_t c2 = static_cast<uint16_t>((frm_nxt[2] << 8) | frm_nxt[3]); |
| 2353 | if ((c2 & 0xFC00) != 0xDC00) |
| 2354 | return codecvt_base::error; |
| 2355 | uint32_t t = static_cast<uint32_t>(((((c1 & 0x03C0) >> 6) + 1) << 16) | ((c1 & 0x003F) << 10) | (c2 & 0x03FF)); |
| 2356 | if (t > Maxcode) |
| 2357 | return codecvt_base::error; |
| 2358 | *to_nxt = t; |
| 2359 | frm_nxt += 4; |
| 2360 | } |
| 2361 | } |
| 2362 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 2363 | } |
| 2364 | |
| 2365 | static int utf16be_to_ucs4_length( |
| 2366 | const uint8_t* frm, |
| 2367 | const uint8_t* frm_end, |
| 2368 | size_t mx, |
| 2369 | unsigned long Maxcode = 0x10FFFF, |
| 2370 | codecvt_mode mode = codecvt_mode(0)) { |
| 2371 | const uint8_t* frm_nxt = frm; |
| 2372 | if (mode & consume_header) { |
| 2373 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFE && frm_nxt[1] == 0xFF) |
| 2374 | frm_nxt += 2; |
| 2375 | } |
| 2376 | for (size_t nchar32_t = 0; frm_nxt < frm_end - 1 && nchar32_t < mx; ++nchar32_t) { |
| 2377 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[0] << 8) | frm_nxt[1]); |
| 2378 | if ((c1 & 0xFC00) == 0xDC00) |
| 2379 | break; |
| 2380 | if ((c1 & 0xFC00) != 0xD800) { |
| 2381 | if (c1 > Maxcode) |
| 2382 | break; |
| 2383 | frm_nxt += 2; |
| 2384 | } else { |
| 2385 | if (frm_end - frm_nxt < 4) |
| 2386 | break; |
| 2387 | uint16_t c2 = static_cast<uint16_t>((frm_nxt[2] << 8) | frm_nxt[3]); |
| 2388 | if ((c2 & 0xFC00) != 0xDC00) |
| 2389 | break; |
| 2390 | uint32_t t = static_cast<uint32_t>(((((c1 & 0x03C0) >> 6) + 1) << 16) | ((c1 & 0x003F) << 10) | (c2 & 0x03FF)); |
| 2391 | if (t > Maxcode) |
| 2392 | break; |
| 2393 | frm_nxt += 4; |
| 2394 | } |
| 2395 | } |
| 2396 | return static_cast<int>(frm_nxt - frm); |
| 2397 | } |
| 2398 | |
| 2399 | static codecvt_base::result ucs4_to_utf16le( |
| 2400 | const uint32_t* frm, |
| 2401 | const uint32_t* frm_end, |
| 2402 | const uint32_t*& frm_nxt, |
| 2403 | uint8_t* to, |
| 2404 | uint8_t* to_end, |
| 2405 | uint8_t*& to_nxt, |
| 2406 | unsigned long Maxcode = 0x10FFFF, |
| 2407 | codecvt_mode mode = codecvt_mode(0)) { |
| 2408 | frm_nxt = frm; |
| 2409 | to_nxt = to; |
| 2410 | if (mode & generate_header) { |
| 2411 | if (to_end - to_nxt < 2) |
| 2412 | return codecvt_base::partial; |
| 2413 | *to_nxt++ = static_cast<uint8_t>(0xFF); |
| 2414 | *to_nxt++ = static_cast<uint8_t>(0xFE); |
| 2415 | } |
| 2416 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 2417 | uint32_t wc = *frm_nxt; |
| 2418 | if ((wc & 0xFFFFF800) == 0x00D800 || wc > Maxcode) |
| 2419 | return codecvt_base::error; |
| 2420 | if (wc < 0x010000) { |
| 2421 | if (to_end - to_nxt < 2) |
| 2422 | return codecvt_base::partial; |
| 2423 | *to_nxt++ = static_cast<uint8_t>(wc); |
| 2424 | *to_nxt++ = static_cast<uint8_t>(wc >> 8); |
| 2425 | } else { |
| 2426 | if (to_end - to_nxt < 4) |
| 2427 | return codecvt_base::partial; |
| 2428 | uint16_t t = static_cast<uint16_t>(0xD800 | ((((wc & 0x1F0000) >> 16) - 1) << 6) | ((wc & 0x00FC00) >> 10)); |
| 2429 | *to_nxt++ = static_cast<uint8_t>(t); |
| 2430 | *to_nxt++ = static_cast<uint8_t>(t >> 8); |
| 2431 | t = static_cast<uint16_t>(0xDC00 | (wc & 0x03FF)); |
| 2432 | *to_nxt++ = static_cast<uint8_t>(t); |
| 2433 | *to_nxt++ = static_cast<uint8_t>(t >> 8); |
| 2434 | } |
| 2435 | } |
| 2436 | return codecvt_base::ok; |
| 2437 | } |
| 2438 | |
| 2439 | static codecvt_base::result utf16le_to_ucs4( |
| 2440 | const uint8_t* frm, |
| 2441 | const uint8_t* frm_end, |
| 2442 | const uint8_t*& frm_nxt, |
| 2443 | uint32_t* to, |
| 2444 | uint32_t* to_end, |
| 2445 | uint32_t*& to_nxt, |
| 2446 | unsigned long Maxcode = 0x10FFFF, |
| 2447 | codecvt_mode mode = codecvt_mode(0)) { |
| 2448 | frm_nxt = frm; |
| 2449 | to_nxt = to; |
| 2450 | if (mode & consume_header) { |
| 2451 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFF && frm_nxt[1] == 0xFE) |
| 2452 | frm_nxt += 2; |
| 2453 | } |
| 2454 | for (; frm_nxt < frm_end - 1 && to_nxt < to_end; ++to_nxt) { |
| 2455 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[1] << 8) | frm_nxt[0]); |
| 2456 | if ((c1 & 0xFC00) == 0xDC00) |
| 2457 | return codecvt_base::error; |
| 2458 | if ((c1 & 0xFC00) != 0xD800) { |
| 2459 | if (c1 > Maxcode) |
| 2460 | return codecvt_base::error; |
| 2461 | *to_nxt = static_cast<uint32_t>(c1); |
| 2462 | frm_nxt += 2; |
| 2463 | } else { |
| 2464 | if (frm_end - frm_nxt < 4) |
| 2465 | return codecvt_base::partial; |
| 2466 | uint16_t c2 = static_cast<uint16_t>((frm_nxt[3] << 8) | frm_nxt[2]); |
| 2467 | if ((c2 & 0xFC00) != 0xDC00) |
| 2468 | return codecvt_base::error; |
| 2469 | uint32_t t = static_cast<uint32_t>(((((c1 & 0x03C0) >> 6) + 1) << 16) | ((c1 & 0x003F) << 10) | (c2 & 0x03FF)); |
| 2470 | if (t > Maxcode) |
| 2471 | return codecvt_base::error; |
| 2472 | *to_nxt = t; |
| 2473 | frm_nxt += 4; |
| 2474 | } |
| 2475 | } |
| 2476 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 2477 | } |
| 2478 | |
| 2479 | static int utf16le_to_ucs4_length( |
| 2480 | const uint8_t* frm, |
| 2481 | const uint8_t* frm_end, |
| 2482 | size_t mx, |
| 2483 | unsigned long Maxcode = 0x10FFFF, |
| 2484 | codecvt_mode mode = codecvt_mode(0)) { |
| 2485 | const uint8_t* frm_nxt = frm; |
| 2486 | if (mode & consume_header) { |
| 2487 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFF && frm_nxt[1] == 0xFE) |
| 2488 | frm_nxt += 2; |
| 2489 | } |
| 2490 | for (size_t nchar32_t = 0; frm_nxt < frm_end - 1 && nchar32_t < mx; ++nchar32_t) { |
| 2491 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[1] << 8) | frm_nxt[0]); |
| 2492 | if ((c1 & 0xFC00) == 0xDC00) |
| 2493 | break; |
| 2494 | if ((c1 & 0xFC00) != 0xD800) { |
| 2495 | if (c1 > Maxcode) |
| 2496 | break; |
| 2497 | frm_nxt += 2; |
| 2498 | } else { |
| 2499 | if (frm_end - frm_nxt < 4) |
| 2500 | break; |
| 2501 | uint16_t c2 = static_cast<uint16_t>((frm_nxt[3] << 8) | frm_nxt[2]); |
| 2502 | if ((c2 & 0xFC00) != 0xDC00) |
| 2503 | break; |
| 2504 | uint32_t t = static_cast<uint32_t>(((((c1 & 0x03C0) >> 6) + 1) << 16) | ((c1 & 0x003F) << 10) | (c2 & 0x03FF)); |
| 2505 | if (t > Maxcode) |
| 2506 | break; |
| 2507 | frm_nxt += 4; |
| 2508 | } |
| 2509 | } |
| 2510 | return static_cast<int>(frm_nxt - frm); |
| 2511 | } |
| 2512 | |
| 2513 | static codecvt_base::result ucs2_to_utf16be( |
| 2514 | const uint16_t* frm, |
| 2515 | const uint16_t* frm_end, |
| 2516 | const uint16_t*& frm_nxt, |
| 2517 | uint8_t* to, |
| 2518 | uint8_t* to_end, |
| 2519 | uint8_t*& to_nxt, |
| 2520 | unsigned long Maxcode = 0x10FFFF, |
| 2521 | codecvt_mode mode = codecvt_mode(0)) { |
| 2522 | frm_nxt = frm; |
| 2523 | to_nxt = to; |
| 2524 | if (mode & generate_header) { |
| 2525 | if (to_end - to_nxt < 2) |
| 2526 | return codecvt_base::partial; |
| 2527 | *to_nxt++ = static_cast<uint8_t>(0xFE); |
| 2528 | *to_nxt++ = static_cast<uint8_t>(0xFF); |
| 2529 | } |
| 2530 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 2531 | uint16_t wc = *frm_nxt; |
| 2532 | if ((wc & 0xF800) == 0xD800 || wc > Maxcode) |
| 2533 | return codecvt_base::error; |
| 2534 | if (to_end - to_nxt < 2) |
| 2535 | return codecvt_base::partial; |
| 2536 | *to_nxt++ = static_cast<uint8_t>(wc >> 8); |
| 2537 | *to_nxt++ = static_cast<uint8_t>(wc); |
| 2538 | } |
| 2539 | return codecvt_base::ok; |
| 2540 | } |
| 2541 | |
| 2542 | static codecvt_base::result utf16be_to_ucs2( |
| 2543 | const uint8_t* frm, |
| 2544 | const uint8_t* frm_end, |
| 2545 | const uint8_t*& frm_nxt, |
| 2546 | uint16_t* to, |
| 2547 | uint16_t* to_end, |
| 2548 | uint16_t*& to_nxt, |
| 2549 | unsigned long Maxcode = 0x10FFFF, |
| 2550 | codecvt_mode mode = codecvt_mode(0)) { |
| 2551 | frm_nxt = frm; |
| 2552 | to_nxt = to; |
| 2553 | if (mode & consume_header) { |
| 2554 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFE && frm_nxt[1] == 0xFF) |
| 2555 | frm_nxt += 2; |
| 2556 | } |
| 2557 | for (; frm_nxt < frm_end - 1 && to_nxt < to_end; ++to_nxt) { |
| 2558 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[0] << 8) | frm_nxt[1]); |
| 2559 | if ((c1 & 0xF800) == 0xD800 || c1 > Maxcode) |
| 2560 | return codecvt_base::error; |
| 2561 | *to_nxt = c1; |
| 2562 | frm_nxt += 2; |
| 2563 | } |
| 2564 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 2565 | } |
| 2566 | |
| 2567 | static int utf16be_to_ucs2_length( |
| 2568 | const uint8_t* frm, |
| 2569 | const uint8_t* frm_end, |
| 2570 | size_t mx, |
| 2571 | unsigned long Maxcode = 0x10FFFF, |
| 2572 | codecvt_mode mode = codecvt_mode(0)) { |
| 2573 | const uint8_t* frm_nxt = frm; |
| 2574 | if (mode & consume_header) { |
| 2575 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFE && frm_nxt[1] == 0xFF) |
| 2576 | frm_nxt += 2; |
| 2577 | } |
| 2578 | for (size_t nchar16_t = 0; frm_nxt < frm_end - 1 && nchar16_t < mx; ++nchar16_t) { |
| 2579 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[0] << 8) | frm_nxt[1]); |
| 2580 | if ((c1 & 0xF800) == 0xD800 || c1 > Maxcode) |
| 2581 | break; |
| 2582 | frm_nxt += 2; |
| 2583 | } |
| 2584 | return static_cast<int>(frm_nxt - frm); |
| 2585 | } |
| 2586 | |
| 2587 | static codecvt_base::result ucs2_to_utf16le( |
| 2588 | const uint16_t* frm, |
| 2589 | const uint16_t* frm_end, |
| 2590 | const uint16_t*& frm_nxt, |
| 2591 | uint8_t* to, |
| 2592 | uint8_t* to_end, |
| 2593 | uint8_t*& to_nxt, |
| 2594 | unsigned long Maxcode = 0x10FFFF, |
| 2595 | codecvt_mode mode = codecvt_mode(0)) { |
| 2596 | frm_nxt = frm; |
| 2597 | to_nxt = to; |
| 2598 | if (mode & generate_header) { |
| 2599 | if (to_end - to_nxt < 2) |
| 2600 | return codecvt_base::partial; |
| 2601 | *to_nxt++ = static_cast<uint8_t>(0xFF); |
| 2602 | *to_nxt++ = static_cast<uint8_t>(0xFE); |
| 2603 | } |
| 2604 | for (; frm_nxt < frm_end; ++frm_nxt) { |
| 2605 | uint16_t wc = *frm_nxt; |
| 2606 | if ((wc & 0xF800) == 0xD800 || wc > Maxcode) |
| 2607 | return codecvt_base::error; |
| 2608 | if (to_end - to_nxt < 2) |
| 2609 | return codecvt_base::partial; |
| 2610 | *to_nxt++ = static_cast<uint8_t>(wc); |
| 2611 | *to_nxt++ = static_cast<uint8_t>(wc >> 8); |
| 2612 | } |
| 2613 | return codecvt_base::ok; |
| 2614 | } |
| 2615 | |
| 2616 | static codecvt_base::result utf16le_to_ucs2( |
| 2617 | const uint8_t* frm, |
| 2618 | const uint8_t* frm_end, |
| 2619 | const uint8_t*& frm_nxt, |
| 2620 | uint16_t* to, |
| 2621 | uint16_t* to_end, |
| 2622 | uint16_t*& to_nxt, |
| 2623 | unsigned long Maxcode = 0x10FFFF, |
| 2624 | codecvt_mode mode = codecvt_mode(0)) { |
| 2625 | frm_nxt = frm; |
| 2626 | to_nxt = to; |
| 2627 | if (mode & consume_header) { |
| 2628 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFF && frm_nxt[1] == 0xFE) |
| 2629 | frm_nxt += 2; |
| 2630 | } |
| 2631 | for (; frm_nxt < frm_end - 1 && to_nxt < to_end; ++to_nxt) { |
| 2632 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[1] << 8) | frm_nxt[0]); |
| 2633 | if ((c1 & 0xF800) == 0xD800 || c1 > Maxcode) |
| 2634 | return codecvt_base::error; |
| 2635 | *to_nxt = c1; |
| 2636 | frm_nxt += 2; |
| 2637 | } |
| 2638 | return frm_nxt < frm_end ? codecvt_base::partial : codecvt_base::ok; |
| 2639 | } |
| 2640 | |
| 2641 | static int utf16le_to_ucs2_length( |
| 2642 | const uint8_t* frm, |
| 2643 | const uint8_t* frm_end, |
| 2644 | size_t mx, |
| 2645 | unsigned long Maxcode = 0x10FFFF, |
| 2646 | codecvt_mode mode = codecvt_mode(0)) { |
| 2647 | const uint8_t* frm_nxt = frm; |
| 2648 | frm_nxt = frm; |
| 2649 | if (mode & consume_header) { |
| 2650 | if (frm_end - frm_nxt >= 2 && frm_nxt[0] == 0xFF && frm_nxt[1] == 0xFE) |
| 2651 | frm_nxt += 2; |
| 2652 | } |
| 2653 | for (size_t nchar16_t = 0; frm_nxt < frm_end - 1 && nchar16_t < mx; ++nchar16_t) { |
| 2654 | uint16_t c1 = static_cast<uint16_t>((frm_nxt[1] << 8) | frm_nxt[0]); |
| 2655 | if ((c1 & 0xF800) == 0xD800 || c1 > Maxcode) |
| 2656 | break; |
| 2657 | frm_nxt += 2; |
| 2658 | } |
| 2659 | return static_cast<int>(frm_nxt - frm); |
| 2660 | } |
| 2661 | |
| 2662 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 2663 | |
| 2664 | // template <> class codecvt<char16_t, char, mbstate_t> |
| 2665 | |
| 2666 | constinit locale::id codecvt<char16_t, char, mbstate_t>::id; |
| 2667 | |
| 2668 | codecvt<char16_t, char, mbstate_t>::~codecvt() {} |
| 2669 | |
| 2670 | codecvt<char16_t, char, mbstate_t>::result codecvt<char16_t, char, mbstate_t>::do_out( |
| 2671 | state_type&, |
| 2672 | const intern_type* frm, |
| 2673 | const intern_type* frm_end, |
| 2674 | const intern_type*& frm_nxt, |
| 2675 | extern_type* to, |
| 2676 | extern_type* to_end, |
| 2677 | extern_type*& to_nxt) const { |
| 2678 | const uint16_t* _frm = reinterpret_cast<const uint16_t*>(frm); |
| 2679 | const uint16_t* _frm_end = reinterpret_cast<const uint16_t*>(frm_end); |
| 2680 | const uint16_t* _frm_nxt = _frm; |
| 2681 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 2682 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 2683 | uint8_t* _to_nxt = _to; |
| 2684 | result r = utf16_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2685 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2686 | to_nxt = to + (_to_nxt - _to); |
| 2687 | return r; |
| 2688 | } |
| 2689 | |
| 2690 | codecvt<char16_t, char, mbstate_t>::result codecvt<char16_t, char, mbstate_t>::do_in( |
| 2691 | state_type&, |
| 2692 | const extern_type* frm, |
| 2693 | const extern_type* frm_end, |
| 2694 | const extern_type*& frm_nxt, |
| 2695 | intern_type* to, |
| 2696 | intern_type* to_end, |
| 2697 | intern_type*& to_nxt) const { |
| 2698 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2699 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2700 | const uint8_t* _frm_nxt = _frm; |
| 2701 | uint16_t* _to = reinterpret_cast<uint16_t*>(to); |
| 2702 | uint16_t* _to_end = reinterpret_cast<uint16_t*>(to_end); |
| 2703 | uint16_t* _to_nxt = _to; |
| 2704 | result r = utf8_to_utf16(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2705 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2706 | to_nxt = to + (_to_nxt - _to); |
| 2707 | return r; |
| 2708 | } |
| 2709 | |
| 2710 | codecvt<char16_t, char, mbstate_t>::result |
| 2711 | codecvt<char16_t, char, mbstate_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 2712 | to_nxt = to; |
| 2713 | return noconv; |
| 2714 | } |
| 2715 | |
| 2716 | int codecvt<char16_t, char, mbstate_t>::do_encoding() const noexcept { return 0; } |
| 2717 | |
| 2718 | bool codecvt<char16_t, char, mbstate_t>::do_always_noconv() const noexcept { return false; } |
| 2719 | |
| 2720 | int codecvt<char16_t, char, mbstate_t>::do_length( |
| 2721 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 2722 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2723 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2724 | return utf8_to_utf16_length(frm: _frm, frm_end: _frm_end, mx); |
| 2725 | } |
| 2726 | |
| 2727 | int codecvt<char16_t, char, mbstate_t>::do_max_length() const noexcept { return 4; } |
| 2728 | |
| 2729 | #if _LIBCPP_HAS_CHAR8_T |
| 2730 | |
| 2731 | // template <> class codecvt<char16_t, char8_t, mbstate_t> |
| 2732 | |
| 2733 | constinit locale::id codecvt<char16_t, char8_t, mbstate_t>::id; |
| 2734 | |
| 2735 | codecvt<char16_t, char8_t, mbstate_t>::~codecvt() {} |
| 2736 | |
| 2737 | codecvt<char16_t, char8_t, mbstate_t>::result codecvt<char16_t, char8_t, mbstate_t>::do_out( |
| 2738 | state_type&, |
| 2739 | const intern_type* frm, |
| 2740 | const intern_type* frm_end, |
| 2741 | const intern_type*& frm_nxt, |
| 2742 | extern_type* to, |
| 2743 | extern_type* to_end, |
| 2744 | extern_type*& to_nxt) const { |
| 2745 | const uint16_t* _frm = reinterpret_cast<const uint16_t*>(frm); |
| 2746 | const uint16_t* _frm_end = reinterpret_cast<const uint16_t*>(frm_end); |
| 2747 | const uint16_t* _frm_nxt = _frm; |
| 2748 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 2749 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 2750 | uint8_t* _to_nxt = _to; |
| 2751 | result r = utf16_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2752 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2753 | to_nxt = to + (_to_nxt - _to); |
| 2754 | return r; |
| 2755 | } |
| 2756 | |
| 2757 | codecvt<char16_t, char8_t, mbstate_t>::result codecvt<char16_t, char8_t, mbstate_t>::do_in( |
| 2758 | state_type&, |
| 2759 | const extern_type* frm, |
| 2760 | const extern_type* frm_end, |
| 2761 | const extern_type*& frm_nxt, |
| 2762 | intern_type* to, |
| 2763 | intern_type* to_end, |
| 2764 | intern_type*& to_nxt) const { |
| 2765 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2766 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2767 | const uint8_t* _frm_nxt = _frm; |
| 2768 | uint16_t* _to = reinterpret_cast<uint16_t*>(to); |
| 2769 | uint16_t* _to_end = reinterpret_cast<uint16_t*>(to_end); |
| 2770 | uint16_t* _to_nxt = _to; |
| 2771 | result r = utf8_to_utf16(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2772 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2773 | to_nxt = to + (_to_nxt - _to); |
| 2774 | return r; |
| 2775 | } |
| 2776 | |
| 2777 | codecvt<char16_t, char8_t, mbstate_t>::result codecvt<char16_t, char8_t, mbstate_t>::do_unshift( |
| 2778 | state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 2779 | to_nxt = to; |
| 2780 | return noconv; |
| 2781 | } |
| 2782 | |
| 2783 | int codecvt<char16_t, char8_t, mbstate_t>::do_encoding() const noexcept { return 0; } |
| 2784 | |
| 2785 | bool codecvt<char16_t, char8_t, mbstate_t>::do_always_noconv() const noexcept { return false; } |
| 2786 | |
| 2787 | int codecvt<char16_t, char8_t, mbstate_t>::do_length( |
| 2788 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 2789 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2790 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2791 | return utf8_to_utf16_length(frm: _frm, frm_end: _frm_end, mx); |
| 2792 | } |
| 2793 | |
| 2794 | int codecvt<char16_t, char8_t, mbstate_t>::do_max_length() const noexcept { return 4; } |
| 2795 | |
| 2796 | #endif |
| 2797 | |
| 2798 | // template <> class codecvt<char32_t, char, mbstate_t> |
| 2799 | |
| 2800 | constinit locale::id codecvt<char32_t, char, mbstate_t>::id; |
| 2801 | |
| 2802 | codecvt<char32_t, char, mbstate_t>::~codecvt() {} |
| 2803 | |
| 2804 | codecvt<char32_t, char, mbstate_t>::result codecvt<char32_t, char, mbstate_t>::do_out( |
| 2805 | state_type&, |
| 2806 | const intern_type* frm, |
| 2807 | const intern_type* frm_end, |
| 2808 | const intern_type*& frm_nxt, |
| 2809 | extern_type* to, |
| 2810 | extern_type* to_end, |
| 2811 | extern_type*& to_nxt) const { |
| 2812 | const uint32_t* _frm = reinterpret_cast<const uint32_t*>(frm); |
| 2813 | const uint32_t* _frm_end = reinterpret_cast<const uint32_t*>(frm_end); |
| 2814 | const uint32_t* _frm_nxt = _frm; |
| 2815 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 2816 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 2817 | uint8_t* _to_nxt = _to; |
| 2818 | result r = ucs4_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2819 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2820 | to_nxt = to + (_to_nxt - _to); |
| 2821 | return r; |
| 2822 | } |
| 2823 | |
| 2824 | codecvt<char32_t, char, mbstate_t>::result codecvt<char32_t, char, mbstate_t>::do_in( |
| 2825 | state_type&, |
| 2826 | const extern_type* frm, |
| 2827 | const extern_type* frm_end, |
| 2828 | const extern_type*& frm_nxt, |
| 2829 | intern_type* to, |
| 2830 | intern_type* to_end, |
| 2831 | intern_type*& to_nxt) const { |
| 2832 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2833 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2834 | const uint8_t* _frm_nxt = _frm; |
| 2835 | uint32_t* _to = reinterpret_cast<uint32_t*>(to); |
| 2836 | uint32_t* _to_end = reinterpret_cast<uint32_t*>(to_end); |
| 2837 | uint32_t* _to_nxt = _to; |
| 2838 | result r = utf8_to_ucs4(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2839 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2840 | to_nxt = to + (_to_nxt - _to); |
| 2841 | return r; |
| 2842 | } |
| 2843 | |
| 2844 | codecvt<char32_t, char, mbstate_t>::result |
| 2845 | codecvt<char32_t, char, mbstate_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 2846 | to_nxt = to; |
| 2847 | return noconv; |
| 2848 | } |
| 2849 | |
| 2850 | int codecvt<char32_t, char, mbstate_t>::do_encoding() const noexcept { return 0; } |
| 2851 | |
| 2852 | bool codecvt<char32_t, char, mbstate_t>::do_always_noconv() const noexcept { return false; } |
| 2853 | |
| 2854 | int codecvt<char32_t, char, mbstate_t>::do_length( |
| 2855 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 2856 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2857 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2858 | return utf8_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx); |
| 2859 | } |
| 2860 | |
| 2861 | int codecvt<char32_t, char, mbstate_t>::do_max_length() const noexcept { return 4; } |
| 2862 | |
| 2863 | #if _LIBCPP_HAS_CHAR8_T |
| 2864 | |
| 2865 | // template <> class codecvt<char32_t, char8_t, mbstate_t> |
| 2866 | |
| 2867 | constinit locale::id codecvt<char32_t, char8_t, mbstate_t>::id; |
| 2868 | |
| 2869 | codecvt<char32_t, char8_t, mbstate_t>::~codecvt() {} |
| 2870 | |
| 2871 | codecvt<char32_t, char8_t, mbstate_t>::result codecvt<char32_t, char8_t, mbstate_t>::do_out( |
| 2872 | state_type&, |
| 2873 | const intern_type* frm, |
| 2874 | const intern_type* frm_end, |
| 2875 | const intern_type*& frm_nxt, |
| 2876 | extern_type* to, |
| 2877 | extern_type* to_end, |
| 2878 | extern_type*& to_nxt) const { |
| 2879 | const uint32_t* _frm = reinterpret_cast<const uint32_t*>(frm); |
| 2880 | const uint32_t* _frm_end = reinterpret_cast<const uint32_t*>(frm_end); |
| 2881 | const uint32_t* _frm_nxt = _frm; |
| 2882 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 2883 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 2884 | uint8_t* _to_nxt = _to; |
| 2885 | result r = ucs4_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2886 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2887 | to_nxt = to + (_to_nxt - _to); |
| 2888 | return r; |
| 2889 | } |
| 2890 | |
| 2891 | codecvt<char32_t, char8_t, mbstate_t>::result codecvt<char32_t, char8_t, mbstate_t>::do_in( |
| 2892 | state_type&, |
| 2893 | const extern_type* frm, |
| 2894 | const extern_type* frm_end, |
| 2895 | const extern_type*& frm_nxt, |
| 2896 | intern_type* to, |
| 2897 | intern_type* to_end, |
| 2898 | intern_type*& to_nxt) const { |
| 2899 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2900 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2901 | const uint8_t* _frm_nxt = _frm; |
| 2902 | uint32_t* _to = reinterpret_cast<uint32_t*>(to); |
| 2903 | uint32_t* _to_end = reinterpret_cast<uint32_t*>(to_end); |
| 2904 | uint32_t* _to_nxt = _to; |
| 2905 | result r = utf8_to_ucs4(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt); |
| 2906 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2907 | to_nxt = to + (_to_nxt - _to); |
| 2908 | return r; |
| 2909 | } |
| 2910 | |
| 2911 | codecvt<char32_t, char8_t, mbstate_t>::result codecvt<char32_t, char8_t, mbstate_t>::do_unshift( |
| 2912 | state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 2913 | to_nxt = to; |
| 2914 | return noconv; |
| 2915 | } |
| 2916 | |
| 2917 | int codecvt<char32_t, char8_t, mbstate_t>::do_encoding() const noexcept { return 0; } |
| 2918 | |
| 2919 | bool codecvt<char32_t, char8_t, mbstate_t>::do_always_noconv() const noexcept { return false; } |
| 2920 | |
| 2921 | int codecvt<char32_t, char8_t, mbstate_t>::do_length( |
| 2922 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 2923 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2924 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2925 | return utf8_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx); |
| 2926 | } |
| 2927 | |
| 2928 | int codecvt<char32_t, char8_t, mbstate_t>::do_max_length() const noexcept { return 4; } |
| 2929 | |
| 2930 | #endif |
| 2931 | |
| 2932 | // __codecvt_utf8<wchar_t> |
| 2933 | |
| 2934 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 2935 | |
| 2936 | constexpr bool is_short_wchar = sizeof(wchar_t) == sizeof(uint16_t); |
| 2937 | using underlying_wchar_t = conditional_t<is_short_wchar, uint16_t, uint32_t>; |
| 2938 | static_assert(sizeof(underlying_wchar_t) == sizeof(wchar_t)); |
| 2939 | |
| 2940 | template <auto callable> |
| 2941 | struct call { |
| 2942 | template <class... Args> |
| 2943 | static auto operator()(Args&&... args) -> decltype(callable(std::forward<Args>(args)...)) { |
| 2944 | return callable(std::forward<Args>(args)...); |
| 2945 | } |
| 2946 | }; |
| 2947 | |
| 2948 | template <auto... Args, class... Args2> |
| 2949 | auto overload(Args2&&... args) { |
| 2950 | struct overload_t : call<Args>... { |
| 2951 | using call<Args>::operator()...; |
| 2952 | }; |
| 2953 | |
| 2954 | return overload_t{}(std::forward<Args2>(args)...); |
| 2955 | } |
| 2956 | |
| 2957 | __codecvt_utf8<wchar_t>::result __codecvt_utf8<wchar_t>::do_out( |
| 2958 | state_type&, |
| 2959 | const intern_type* frm, |
| 2960 | const intern_type* frm_end, |
| 2961 | const intern_type*& frm_nxt, |
| 2962 | extern_type* to, |
| 2963 | extern_type* to_end, |
| 2964 | extern_type*& to_nxt) const { |
| 2965 | const auto* _frm = reinterpret_cast<const underlying_wchar_t*>(frm); |
| 2966 | const auto* _frm_end = reinterpret_cast<const underlying_wchar_t*>(frm_end); |
| 2967 | const auto* _frm_nxt = _frm; |
| 2968 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 2969 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 2970 | uint8_t* _to_nxt = _to; |
| 2971 | result r = overload<ucs2_to_utf8, ucs4_to_utf8>(args&: _frm, args&: _frm_end, args&: _frm_nxt, args&: _to, args&: _to_end, args&: _to_nxt, args: __maxcode_, args: __mode_); |
| 2972 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2973 | to_nxt = to + (_to_nxt - _to); |
| 2974 | return r; |
| 2975 | } |
| 2976 | |
| 2977 | __codecvt_utf8<wchar_t>::result __codecvt_utf8<wchar_t>::do_in( |
| 2978 | state_type&, |
| 2979 | const extern_type* frm, |
| 2980 | const extern_type* frm_end, |
| 2981 | const extern_type*& frm_nxt, |
| 2982 | intern_type* to, |
| 2983 | intern_type* to_end, |
| 2984 | intern_type*& to_nxt) const { |
| 2985 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 2986 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 2987 | const uint8_t* _frm_nxt = _frm; |
| 2988 | auto* _to = reinterpret_cast<underlying_wchar_t*>(to); |
| 2989 | auto* _to_end = reinterpret_cast<underlying_wchar_t*>(to_end); |
| 2990 | auto* _to_nxt = _to; |
| 2991 | result r = overload<utf8_to_ucs2, utf8_to_ucs4>(args&: _frm, args&: _frm_end, args&: _frm_nxt, args&: _to, args&: _to_end, args&: _to_nxt, args: __maxcode_, args: __mode_); |
| 2992 | frm_nxt = frm + (_frm_nxt - _frm); |
| 2993 | to_nxt = to + (_to_nxt - _to); |
| 2994 | return r; |
| 2995 | } |
| 2996 | |
| 2997 | __codecvt_utf8<wchar_t>::result |
| 2998 | __codecvt_utf8<wchar_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 2999 | to_nxt = to; |
| 3000 | return noconv; |
| 3001 | } |
| 3002 | |
| 3003 | int __codecvt_utf8<wchar_t>::do_encoding() const noexcept { return 0; } |
| 3004 | |
| 3005 | bool __codecvt_utf8<wchar_t>::do_always_noconv() const noexcept { return false; } |
| 3006 | |
| 3007 | int __codecvt_utf8<wchar_t>::do_length( |
| 3008 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3009 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3010 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3011 | if constexpr (is_short_wchar) |
| 3012 | return utf8_to_ucs2_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3013 | else |
| 3014 | return utf8_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3015 | } |
| 3016 | |
| 3017 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3018 | int __codecvt_utf8<wchar_t>::do_max_length() const noexcept { |
| 3019 | if constexpr (is_short_wchar) |
| 3020 | return (__mode_ & consume_header) ? 6 : 3; |
| 3021 | else |
| 3022 | return (__mode_ & consume_header) ? 7 : 4; |
| 3023 | } |
| 3024 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3025 | |
| 3026 | // __codecvt_utf8<char16_t> |
| 3027 | |
| 3028 | __codecvt_utf8<char16_t>::result __codecvt_utf8<char16_t>::do_out( |
| 3029 | state_type&, |
| 3030 | const intern_type* frm, |
| 3031 | const intern_type* frm_end, |
| 3032 | const intern_type*& frm_nxt, |
| 3033 | extern_type* to, |
| 3034 | extern_type* to_end, |
| 3035 | extern_type*& to_nxt) const { |
| 3036 | const uint16_t* _frm = reinterpret_cast<const uint16_t*>(frm); |
| 3037 | const uint16_t* _frm_end = reinterpret_cast<const uint16_t*>(frm_end); |
| 3038 | const uint16_t* _frm_nxt = _frm; |
| 3039 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3040 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3041 | uint8_t* _to_nxt = _to; |
| 3042 | result r = ucs2_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3043 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3044 | to_nxt = to + (_to_nxt - _to); |
| 3045 | return r; |
| 3046 | } |
| 3047 | |
| 3048 | __codecvt_utf8<char16_t>::result __codecvt_utf8<char16_t>::do_in( |
| 3049 | state_type&, |
| 3050 | const extern_type* frm, |
| 3051 | const extern_type* frm_end, |
| 3052 | const extern_type*& frm_nxt, |
| 3053 | intern_type* to, |
| 3054 | intern_type* to_end, |
| 3055 | intern_type*& to_nxt) const { |
| 3056 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3057 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3058 | const uint8_t* _frm_nxt = _frm; |
| 3059 | uint16_t* _to = reinterpret_cast<uint16_t*>(to); |
| 3060 | uint16_t* _to_end = reinterpret_cast<uint16_t*>(to_end); |
| 3061 | uint16_t* _to_nxt = _to; |
| 3062 | result r = utf8_to_ucs2(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3063 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3064 | to_nxt = to + (_to_nxt - _to); |
| 3065 | return r; |
| 3066 | } |
| 3067 | |
| 3068 | __codecvt_utf8<char16_t>::result |
| 3069 | __codecvt_utf8<char16_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3070 | to_nxt = to; |
| 3071 | return noconv; |
| 3072 | } |
| 3073 | |
| 3074 | int __codecvt_utf8<char16_t>::do_encoding() const noexcept { return 0; } |
| 3075 | |
| 3076 | bool __codecvt_utf8<char16_t>::do_always_noconv() const noexcept { return false; } |
| 3077 | |
| 3078 | int __codecvt_utf8<char16_t>::do_length( |
| 3079 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3080 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3081 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3082 | return utf8_to_ucs2_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3083 | } |
| 3084 | |
| 3085 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3086 | int __codecvt_utf8<char16_t>::do_max_length() const noexcept { |
| 3087 | if (__mode_ & consume_header) |
| 3088 | return 6; |
| 3089 | return 3; |
| 3090 | } |
| 3091 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3092 | |
| 3093 | // __codecvt_utf8<char32_t> |
| 3094 | |
| 3095 | __codecvt_utf8<char32_t>::result __codecvt_utf8<char32_t>::do_out( |
| 3096 | state_type&, |
| 3097 | const intern_type* frm, |
| 3098 | const intern_type* frm_end, |
| 3099 | const intern_type*& frm_nxt, |
| 3100 | extern_type* to, |
| 3101 | extern_type* to_end, |
| 3102 | extern_type*& to_nxt) const { |
| 3103 | const uint32_t* _frm = reinterpret_cast<const uint32_t*>(frm); |
| 3104 | const uint32_t* _frm_end = reinterpret_cast<const uint32_t*>(frm_end); |
| 3105 | const uint32_t* _frm_nxt = _frm; |
| 3106 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3107 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3108 | uint8_t* _to_nxt = _to; |
| 3109 | result r = ucs4_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3110 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3111 | to_nxt = to + (_to_nxt - _to); |
| 3112 | return r; |
| 3113 | } |
| 3114 | |
| 3115 | __codecvt_utf8<char32_t>::result __codecvt_utf8<char32_t>::do_in( |
| 3116 | state_type&, |
| 3117 | const extern_type* frm, |
| 3118 | const extern_type* frm_end, |
| 3119 | const extern_type*& frm_nxt, |
| 3120 | intern_type* to, |
| 3121 | intern_type* to_end, |
| 3122 | intern_type*& to_nxt) const { |
| 3123 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3124 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3125 | const uint8_t* _frm_nxt = _frm; |
| 3126 | uint32_t* _to = reinterpret_cast<uint32_t*>(to); |
| 3127 | uint32_t* _to_end = reinterpret_cast<uint32_t*>(to_end); |
| 3128 | uint32_t* _to_nxt = _to; |
| 3129 | result r = utf8_to_ucs4(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3130 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3131 | to_nxt = to + (_to_nxt - _to); |
| 3132 | return r; |
| 3133 | } |
| 3134 | |
| 3135 | __codecvt_utf8<char32_t>::result |
| 3136 | __codecvt_utf8<char32_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3137 | to_nxt = to; |
| 3138 | return noconv; |
| 3139 | } |
| 3140 | |
| 3141 | int __codecvt_utf8<char32_t>::do_encoding() const noexcept { return 0; } |
| 3142 | |
| 3143 | bool __codecvt_utf8<char32_t>::do_always_noconv() const noexcept { return false; } |
| 3144 | |
| 3145 | int __codecvt_utf8<char32_t>::do_length( |
| 3146 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3147 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3148 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3149 | return utf8_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3150 | } |
| 3151 | |
| 3152 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3153 | int __codecvt_utf8<char32_t>::do_max_length() const noexcept { |
| 3154 | if (__mode_ & consume_header) |
| 3155 | return 7; |
| 3156 | return 4; |
| 3157 | } |
| 3158 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3159 | |
| 3160 | // __codecvt_utf16<wchar_t, false> |
| 3161 | |
| 3162 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3163 | __codecvt_utf16<wchar_t, false>::result __codecvt_utf16<wchar_t, false>::do_out( |
| 3164 | state_type&, |
| 3165 | const intern_type* frm, |
| 3166 | const intern_type* frm_end, |
| 3167 | const intern_type*& frm_nxt, |
| 3168 | extern_type* to, |
| 3169 | extern_type* to_end, |
| 3170 | extern_type*& to_nxt) const { |
| 3171 | const auto* _frm = reinterpret_cast<const underlying_wchar_t*>(frm); |
| 3172 | const auto* _frm_end = reinterpret_cast<const underlying_wchar_t*>(frm_end); |
| 3173 | const auto* _frm_nxt = _frm; |
| 3174 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3175 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3176 | uint8_t* _to_nxt = _to; |
| 3177 | result r = |
| 3178 | overload<ucs2_to_utf16be, ucs4_to_utf16be>(args&: _frm, args&: _frm_end, args&: _frm_nxt, args&: _to, args&: _to_end, args&: _to_nxt, args: __maxcode_, args: __mode_); |
| 3179 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3180 | to_nxt = to + (_to_nxt - _to); |
| 3181 | return r; |
| 3182 | } |
| 3183 | |
| 3184 | __codecvt_utf16<wchar_t, false>::result __codecvt_utf16<wchar_t, false>::do_in( |
| 3185 | state_type&, |
| 3186 | const extern_type* frm, |
| 3187 | const extern_type* frm_end, |
| 3188 | const extern_type*& frm_nxt, |
| 3189 | intern_type* to, |
| 3190 | intern_type* to_end, |
| 3191 | intern_type*& to_nxt) const { |
| 3192 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3193 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3194 | const uint8_t* _frm_nxt = _frm; |
| 3195 | auto* _to = reinterpret_cast<underlying_wchar_t*>(to); |
| 3196 | auto* _to_end = reinterpret_cast<underlying_wchar_t*>(to_end); |
| 3197 | auto* _to_nxt = _to; |
| 3198 | result r = |
| 3199 | overload<utf16be_to_ucs2, utf16be_to_ucs4>(args&: _frm, args&: _frm_end, args&: _frm_nxt, args&: _to, args&: _to_end, args&: _to_nxt, args: __maxcode_, args: __mode_); |
| 3200 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3201 | to_nxt = to + (_to_nxt - _to); |
| 3202 | return r; |
| 3203 | } |
| 3204 | |
| 3205 | __codecvt_utf16<wchar_t, false>::result |
| 3206 | __codecvt_utf16<wchar_t, false>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3207 | to_nxt = to; |
| 3208 | return noconv; |
| 3209 | } |
| 3210 | |
| 3211 | int __codecvt_utf16<wchar_t, false>::do_encoding() const noexcept { return 0; } |
| 3212 | |
| 3213 | bool __codecvt_utf16<wchar_t, false>::do_always_noconv() const noexcept { return false; } |
| 3214 | |
| 3215 | int __codecvt_utf16<wchar_t, false>::do_length( |
| 3216 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3217 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3218 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3219 | if constexpr (is_short_wchar) |
| 3220 | return utf16be_to_ucs2_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3221 | else |
| 3222 | return utf16be_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3223 | } |
| 3224 | |
| 3225 | int __codecvt_utf16<wchar_t, false>::do_max_length() const noexcept { |
| 3226 | if constexpr (is_short_wchar) |
| 3227 | return (__mode_ & consume_header) ? 4 : 2; |
| 3228 | else |
| 3229 | return (__mode_ & consume_header) ? 6 : 4; |
| 3230 | } |
| 3231 | |
| 3232 | // __codecvt_utf16<wchar_t, true> |
| 3233 | |
| 3234 | __codecvt_utf16<wchar_t, true>::result __codecvt_utf16<wchar_t, true>::do_out( |
| 3235 | state_type&, |
| 3236 | const intern_type* frm, |
| 3237 | const intern_type* frm_end, |
| 3238 | const intern_type*& frm_nxt, |
| 3239 | extern_type* to, |
| 3240 | extern_type* to_end, |
| 3241 | extern_type*& to_nxt) const { |
| 3242 | const auto* _frm = reinterpret_cast<const underlying_wchar_t*>(frm); |
| 3243 | const auto* _frm_end = reinterpret_cast<const underlying_wchar_t*>(frm_end); |
| 3244 | const auto* _frm_nxt = _frm; |
| 3245 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3246 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3247 | uint8_t* _to_nxt = _to; |
| 3248 | result r = |
| 3249 | overload<ucs2_to_utf16le, ucs4_to_utf16le>(args&: _frm, args&: _frm_end, args&: _frm_nxt, args&: _to, args&: _to_end, args&: _to_nxt, args: __maxcode_, args: __mode_); |
| 3250 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3251 | to_nxt = to + (_to_nxt - _to); |
| 3252 | return r; |
| 3253 | } |
| 3254 | |
| 3255 | __codecvt_utf16<wchar_t, true>::result __codecvt_utf16<wchar_t, true>::do_in( |
| 3256 | state_type&, |
| 3257 | const extern_type* frm, |
| 3258 | const extern_type* frm_end, |
| 3259 | const extern_type*& frm_nxt, |
| 3260 | intern_type* to, |
| 3261 | intern_type* to_end, |
| 3262 | intern_type*& to_nxt) const { |
| 3263 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3264 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3265 | const uint8_t* _frm_nxt = _frm; |
| 3266 | auto* _to = reinterpret_cast<underlying_wchar_t*>(to); |
| 3267 | auto* _to_end = reinterpret_cast<underlying_wchar_t*>(to_end); |
| 3268 | auto* _to_nxt = _to; |
| 3269 | result r = |
| 3270 | overload<utf16le_to_ucs2, utf16le_to_ucs4>(args&: _frm, args&: _frm_end, args&: _frm_nxt, args&: _to, args&: _to_end, args&: _to_nxt, args: __maxcode_, args: __mode_); |
| 3271 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3272 | to_nxt = to + (_to_nxt - _to); |
| 3273 | return r; |
| 3274 | } |
| 3275 | |
| 3276 | __codecvt_utf16<wchar_t, true>::result |
| 3277 | __codecvt_utf16<wchar_t, true>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3278 | to_nxt = to; |
| 3279 | return noconv; |
| 3280 | } |
| 3281 | |
| 3282 | int __codecvt_utf16<wchar_t, true>::do_encoding() const noexcept { return 0; } |
| 3283 | |
| 3284 | bool __codecvt_utf16<wchar_t, true>::do_always_noconv() const noexcept { return false; } |
| 3285 | |
| 3286 | int __codecvt_utf16<wchar_t, true>::do_length( |
| 3287 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3288 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3289 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3290 | if constexpr (is_short_wchar) |
| 3291 | return utf16le_to_ucs2_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3292 | else |
| 3293 | return utf16le_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3294 | } |
| 3295 | |
| 3296 | int __codecvt_utf16<wchar_t, true>::do_max_length() const noexcept { |
| 3297 | if constexpr (is_short_wchar) |
| 3298 | return (__mode_ & consume_header) ? 4 : 2; |
| 3299 | else |
| 3300 | return (__mode_ & consume_header) ? 6 : 4; |
| 3301 | } |
| 3302 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3303 | |
| 3304 | // __codecvt_utf16<char16_t, false> |
| 3305 | |
| 3306 | __codecvt_utf16<char16_t, false>::result __codecvt_utf16<char16_t, false>::do_out( |
| 3307 | state_type&, |
| 3308 | const intern_type* frm, |
| 3309 | const intern_type* frm_end, |
| 3310 | const intern_type*& frm_nxt, |
| 3311 | extern_type* to, |
| 3312 | extern_type* to_end, |
| 3313 | extern_type*& to_nxt) const { |
| 3314 | const uint16_t* _frm = reinterpret_cast<const uint16_t*>(frm); |
| 3315 | const uint16_t* _frm_end = reinterpret_cast<const uint16_t*>(frm_end); |
| 3316 | const uint16_t* _frm_nxt = _frm; |
| 3317 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3318 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3319 | uint8_t* _to_nxt = _to; |
| 3320 | result r = ucs2_to_utf16be(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3321 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3322 | to_nxt = to + (_to_nxt - _to); |
| 3323 | return r; |
| 3324 | } |
| 3325 | |
| 3326 | __codecvt_utf16<char16_t, false>::result __codecvt_utf16<char16_t, false>::do_in( |
| 3327 | state_type&, |
| 3328 | const extern_type* frm, |
| 3329 | const extern_type* frm_end, |
| 3330 | const extern_type*& frm_nxt, |
| 3331 | intern_type* to, |
| 3332 | intern_type* to_end, |
| 3333 | intern_type*& to_nxt) const { |
| 3334 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3335 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3336 | const uint8_t* _frm_nxt = _frm; |
| 3337 | uint16_t* _to = reinterpret_cast<uint16_t*>(to); |
| 3338 | uint16_t* _to_end = reinterpret_cast<uint16_t*>(to_end); |
| 3339 | uint16_t* _to_nxt = _to; |
| 3340 | result r = utf16be_to_ucs2(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3341 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3342 | to_nxt = to + (_to_nxt - _to); |
| 3343 | return r; |
| 3344 | } |
| 3345 | |
| 3346 | __codecvt_utf16<char16_t, false>::result |
| 3347 | __codecvt_utf16<char16_t, false>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3348 | to_nxt = to; |
| 3349 | return noconv; |
| 3350 | } |
| 3351 | |
| 3352 | int __codecvt_utf16<char16_t, false>::do_encoding() const noexcept { return 0; } |
| 3353 | |
| 3354 | bool __codecvt_utf16<char16_t, false>::do_always_noconv() const noexcept { return false; } |
| 3355 | |
| 3356 | int __codecvt_utf16<char16_t, false>::do_length( |
| 3357 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3358 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3359 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3360 | return utf16be_to_ucs2_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3361 | } |
| 3362 | |
| 3363 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3364 | int __codecvt_utf16<char16_t, false>::do_max_length() const noexcept { |
| 3365 | if (__mode_ & consume_header) |
| 3366 | return 4; |
| 3367 | return 2; |
| 3368 | } |
| 3369 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3370 | |
| 3371 | // __codecvt_utf16<char16_t, true> |
| 3372 | |
| 3373 | __codecvt_utf16<char16_t, true>::result __codecvt_utf16<char16_t, true>::do_out( |
| 3374 | state_type&, |
| 3375 | const intern_type* frm, |
| 3376 | const intern_type* frm_end, |
| 3377 | const intern_type*& frm_nxt, |
| 3378 | extern_type* to, |
| 3379 | extern_type* to_end, |
| 3380 | extern_type*& to_nxt) const { |
| 3381 | const uint16_t* _frm = reinterpret_cast<const uint16_t*>(frm); |
| 3382 | const uint16_t* _frm_end = reinterpret_cast<const uint16_t*>(frm_end); |
| 3383 | const uint16_t* _frm_nxt = _frm; |
| 3384 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3385 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3386 | uint8_t* _to_nxt = _to; |
| 3387 | result r = ucs2_to_utf16le(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3388 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3389 | to_nxt = to + (_to_nxt - _to); |
| 3390 | return r; |
| 3391 | } |
| 3392 | |
| 3393 | __codecvt_utf16<char16_t, true>::result __codecvt_utf16<char16_t, true>::do_in( |
| 3394 | state_type&, |
| 3395 | const extern_type* frm, |
| 3396 | const extern_type* frm_end, |
| 3397 | const extern_type*& frm_nxt, |
| 3398 | intern_type* to, |
| 3399 | intern_type* to_end, |
| 3400 | intern_type*& to_nxt) const { |
| 3401 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3402 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3403 | const uint8_t* _frm_nxt = _frm; |
| 3404 | uint16_t* _to = reinterpret_cast<uint16_t*>(to); |
| 3405 | uint16_t* _to_end = reinterpret_cast<uint16_t*>(to_end); |
| 3406 | uint16_t* _to_nxt = _to; |
| 3407 | result r = utf16le_to_ucs2(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3408 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3409 | to_nxt = to + (_to_nxt - _to); |
| 3410 | return r; |
| 3411 | } |
| 3412 | |
| 3413 | __codecvt_utf16<char16_t, true>::result |
| 3414 | __codecvt_utf16<char16_t, true>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3415 | to_nxt = to; |
| 3416 | return noconv; |
| 3417 | } |
| 3418 | |
| 3419 | int __codecvt_utf16<char16_t, true>::do_encoding() const noexcept { return 0; } |
| 3420 | |
| 3421 | bool __codecvt_utf16<char16_t, true>::do_always_noconv() const noexcept { return false; } |
| 3422 | |
| 3423 | int __codecvt_utf16<char16_t, true>::do_length( |
| 3424 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3425 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3426 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3427 | return utf16le_to_ucs2_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3428 | } |
| 3429 | |
| 3430 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3431 | int __codecvt_utf16<char16_t, true>::do_max_length() const noexcept { |
| 3432 | if (__mode_ & consume_header) |
| 3433 | return 4; |
| 3434 | return 2; |
| 3435 | } |
| 3436 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3437 | |
| 3438 | // __codecvt_utf16<char32_t, false> |
| 3439 | |
| 3440 | __codecvt_utf16<char32_t, false>::result __codecvt_utf16<char32_t, false>::do_out( |
| 3441 | state_type&, |
| 3442 | const intern_type* frm, |
| 3443 | const intern_type* frm_end, |
| 3444 | const intern_type*& frm_nxt, |
| 3445 | extern_type* to, |
| 3446 | extern_type* to_end, |
| 3447 | extern_type*& to_nxt) const { |
| 3448 | const uint32_t* _frm = reinterpret_cast<const uint32_t*>(frm); |
| 3449 | const uint32_t* _frm_end = reinterpret_cast<const uint32_t*>(frm_end); |
| 3450 | const uint32_t* _frm_nxt = _frm; |
| 3451 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3452 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3453 | uint8_t* _to_nxt = _to; |
| 3454 | result r = ucs4_to_utf16be(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3455 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3456 | to_nxt = to + (_to_nxt - _to); |
| 3457 | return r; |
| 3458 | } |
| 3459 | |
| 3460 | __codecvt_utf16<char32_t, false>::result __codecvt_utf16<char32_t, false>::do_in( |
| 3461 | state_type&, |
| 3462 | const extern_type* frm, |
| 3463 | const extern_type* frm_end, |
| 3464 | const extern_type*& frm_nxt, |
| 3465 | intern_type* to, |
| 3466 | intern_type* to_end, |
| 3467 | intern_type*& to_nxt) const { |
| 3468 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3469 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3470 | const uint8_t* _frm_nxt = _frm; |
| 3471 | uint32_t* _to = reinterpret_cast<uint32_t*>(to); |
| 3472 | uint32_t* _to_end = reinterpret_cast<uint32_t*>(to_end); |
| 3473 | uint32_t* _to_nxt = _to; |
| 3474 | result r = utf16be_to_ucs4(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3475 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3476 | to_nxt = to + (_to_nxt - _to); |
| 3477 | return r; |
| 3478 | } |
| 3479 | |
| 3480 | __codecvt_utf16<char32_t, false>::result |
| 3481 | __codecvt_utf16<char32_t, false>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3482 | to_nxt = to; |
| 3483 | return noconv; |
| 3484 | } |
| 3485 | |
| 3486 | int __codecvt_utf16<char32_t, false>::do_encoding() const noexcept { return 0; } |
| 3487 | |
| 3488 | bool __codecvt_utf16<char32_t, false>::do_always_noconv() const noexcept { return false; } |
| 3489 | |
| 3490 | int __codecvt_utf16<char32_t, false>::do_length( |
| 3491 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3492 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3493 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3494 | return utf16be_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3495 | } |
| 3496 | |
| 3497 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3498 | int __codecvt_utf16<char32_t, false>::do_max_length() const noexcept { |
| 3499 | if (__mode_ & consume_header) |
| 3500 | return 6; |
| 3501 | return 4; |
| 3502 | } |
| 3503 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3504 | |
| 3505 | // __codecvt_utf16<char32_t, true> |
| 3506 | |
| 3507 | __codecvt_utf16<char32_t, true>::result __codecvt_utf16<char32_t, true>::do_out( |
| 3508 | state_type&, |
| 3509 | const intern_type* frm, |
| 3510 | const intern_type* frm_end, |
| 3511 | const intern_type*& frm_nxt, |
| 3512 | extern_type* to, |
| 3513 | extern_type* to_end, |
| 3514 | extern_type*& to_nxt) const { |
| 3515 | const uint32_t* _frm = reinterpret_cast<const uint32_t*>(frm); |
| 3516 | const uint32_t* _frm_end = reinterpret_cast<const uint32_t*>(frm_end); |
| 3517 | const uint32_t* _frm_nxt = _frm; |
| 3518 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3519 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3520 | uint8_t* _to_nxt = _to; |
| 3521 | result r = ucs4_to_utf16le(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3522 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3523 | to_nxt = to + (_to_nxt - _to); |
| 3524 | return r; |
| 3525 | } |
| 3526 | |
| 3527 | __codecvt_utf16<char32_t, true>::result __codecvt_utf16<char32_t, true>::do_in( |
| 3528 | state_type&, |
| 3529 | const extern_type* frm, |
| 3530 | const extern_type* frm_end, |
| 3531 | const extern_type*& frm_nxt, |
| 3532 | intern_type* to, |
| 3533 | intern_type* to_end, |
| 3534 | intern_type*& to_nxt) const { |
| 3535 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3536 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3537 | const uint8_t* _frm_nxt = _frm; |
| 3538 | uint32_t* _to = reinterpret_cast<uint32_t*>(to); |
| 3539 | uint32_t* _to_end = reinterpret_cast<uint32_t*>(to_end); |
| 3540 | uint32_t* _to_nxt = _to; |
| 3541 | result r = utf16le_to_ucs4(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3542 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3543 | to_nxt = to + (_to_nxt - _to); |
| 3544 | return r; |
| 3545 | } |
| 3546 | |
| 3547 | __codecvt_utf16<char32_t, true>::result |
| 3548 | __codecvt_utf16<char32_t, true>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3549 | to_nxt = to; |
| 3550 | return noconv; |
| 3551 | } |
| 3552 | |
| 3553 | int __codecvt_utf16<char32_t, true>::do_encoding() const noexcept { return 0; } |
| 3554 | |
| 3555 | bool __codecvt_utf16<char32_t, true>::do_always_noconv() const noexcept { return false; } |
| 3556 | |
| 3557 | int __codecvt_utf16<char32_t, true>::do_length( |
| 3558 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3559 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3560 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3561 | return utf16le_to_ucs4_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3562 | } |
| 3563 | |
| 3564 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3565 | int __codecvt_utf16<char32_t, true>::do_max_length() const noexcept { |
| 3566 | if (__mode_ & consume_header) |
| 3567 | return 6; |
| 3568 | return 4; |
| 3569 | } |
| 3570 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3571 | |
| 3572 | // __codecvt_utf8_utf16<wchar_t> |
| 3573 | |
| 3574 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3575 | __codecvt_utf8_utf16<wchar_t>::result __codecvt_utf8_utf16<wchar_t>::do_out( |
| 3576 | state_type&, |
| 3577 | const intern_type* frm, |
| 3578 | const intern_type* frm_end, |
| 3579 | const intern_type*& frm_nxt, |
| 3580 | extern_type* to, |
| 3581 | extern_type* to_end, |
| 3582 | extern_type*& to_nxt) const { |
| 3583 | const auto* _frm = reinterpret_cast<const underlying_wchar_t*>(frm); |
| 3584 | const auto* _frm_end = reinterpret_cast<const underlying_wchar_t*>(frm_end); |
| 3585 | const auto* _frm_nxt = _frm; |
| 3586 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3587 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3588 | uint8_t* _to_nxt = _to; |
| 3589 | result r = utf16_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3590 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3591 | to_nxt = to + (_to_nxt - _to); |
| 3592 | return r; |
| 3593 | } |
| 3594 | |
| 3595 | __codecvt_utf8_utf16<wchar_t>::result __codecvt_utf8_utf16<wchar_t>::do_in( |
| 3596 | state_type&, |
| 3597 | const extern_type* frm, |
| 3598 | const extern_type* frm_end, |
| 3599 | const extern_type*& frm_nxt, |
| 3600 | intern_type* to, |
| 3601 | intern_type* to_end, |
| 3602 | intern_type*& to_nxt) const { |
| 3603 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3604 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3605 | const uint8_t* _frm_nxt = _frm; |
| 3606 | auto* _to = reinterpret_cast<underlying_wchar_t*>(to); |
| 3607 | auto* _to_end = reinterpret_cast<underlying_wchar_t*>(to_end); |
| 3608 | auto* _to_nxt = _to; |
| 3609 | result r = utf8_to_utf16(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3610 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3611 | to_nxt = to + (_to_nxt - _to); |
| 3612 | return r; |
| 3613 | } |
| 3614 | |
| 3615 | __codecvt_utf8_utf16<wchar_t>::result |
| 3616 | __codecvt_utf8_utf16<wchar_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3617 | to_nxt = to; |
| 3618 | return noconv; |
| 3619 | } |
| 3620 | |
| 3621 | int __codecvt_utf8_utf16<wchar_t>::do_encoding() const noexcept { return 0; } |
| 3622 | |
| 3623 | bool __codecvt_utf8_utf16<wchar_t>::do_always_noconv() const noexcept { return false; } |
| 3624 | |
| 3625 | int __codecvt_utf8_utf16<wchar_t>::do_length( |
| 3626 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3627 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3628 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3629 | return utf8_to_utf16_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3630 | } |
| 3631 | |
| 3632 | int __codecvt_utf8_utf16<wchar_t>::do_max_length() const noexcept { |
| 3633 | if (__mode_ & consume_header) |
| 3634 | return 7; |
| 3635 | return 4; |
| 3636 | } |
| 3637 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3638 | |
| 3639 | // __codecvt_utf8_utf16<char16_t> |
| 3640 | |
| 3641 | __codecvt_utf8_utf16<char16_t>::result __codecvt_utf8_utf16<char16_t>::do_out( |
| 3642 | state_type&, |
| 3643 | const intern_type* frm, |
| 3644 | const intern_type* frm_end, |
| 3645 | const intern_type*& frm_nxt, |
| 3646 | extern_type* to, |
| 3647 | extern_type* to_end, |
| 3648 | extern_type*& to_nxt) const { |
| 3649 | const uint16_t* _frm = reinterpret_cast<const uint16_t*>(frm); |
| 3650 | const uint16_t* _frm_end = reinterpret_cast<const uint16_t*>(frm_end); |
| 3651 | const uint16_t* _frm_nxt = _frm; |
| 3652 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3653 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3654 | uint8_t* _to_nxt = _to; |
| 3655 | result r = utf16_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3656 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3657 | to_nxt = to + (_to_nxt - _to); |
| 3658 | return r; |
| 3659 | } |
| 3660 | |
| 3661 | __codecvt_utf8_utf16<char16_t>::result __codecvt_utf8_utf16<char16_t>::do_in( |
| 3662 | state_type&, |
| 3663 | const extern_type* frm, |
| 3664 | const extern_type* frm_end, |
| 3665 | const extern_type*& frm_nxt, |
| 3666 | intern_type* to, |
| 3667 | intern_type* to_end, |
| 3668 | intern_type*& to_nxt) const { |
| 3669 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3670 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3671 | const uint8_t* _frm_nxt = _frm; |
| 3672 | uint16_t* _to = reinterpret_cast<uint16_t*>(to); |
| 3673 | uint16_t* _to_end = reinterpret_cast<uint16_t*>(to_end); |
| 3674 | uint16_t* _to_nxt = _to; |
| 3675 | result r = utf8_to_utf16(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3676 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3677 | to_nxt = to + (_to_nxt - _to); |
| 3678 | return r; |
| 3679 | } |
| 3680 | |
| 3681 | __codecvt_utf8_utf16<char16_t>::result |
| 3682 | __codecvt_utf8_utf16<char16_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3683 | to_nxt = to; |
| 3684 | return noconv; |
| 3685 | } |
| 3686 | |
| 3687 | int __codecvt_utf8_utf16<char16_t>::do_encoding() const noexcept { return 0; } |
| 3688 | |
| 3689 | bool __codecvt_utf8_utf16<char16_t>::do_always_noconv() const noexcept { return false; } |
| 3690 | |
| 3691 | int __codecvt_utf8_utf16<char16_t>::do_length( |
| 3692 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3693 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3694 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3695 | return utf8_to_utf16_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3696 | } |
| 3697 | |
| 3698 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3699 | int __codecvt_utf8_utf16<char16_t>::do_max_length() const noexcept { |
| 3700 | if (__mode_ & consume_header) |
| 3701 | return 7; |
| 3702 | return 4; |
| 3703 | } |
| 3704 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3705 | |
| 3706 | // __codecvt_utf8_utf16<char32_t> |
| 3707 | |
| 3708 | __codecvt_utf8_utf16<char32_t>::result __codecvt_utf8_utf16<char32_t>::do_out( |
| 3709 | state_type&, |
| 3710 | const intern_type* frm, |
| 3711 | const intern_type* frm_end, |
| 3712 | const intern_type*& frm_nxt, |
| 3713 | extern_type* to, |
| 3714 | extern_type* to_end, |
| 3715 | extern_type*& to_nxt) const { |
| 3716 | const uint32_t* _frm = reinterpret_cast<const uint32_t*>(frm); |
| 3717 | const uint32_t* _frm_end = reinterpret_cast<const uint32_t*>(frm_end); |
| 3718 | const uint32_t* _frm_nxt = _frm; |
| 3719 | uint8_t* _to = reinterpret_cast<uint8_t*>(to); |
| 3720 | uint8_t* _to_end = reinterpret_cast<uint8_t*>(to_end); |
| 3721 | uint8_t* _to_nxt = _to; |
| 3722 | result r = utf16_to_utf8(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3723 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3724 | to_nxt = to + (_to_nxt - _to); |
| 3725 | return r; |
| 3726 | } |
| 3727 | |
| 3728 | __codecvt_utf8_utf16<char32_t>::result __codecvt_utf8_utf16<char32_t>::do_in( |
| 3729 | state_type&, |
| 3730 | const extern_type* frm, |
| 3731 | const extern_type* frm_end, |
| 3732 | const extern_type*& frm_nxt, |
| 3733 | intern_type* to, |
| 3734 | intern_type* to_end, |
| 3735 | intern_type*& to_nxt) const { |
| 3736 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3737 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3738 | const uint8_t* _frm_nxt = _frm; |
| 3739 | uint32_t* _to = reinterpret_cast<uint32_t*>(to); |
| 3740 | uint32_t* _to_end = reinterpret_cast<uint32_t*>(to_end); |
| 3741 | uint32_t* _to_nxt = _to; |
| 3742 | result r = utf8_to_utf16(frm: _frm, frm_end: _frm_end, frm_nxt&: _frm_nxt, to: _to, to_end: _to_end, to_nxt&: _to_nxt, Maxcode: __maxcode_, mode: __mode_); |
| 3743 | frm_nxt = frm + (_frm_nxt - _frm); |
| 3744 | to_nxt = to + (_to_nxt - _to); |
| 3745 | return r; |
| 3746 | } |
| 3747 | |
| 3748 | __codecvt_utf8_utf16<char32_t>::result |
| 3749 | __codecvt_utf8_utf16<char32_t>::do_unshift(state_type&, extern_type* to, extern_type*, extern_type*& to_nxt) const { |
| 3750 | to_nxt = to; |
| 3751 | return noconv; |
| 3752 | } |
| 3753 | |
| 3754 | int __codecvt_utf8_utf16<char32_t>::do_encoding() const noexcept { return 0; } |
| 3755 | |
| 3756 | bool __codecvt_utf8_utf16<char32_t>::do_always_noconv() const noexcept { return false; } |
| 3757 | |
| 3758 | int __codecvt_utf8_utf16<char32_t>::do_length( |
| 3759 | state_type&, const extern_type* frm, const extern_type* frm_end, size_t mx) const { |
| 3760 | const uint8_t* _frm = reinterpret_cast<const uint8_t*>(frm); |
| 3761 | const uint8_t* _frm_end = reinterpret_cast<const uint8_t*>(frm_end); |
| 3762 | return utf8_to_utf16_length(frm: _frm, frm_end: _frm_end, mx, Maxcode: __maxcode_, mode: __mode_); |
| 3763 | } |
| 3764 | |
| 3765 | _LIBCPP_SUPPRESS_DEPRECATED_PUSH |
| 3766 | int __codecvt_utf8_utf16<char32_t>::do_max_length() const noexcept { |
| 3767 | if (__mode_ & consume_header) |
| 3768 | return 7; |
| 3769 | return 4; |
| 3770 | } |
| 3771 | _LIBCPP_SUPPRESS_DEPRECATED_POP |
| 3772 | |
| 3773 | // __narrow_to_utf8<16> |
| 3774 | |
| 3775 | __narrow_to_utf8<16>::~__narrow_to_utf8() {} |
| 3776 | |
| 3777 | // __narrow_to_utf8<32> |
| 3778 | |
| 3779 | __narrow_to_utf8<32>::~__narrow_to_utf8() {} |
| 3780 | |
| 3781 | // __widen_from_utf8<16> |
| 3782 | |
| 3783 | __widen_from_utf8<16>::~__widen_from_utf8() {} |
| 3784 | |
| 3785 | // __widen_from_utf8<32> |
| 3786 | |
| 3787 | __widen_from_utf8<32>::~__widen_from_utf8() {} |
| 3788 | |
| 3789 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3790 | static bool checked_string_to_wchar_convert(wchar_t& dest, const char* ptr, __locale::__locale_t loc) { |
| 3791 | if (*ptr == '\0') |
| 3792 | return false; |
| 3793 | mbstate_t mb = {}; |
| 3794 | wchar_t out; |
| 3795 | size_t ret = __locale::__mbrtowc(pwc: &out, s: ptr, n: strlen(s: ptr), ps: &mb, loc: loc); |
| 3796 | if (ret == static_cast<size_t>(-1) || ret == static_cast<size_t>(-2)) { |
| 3797 | return false; |
| 3798 | } |
| 3799 | dest = out; |
| 3800 | return true; |
| 3801 | } |
| 3802 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3803 | |
| 3804 | #if !_LIBCPP_HAS_WIDE_CHARACTERS |
| 3805 | static bool is_narrow_non_breaking_space(const char* ptr) { |
| 3806 | // https://www.fileformat.info/info/unicode/char/202f/index.htm |
| 3807 | return ptr[0] == '\xe2' && ptr[1] == '\x80' && ptr[2] == '\xaf'; |
| 3808 | } |
| 3809 | |
| 3810 | static bool is_non_breaking_space(const char* ptr) { |
| 3811 | // https://www.fileformat.info/info/unicode/char/a0/index.htm |
| 3812 | return ptr[0] == '\xc2' && ptr[1] == '\xa0'; |
| 3813 | } |
| 3814 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3815 | |
| 3816 | static bool checked_string_to_char_convert(char& dest, const char* ptr, __locale::__locale_t __loc) { |
| 3817 | if (*ptr == '\0') |
| 3818 | return false; |
| 3819 | if (!ptr[1]) { |
| 3820 | dest = *ptr; |
| 3821 | return true; |
| 3822 | } |
| 3823 | |
| 3824 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3825 | // First convert the MBS into a wide char then attempt to narrow it using |
| 3826 | // wctob_l. |
| 3827 | wchar_t wout; |
| 3828 | if (!checked_string_to_wchar_convert(dest&: wout, ptr, loc: __loc)) |
| 3829 | return false; |
| 3830 | int res; |
| 3831 | if ((res = __locale::__wctob(c: wout, __loc)) != char_traits<char>::eof()) { |
| 3832 | dest = res; |
| 3833 | return true; |
| 3834 | } |
| 3835 | // FIXME: Work around specific multibyte sequences that we can reasonably |
| 3836 | // translate into a different single byte. |
| 3837 | switch (wout) { |
| 3838 | case L'\u202F': // narrow non-breaking space |
| 3839 | case L'\u00A0': // non-breaking space |
| 3840 | dest = ' '; |
| 3841 | return true; |
| 3842 | default: |
| 3843 | return false; |
| 3844 | } |
| 3845 | #else // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3846 | // FIXME: Work around specific multibyte sequences that we can reasonably |
| 3847 | // translate into a different single byte. |
| 3848 | if (is_narrow_non_breaking_space(ptr) || is_non_breaking_space(ptr)) { |
| 3849 | dest = ' '; |
| 3850 | return true; |
| 3851 | } |
| 3852 | |
| 3853 | return false; |
| 3854 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3855 | __libcpp_unreachable(); |
| 3856 | } |
| 3857 | |
| 3858 | // numpunct<char> && numpunct<wchar_t> |
| 3859 | |
| 3860 | constinit locale::id numpunct<char>::id; |
| 3861 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3862 | constinit locale::id numpunct<wchar_t>::id; |
| 3863 | #endif |
| 3864 | |
| 3865 | numpunct<char>::numpunct(size_t refs) : locale::facet(refs), __decimal_point_('.'), __thousands_sep_(',') {} |
| 3866 | |
| 3867 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3868 | numpunct<wchar_t>::numpunct(size_t refs) : locale::facet(refs), __decimal_point_(L'.'), __thousands_sep_(L',') {} |
| 3869 | #endif |
| 3870 | |
| 3871 | numpunct<char>::~numpunct() {} |
| 3872 | |
| 3873 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3874 | numpunct<wchar_t>::~numpunct() {} |
| 3875 | #endif |
| 3876 | |
| 3877 | char numpunct< char >::do_decimal_point() const { return __decimal_point_; } |
| 3878 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3879 | wchar_t numpunct<wchar_t>::do_decimal_point() const { return __decimal_point_; } |
| 3880 | #endif |
| 3881 | |
| 3882 | char numpunct< char >::do_thousands_sep() const { return __thousands_sep_; } |
| 3883 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3884 | wchar_t numpunct<wchar_t>::do_thousands_sep() const { return __thousands_sep_; } |
| 3885 | #endif |
| 3886 | |
| 3887 | string numpunct< char >::do_grouping() const { return __grouping_; } |
| 3888 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3889 | string numpunct<wchar_t>::do_grouping() const { return __grouping_; } |
| 3890 | #endif |
| 3891 | |
| 3892 | string numpunct< char >::do_truename() const { return "true"; } |
| 3893 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3894 | wstring numpunct<wchar_t>::do_truename() const { return L"true"; } |
| 3895 | #endif |
| 3896 | |
| 3897 | string numpunct< char >::do_falsename() const { return "false"; } |
| 3898 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3899 | wstring numpunct<wchar_t>::do_falsename() const { return L"false"; } |
| 3900 | #endif |
| 3901 | |
| 3902 | // numpunct_byname<char> |
| 3903 | |
| 3904 | numpunct_byname<char>::numpunct_byname(const char* nm, size_t refs) : numpunct<char>(refs) { |
| 3905 | typedef numpunct<char> base; |
| 3906 | if (strcmp(s1: nm, s2: "C") != 0) { |
| 3907 | __libcpp_unique_locale loc(nm); |
| 3908 | if (!loc) |
| 3909 | std::__throw_runtime_error( |
| 3910 | ("numpunct_byname<char>::numpunct_byname" |
| 3911 | " failed to construct for "+ |
| 3912 | string(nm)) |
| 3913 | .c_str()); |
| 3914 | |
| 3915 | __locale::__lconv_t* lc = __locale::__localeconv(loc&: loc.get()); |
| 3916 | if (!checked_string_to_char_convert(dest&: __decimal_point_, ptr: lc->decimal_point, loc: loc.get())) |
| 3917 | __decimal_point_ = base::do_decimal_point(); |
| 3918 | if (!checked_string_to_char_convert(dest&: __thousands_sep_, ptr: lc->thousands_sep, loc: loc.get())) |
| 3919 | __thousands_sep_ = base::do_thousands_sep(); |
| 3920 | __grouping_ = lc->grouping; |
| 3921 | // localization for truename and falsename is not available |
| 3922 | } |
| 3923 | } |
| 3924 | |
| 3925 | numpunct_byname<char>::numpunct_byname(const string& nm, size_t refs) : numpunct_byname<char>(nm.c_str(), refs) {} |
| 3926 | |
| 3927 | numpunct_byname<char>::~numpunct_byname() {} |
| 3928 | |
| 3929 | // numpunct_byname<wchar_t> |
| 3930 | |
| 3931 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 3932 | numpunct_byname<wchar_t>::numpunct_byname(const char* nm, size_t refs) : numpunct<wchar_t>(refs) { |
| 3933 | if (strcmp(s1: nm, s2: "C") != 0) { |
| 3934 | __libcpp_unique_locale loc(nm); |
| 3935 | if (!loc) |
| 3936 | std::__throw_runtime_error( |
| 3937 | ("numpunct_byname<wchar_t>::numpunct_byname" |
| 3938 | " failed to construct for "+ |
| 3939 | string(nm)) |
| 3940 | .c_str()); |
| 3941 | |
| 3942 | __locale::__lconv_t* lc = __locale::__localeconv(loc&: loc.get()); |
| 3943 | checked_string_to_wchar_convert(dest&: __decimal_point_, ptr: lc->decimal_point, loc: loc.get()); |
| 3944 | checked_string_to_wchar_convert(dest&: __thousands_sep_, ptr: lc->thousands_sep, loc: loc.get()); |
| 3945 | __grouping_ = lc->grouping; |
| 3946 | // localization for truename and falsename is not available |
| 3947 | } |
| 3948 | } |
| 3949 | |
| 3950 | numpunct_byname<wchar_t>::numpunct_byname(const string& nm, size_t refs) : numpunct_byname<wchar_t>(nm.c_str(), refs) {} |
| 3951 | |
| 3952 | numpunct_byname<wchar_t>::~numpunct_byname() {} |
| 3953 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 3954 | |
| 3955 | // num_get helpers |
| 3956 | |
| 3957 | int __num_get_base::__get_base(ios_base& iob) { |
| 3958 | ios_base::fmtflags __basefield = iob.flags() & ios_base::basefield; |
| 3959 | if (__basefield == ios_base::oct) |
| 3960 | return 8; |
| 3961 | else if (__basefield == ios_base::hex) |
| 3962 | return 16; |
| 3963 | else if (__basefield == 0) |
| 3964 | return 0; |
| 3965 | return 10; |
| 3966 | } |
| 3967 | |
| 3968 | const char __num_get_base::__src[33] = "0123456789abcdefABCDEFxX+-pPiInN"; |
| 3969 | |
| 3970 | void __check_grouping(const string& __grouping, unsigned* __g, unsigned* __g_end, ios_base::iostate& __err) { |
| 3971 | // if the grouping pattern is empty _or_ there are no grouping bits, then do nothing |
| 3972 | // we always have at least a single entry in [__g, __g_end); the end of the input sequence |
| 3973 | if (__grouping.size() != 0 && __g_end - __g > 1) { |
| 3974 | reverse(first: __g, last: __g_end); |
| 3975 | const char* __ig = __grouping.data(); |
| 3976 | const char* __eg = __ig + __grouping.size(); |
| 3977 | for (unsigned* __r = __g; __r < __g_end - 1; ++__r) { |
| 3978 | if (0 < *__ig && *__ig < numeric_limits<char>::max()) { |
| 3979 | if (static_cast<unsigned>(*__ig) != *__r) { |
| 3980 | __err = ios_base::failbit; |
| 3981 | return; |
| 3982 | } |
| 3983 | } |
| 3984 | if (__eg - __ig > 1) |
| 3985 | ++__ig; |
| 3986 | } |
| 3987 | if (0 < *__ig && *__ig < numeric_limits<char>::max()) { |
| 3988 | if (static_cast<unsigned>(*__ig) < __g_end[-1] || __g_end[-1] == 0) |
| 3989 | __err = ios_base::failbit; |
| 3990 | } |
| 3991 | } |
| 3992 | } |
| 3993 | |
| 3994 | void __num_put_base::__format_int(char* __fmtp, const char* __len, bool __signd, ios_base::fmtflags __flags) { |
| 3995 | if ((__flags & ios_base::showpos) && (__flags & ios_base::basefield) != ios_base::oct && |
| 3996 | (__flags & ios_base::basefield) != ios_base::hex && __signd) |
| 3997 | *__fmtp++ = '+'; |
| 3998 | if (__flags & ios_base::showbase) |
| 3999 | *__fmtp++ = '#'; |
| 4000 | while (*__len) |
| 4001 | *__fmtp++ = *__len++; |
| 4002 | if ((__flags & ios_base::basefield) == ios_base::oct) |
| 4003 | *__fmtp = 'o'; |
| 4004 | else if ((__flags & ios_base::basefield) == ios_base::hex) { |
| 4005 | if (__flags & ios_base::uppercase) |
| 4006 | *__fmtp = 'X'; |
| 4007 | else |
| 4008 | *__fmtp = 'x'; |
| 4009 | } else if (__signd) |
| 4010 | *__fmtp = 'd'; |
| 4011 | else |
| 4012 | *__fmtp = 'u'; |
| 4013 | } |
| 4014 | |
| 4015 | bool __num_put_base::__format_float(char* __fmtp, const char* __len, ios_base::fmtflags __flags) { |
| 4016 | bool specify_precision = true; |
| 4017 | if (__flags & ios_base::showpos) |
| 4018 | *__fmtp++ = '+'; |
| 4019 | if (__flags & ios_base::showpoint) |
| 4020 | *__fmtp++ = '#'; |
| 4021 | ios_base::fmtflags floatfield = __flags & ios_base::floatfield; |
| 4022 | bool uppercase = (__flags & ios_base::uppercase) != 0; |
| 4023 | if (floatfield == (ios_base::fixed | ios_base::scientific)) |
| 4024 | specify_precision = false; |
| 4025 | else { |
| 4026 | *__fmtp++ = '.'; |
| 4027 | *__fmtp++ = '*'; |
| 4028 | } |
| 4029 | while (*__len) |
| 4030 | *__fmtp++ = *__len++; |
| 4031 | if (floatfield == ios_base::fixed) { |
| 4032 | if (uppercase) |
| 4033 | *__fmtp = 'F'; |
| 4034 | else |
| 4035 | *__fmtp = 'f'; |
| 4036 | } else if (floatfield == ios_base::scientific) { |
| 4037 | if (uppercase) |
| 4038 | *__fmtp = 'E'; |
| 4039 | else |
| 4040 | *__fmtp = 'e'; |
| 4041 | } else if (floatfield == (ios_base::fixed | ios_base::scientific)) { |
| 4042 | if (uppercase) |
| 4043 | *__fmtp = 'A'; |
| 4044 | else |
| 4045 | *__fmtp = 'a'; |
| 4046 | } else { |
| 4047 | if (uppercase) |
| 4048 | *__fmtp = 'G'; |
| 4049 | else |
| 4050 | *__fmtp = 'g'; |
| 4051 | } |
| 4052 | return specify_precision; |
| 4053 | } |
| 4054 | |
| 4055 | char* __num_put_base::__identify_padding(char* __nb, char* __ne, const ios_base& __iob) { |
| 4056 | switch (__iob.flags() & ios_base::adjustfield) { |
| 4057 | case ios_base::internal: |
| 4058 | if (__nb[0] == '-' || __nb[0] == '+') |
| 4059 | return __nb + 1; |
| 4060 | if (__ne - __nb >= 2 && __nb[0] == '0' && (__nb[1] == 'x' || __nb[1] == 'X')) |
| 4061 | return __nb + 2; |
| 4062 | break; |
| 4063 | case ios_base::left: |
| 4064 | return __ne; |
| 4065 | case ios_base::right: |
| 4066 | default: |
| 4067 | break; |
| 4068 | } |
| 4069 | return __nb; |
| 4070 | } |
| 4071 | |
| 4072 | // time_get |
| 4073 | |
| 4074 | static string* init_weeks() { |
| 4075 | static string weeks[14]; |
| 4076 | weeks[0] = "Sunday"; |
| 4077 | weeks[1] = "Monday"; |
| 4078 | weeks[2] = "Tuesday"; |
| 4079 | weeks[3] = "Wednesday"; |
| 4080 | weeks[4] = "Thursday"; |
| 4081 | weeks[5] = "Friday"; |
| 4082 | weeks[6] = "Saturday"; |
| 4083 | weeks[7] = "Sun"; |
| 4084 | weeks[8] = "Mon"; |
| 4085 | weeks[9] = "Tue"; |
| 4086 | weeks[10] = "Wed"; |
| 4087 | weeks[11] = "Thu"; |
| 4088 | weeks[12] = "Fri"; |
| 4089 | weeks[13] = "Sat"; |
| 4090 | return weeks; |
| 4091 | } |
| 4092 | |
| 4093 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4094 | static wstring* init_wweeks() { |
| 4095 | static wstring weeks[14]; |
| 4096 | weeks[0] = L"Sunday"; |
| 4097 | weeks[1] = L"Monday"; |
| 4098 | weeks[2] = L"Tuesday"; |
| 4099 | weeks[3] = L"Wednesday"; |
| 4100 | weeks[4] = L"Thursday"; |
| 4101 | weeks[5] = L"Friday"; |
| 4102 | weeks[6] = L"Saturday"; |
| 4103 | weeks[7] = L"Sun"; |
| 4104 | weeks[8] = L"Mon"; |
| 4105 | weeks[9] = L"Tue"; |
| 4106 | weeks[10] = L"Wed"; |
| 4107 | weeks[11] = L"Thu"; |
| 4108 | weeks[12] = L"Fri"; |
| 4109 | weeks[13] = L"Sat"; |
| 4110 | return weeks; |
| 4111 | } |
| 4112 | #endif |
| 4113 | |
| 4114 | template <> |
| 4115 | const string* __time_get_c_storage<char>::__weeks() const { |
| 4116 | static const string* weeks = init_weeks(); |
| 4117 | return weeks; |
| 4118 | } |
| 4119 | |
| 4120 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4121 | template <> |
| 4122 | const wstring* __time_get_c_storage<wchar_t>::__weeks() const { |
| 4123 | static const wstring* weeks = init_wweeks(); |
| 4124 | return weeks; |
| 4125 | } |
| 4126 | #endif |
| 4127 | |
| 4128 | static string* init_months() { |
| 4129 | static string months[24]; |
| 4130 | months[0] = "January"; |
| 4131 | months[1] = "February"; |
| 4132 | months[2] = "March"; |
| 4133 | months[3] = "April"; |
| 4134 | months[4] = "May"; |
| 4135 | months[5] = "June"; |
| 4136 | months[6] = "July"; |
| 4137 | months[7] = "August"; |
| 4138 | months[8] = "September"; |
| 4139 | months[9] = "October"; |
| 4140 | months[10] = "November"; |
| 4141 | months[11] = "December"; |
| 4142 | months[12] = "Jan"; |
| 4143 | months[13] = "Feb"; |
| 4144 | months[14] = "Mar"; |
| 4145 | months[15] = "Apr"; |
| 4146 | months[16] = "May"; |
| 4147 | months[17] = "Jun"; |
| 4148 | months[18] = "Jul"; |
| 4149 | months[19] = "Aug"; |
| 4150 | months[20] = "Sep"; |
| 4151 | months[21] = "Oct"; |
| 4152 | months[22] = "Nov"; |
| 4153 | months[23] = "Dec"; |
| 4154 | return months; |
| 4155 | } |
| 4156 | |
| 4157 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4158 | static wstring* init_wmonths() { |
| 4159 | static wstring months[24]; |
| 4160 | months[0] = L"January"; |
| 4161 | months[1] = L"February"; |
| 4162 | months[2] = L"March"; |
| 4163 | months[3] = L"April"; |
| 4164 | months[4] = L"May"; |
| 4165 | months[5] = L"June"; |
| 4166 | months[6] = L"July"; |
| 4167 | months[7] = L"August"; |
| 4168 | months[8] = L"September"; |
| 4169 | months[9] = L"October"; |
| 4170 | months[10] = L"November"; |
| 4171 | months[11] = L"December"; |
| 4172 | months[12] = L"Jan"; |
| 4173 | months[13] = L"Feb"; |
| 4174 | months[14] = L"Mar"; |
| 4175 | months[15] = L"Apr"; |
| 4176 | months[16] = L"May"; |
| 4177 | months[17] = L"Jun"; |
| 4178 | months[18] = L"Jul"; |
| 4179 | months[19] = L"Aug"; |
| 4180 | months[20] = L"Sep"; |
| 4181 | months[21] = L"Oct"; |
| 4182 | months[22] = L"Nov"; |
| 4183 | months[23] = L"Dec"; |
| 4184 | return months; |
| 4185 | } |
| 4186 | #endif |
| 4187 | |
| 4188 | template <> |
| 4189 | const string* __time_get_c_storage<char>::__months() const { |
| 4190 | static const string* months = init_months(); |
| 4191 | return months; |
| 4192 | } |
| 4193 | |
| 4194 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4195 | template <> |
| 4196 | const wstring* __time_get_c_storage<wchar_t>::__months() const { |
| 4197 | static const wstring* months = init_wmonths(); |
| 4198 | return months; |
| 4199 | } |
| 4200 | #endif |
| 4201 | |
| 4202 | static string* init_am_pm() { |
| 4203 | static string am_pm[2]; |
| 4204 | am_pm[0] = "AM"; |
| 4205 | am_pm[1] = "PM"; |
| 4206 | return am_pm; |
| 4207 | } |
| 4208 | |
| 4209 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4210 | static wstring* init_wam_pm() { |
| 4211 | static wstring am_pm[2]; |
| 4212 | am_pm[0] = L"AM"; |
| 4213 | am_pm[1] = L"PM"; |
| 4214 | return am_pm; |
| 4215 | } |
| 4216 | #endif |
| 4217 | |
| 4218 | template <> |
| 4219 | const string* __time_get_c_storage<char>::__am_pm() const { |
| 4220 | static const string* am_pm = init_am_pm(); |
| 4221 | return am_pm; |
| 4222 | } |
| 4223 | |
| 4224 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4225 | template <> |
| 4226 | const wstring* __time_get_c_storage<wchar_t>::__am_pm() const { |
| 4227 | static const wstring* am_pm = init_wam_pm(); |
| 4228 | return am_pm; |
| 4229 | } |
| 4230 | #endif |
| 4231 | |
| 4232 | template <> |
| 4233 | const string& __time_get_c_storage<char>::__x() const { |
| 4234 | static string s("%m/%d/%y"); |
| 4235 | return s; |
| 4236 | } |
| 4237 | |
| 4238 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4239 | template <> |
| 4240 | const wstring& __time_get_c_storage<wchar_t>::__x() const { |
| 4241 | static wstring s(L"%m/%d/%y"); |
| 4242 | return s; |
| 4243 | } |
| 4244 | #endif |
| 4245 | |
| 4246 | template <> |
| 4247 | const string& __time_get_c_storage<char>::__X() const { |
| 4248 | static string s("%H:%M:%S"); |
| 4249 | return s; |
| 4250 | } |
| 4251 | |
| 4252 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4253 | template <> |
| 4254 | const wstring& __time_get_c_storage<wchar_t>::__X() const { |
| 4255 | static wstring s(L"%H:%M:%S"); |
| 4256 | return s; |
| 4257 | } |
| 4258 | #endif |
| 4259 | |
| 4260 | template <> |
| 4261 | const string& __time_get_c_storage<char>::__c() const { |
| 4262 | static string s("%a %b %d %H:%M:%S %Y"); |
| 4263 | return s; |
| 4264 | } |
| 4265 | |
| 4266 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4267 | template <> |
| 4268 | const wstring& __time_get_c_storage<wchar_t>::__c() const { |
| 4269 | static wstring s(L"%a %b %d %H:%M:%S %Y"); |
| 4270 | return s; |
| 4271 | } |
| 4272 | #endif |
| 4273 | |
| 4274 | template <> |
| 4275 | const string& __time_get_c_storage<char>::__r() const { |
| 4276 | static string s("%I:%M:%S %p"); |
| 4277 | return s; |
| 4278 | } |
| 4279 | |
| 4280 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4281 | template <> |
| 4282 | const wstring& __time_get_c_storage<wchar_t>::__r() const { |
| 4283 | static wstring s(L"%I:%M:%S %p"); |
| 4284 | return s; |
| 4285 | } |
| 4286 | #endif |
| 4287 | |
| 4288 | // time_get_byname |
| 4289 | |
| 4290 | __time_get::__time_get(const char* nm) : __loc_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: nm, base: 0)) { |
| 4291 | if (__loc_ == 0) |
| 4292 | std::__throw_runtime_error(("time_get_byname failed to construct for "+ string(nm)).c_str()); |
| 4293 | } |
| 4294 | |
| 4295 | __time_get::__time_get(const string& nm) : __loc_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: nm.c_str(), base: 0)) { |
| 4296 | if (__loc_ == 0) |
| 4297 | std::__throw_runtime_error(("time_get_byname failed to construct for "+ nm).c_str()); |
| 4298 | } |
| 4299 | |
| 4300 | __time_get::~__time_get() { __locale::__freelocale(loc: __loc_); } |
| 4301 | |
| 4302 | _LIBCPP_CLANG_DIAGNOSTIC_IGNORED("-Wmissing-field-initializers") |
| 4303 | |
| 4304 | template <> |
| 4305 | string __time_get_storage<char>::__analyze(char fmt, const ctype<char>& ct) { |
| 4306 | tm t = {.tm_sec: 0}; |
| 4307 | t.tm_sec = 59; |
| 4308 | t.tm_min = 55; |
| 4309 | t.tm_hour = 23; |
| 4310 | t.tm_mday = 31; |
| 4311 | t.tm_mon = 11; |
| 4312 | t.tm_year = 161; |
| 4313 | t.tm_wday = 6; |
| 4314 | t.tm_yday = 364; |
| 4315 | t.tm_isdst = -1; |
| 4316 | char buf[100]; |
| 4317 | char f[3] = {0}; |
| 4318 | f[0] = '%'; |
| 4319 | f[1] = fmt; |
| 4320 | size_t n = __locale::__strftime(s: buf, max: std::size(buf), format: f, tm: &t, loc: __loc_); |
| 4321 | char* bb = buf; |
| 4322 | char* be = buf + n; |
| 4323 | string result; |
| 4324 | while (bb != be) { |
| 4325 | if (ct.is(m: ctype_base::space, c: *bb)) { |
| 4326 | result.push_back(c: ' '); |
| 4327 | for (++bb; bb != be && ct.is(m: ctype_base::space, c: *bb); ++bb) |
| 4328 | ; |
| 4329 | continue; |
| 4330 | } |
| 4331 | char* w = bb; |
| 4332 | ios_base::iostate err = ios_base::goodbit; |
| 4333 | ptrdiff_t i = __scan_keyword(b&: w, e: be, kb: this->__weeks_, ke: this->__weeks_ + 14, ct: ct, err&: err, case_sensitive: false) - this->__weeks_; |
| 4334 | if (i < 14) { |
| 4335 | result.push_back(c: '%'); |
| 4336 | if (i < 7) |
| 4337 | result.push_back(c: 'A'); |
| 4338 | else |
| 4339 | result.push_back(c: 'a'); |
| 4340 | bb = w; |
| 4341 | continue; |
| 4342 | } |
| 4343 | w = bb; |
| 4344 | i = __scan_keyword(b&: w, e: be, kb: this->__months_, ke: this->__months_ + 24, ct: ct, err&: err, case_sensitive: false) - this->__months_; |
| 4345 | if (i < 24) { |
| 4346 | result.push_back(c: '%'); |
| 4347 | if (i < 12) |
| 4348 | result.push_back(c: 'B'); |
| 4349 | else |
| 4350 | result.push_back(c: 'b'); |
| 4351 | if (fmt == 'x' && ct.is(m: ctype_base::digit, c: this->__months_[i][0])) |
| 4352 | result.back() = 'm'; |
| 4353 | bb = w; |
| 4354 | continue; |
| 4355 | } |
| 4356 | if (this->__am_pm_[0].size() + this->__am_pm_[1].size() > 0) { |
| 4357 | w = bb; |
| 4358 | i = __scan_keyword(b&: w, e: be, kb: this->__am_pm_, ke: this->__am_pm_ + 2, ct: ct, err&: err, case_sensitive: false) - this->__am_pm_; |
| 4359 | if (i < 2) { |
| 4360 | result.push_back(c: '%'); |
| 4361 | result.push_back(c: 'p'); |
| 4362 | bb = w; |
| 4363 | continue; |
| 4364 | } |
| 4365 | } |
| 4366 | w = bb; |
| 4367 | if (ct.is(m: ctype_base::digit, c: *bb)) { |
| 4368 | switch (__get_up_to_n_digits(b&: bb, e: be, err&: err, ct: ct, n: 4)) { |
| 4369 | case 6: |
| 4370 | result.push_back(c: '%'); |
| 4371 | result.push_back(c: 'w'); |
| 4372 | break; |
| 4373 | case 7: |
| 4374 | result.push_back(c: '%'); |
| 4375 | result.push_back(c: 'u'); |
| 4376 | break; |
| 4377 | case 11: |
| 4378 | result.push_back(c: '%'); |
| 4379 | result.push_back(c: 'I'); |
| 4380 | break; |
| 4381 | case 12: |
| 4382 | result.push_back(c: '%'); |
| 4383 | result.push_back(c: 'm'); |
| 4384 | break; |
| 4385 | case 23: |
| 4386 | result.push_back(c: '%'); |
| 4387 | result.push_back(c: 'H'); |
| 4388 | break; |
| 4389 | case 31: |
| 4390 | result.push_back(c: '%'); |
| 4391 | result.push_back(c: 'd'); |
| 4392 | break; |
| 4393 | case 55: |
| 4394 | result.push_back(c: '%'); |
| 4395 | result.push_back(c: 'M'); |
| 4396 | break; |
| 4397 | case 59: |
| 4398 | result.push_back(c: '%'); |
| 4399 | result.push_back(c: 'S'); |
| 4400 | break; |
| 4401 | case 61: |
| 4402 | result.push_back(c: '%'); |
| 4403 | result.push_back(c: 'y'); |
| 4404 | break; |
| 4405 | case 364: |
| 4406 | result.push_back(c: '%'); |
| 4407 | result.push_back(c: 'j'); |
| 4408 | break; |
| 4409 | case 2061: |
| 4410 | result.push_back(c: '%'); |
| 4411 | result.push_back(c: 'Y'); |
| 4412 | break; |
| 4413 | default: |
| 4414 | for (; w != bb; ++w) |
| 4415 | result.push_back(c: *w); |
| 4416 | break; |
| 4417 | } |
| 4418 | continue; |
| 4419 | } |
| 4420 | if (*bb == '%') { |
| 4421 | result.push_back(c: '%'); |
| 4422 | result.push_back(c: '%'); |
| 4423 | ++bb; |
| 4424 | continue; |
| 4425 | } |
| 4426 | result.push_back(c: *bb); |
| 4427 | ++bb; |
| 4428 | } |
| 4429 | return result; |
| 4430 | } |
| 4431 | |
| 4432 | _LIBCPP_CLANG_DIAGNOSTIC_IGNORED("-Wmissing-braces") |
| 4433 | |
| 4434 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4435 | template <> |
| 4436 | wstring __time_get_storage<wchar_t>::__analyze(char fmt, const ctype<wchar_t>& ct) { |
| 4437 | tm t = {.tm_sec: 0}; |
| 4438 | t.tm_sec = 59; |
| 4439 | t.tm_min = 55; |
| 4440 | t.tm_hour = 23; |
| 4441 | t.tm_mday = 31; |
| 4442 | t.tm_mon = 11; |
| 4443 | t.tm_year = 161; |
| 4444 | t.tm_wday = 6; |
| 4445 | t.tm_yday = 364; |
| 4446 | t.tm_isdst = -1; |
| 4447 | char buf[100]; |
| 4448 | char f[3] = {0}; |
| 4449 | f[0] = '%'; |
| 4450 | f[1] = fmt; |
| 4451 | __locale::__strftime(s: buf, max: std::size(buf), format: f, tm: &t, loc: __loc_); |
| 4452 | wchar_t wbuf[100]; |
| 4453 | wchar_t* wbb = wbuf; |
| 4454 | mbstate_t mb = {0}; |
| 4455 | const char* bb = buf; |
| 4456 | size_t j = __locale::__mbsrtowcs(dest: wbb, src: &bb, len: std::size(wbuf), ps: &mb, loc: __loc_); |
| 4457 | if (j == size_t(-1)) |
| 4458 | std::__throw_runtime_error("locale not supported"); |
| 4459 | wchar_t* wbe = wbb + j; |
| 4460 | wstring result; |
| 4461 | while (wbb != wbe) { |
| 4462 | if (ct.is(m: ctype_base::space, c: *wbb)) { |
| 4463 | result.push_back(c: L' '); |
| 4464 | for (++wbb; wbb != wbe && ct.is(m: ctype_base::space, c: *wbb); ++wbb) |
| 4465 | ; |
| 4466 | continue; |
| 4467 | } |
| 4468 | wchar_t* w = wbb; |
| 4469 | ios_base::iostate err = ios_base::goodbit; |
| 4470 | ptrdiff_t i = __scan_keyword(b&: w, e: wbe, kb: this->__weeks_, ke: this->__weeks_ + 14, ct: ct, err&: err, case_sensitive: false) - this->__weeks_; |
| 4471 | if (i < 14) { |
| 4472 | result.push_back(c: L'%'); |
| 4473 | if (i < 7) |
| 4474 | result.push_back(c: L'A'); |
| 4475 | else |
| 4476 | result.push_back(c: L'a'); |
| 4477 | wbb = w; |
| 4478 | continue; |
| 4479 | } |
| 4480 | w = wbb; |
| 4481 | i = __scan_keyword(b&: w, e: wbe, kb: this->__months_, ke: this->__months_ + 24, ct: ct, err&: err, case_sensitive: false) - this->__months_; |
| 4482 | if (i < 24) { |
| 4483 | result.push_back(c: L'%'); |
| 4484 | if (i < 12) |
| 4485 | result.push_back(c: L'B'); |
| 4486 | else |
| 4487 | result.push_back(c: L'b'); |
| 4488 | if (fmt == 'x' && ct.is(m: ctype_base::digit, c: this->__months_[i][0])) |
| 4489 | result.back() = L'm'; |
| 4490 | wbb = w; |
| 4491 | continue; |
| 4492 | } |
| 4493 | if (this->__am_pm_[0].size() + this->__am_pm_[1].size() > 0) { |
| 4494 | w = wbb; |
| 4495 | i = __scan_keyword(b&: w, e: wbe, kb: this->__am_pm_, ke: this->__am_pm_ + 2, ct: ct, err&: err, case_sensitive: false) - this->__am_pm_; |
| 4496 | if (i < 2) { |
| 4497 | result.push_back(c: L'%'); |
| 4498 | result.push_back(c: L'p'); |
| 4499 | wbb = w; |
| 4500 | continue; |
| 4501 | } |
| 4502 | } |
| 4503 | w = wbb; |
| 4504 | if (ct.is(m: ctype_base::digit, c: *wbb)) { |
| 4505 | switch (__get_up_to_n_digits(b&: wbb, e: wbe, err&: err, ct: ct, n: 4)) { |
| 4506 | case 6: |
| 4507 | result.push_back(c: L'%'); |
| 4508 | result.push_back(c: L'w'); |
| 4509 | break; |
| 4510 | case 7: |
| 4511 | result.push_back(c: L'%'); |
| 4512 | result.push_back(c: L'u'); |
| 4513 | break; |
| 4514 | case 11: |
| 4515 | result.push_back(c: L'%'); |
| 4516 | result.push_back(c: L'I'); |
| 4517 | break; |
| 4518 | case 12: |
| 4519 | result.push_back(c: L'%'); |
| 4520 | result.push_back(c: L'm'); |
| 4521 | break; |
| 4522 | case 23: |
| 4523 | result.push_back(c: L'%'); |
| 4524 | result.push_back(c: L'H'); |
| 4525 | break; |
| 4526 | case 31: |
| 4527 | result.push_back(c: L'%'); |
| 4528 | result.push_back(c: L'd'); |
| 4529 | break; |
| 4530 | case 55: |
| 4531 | result.push_back(c: L'%'); |
| 4532 | result.push_back(c: L'M'); |
| 4533 | break; |
| 4534 | case 59: |
| 4535 | result.push_back(c: L'%'); |
| 4536 | result.push_back(c: L'S'); |
| 4537 | break; |
| 4538 | case 61: |
| 4539 | result.push_back(c: L'%'); |
| 4540 | result.push_back(c: L'y'); |
| 4541 | break; |
| 4542 | case 364: |
| 4543 | result.push_back(c: L'%'); |
| 4544 | result.push_back(c: L'j'); |
| 4545 | break; |
| 4546 | case 2061: |
| 4547 | result.push_back(c: L'%'); |
| 4548 | result.push_back(c: L'Y'); |
| 4549 | break; |
| 4550 | default: |
| 4551 | for (; w != wbb; ++w) |
| 4552 | result.push_back(c: *w); |
| 4553 | break; |
| 4554 | } |
| 4555 | continue; |
| 4556 | } |
| 4557 | if (ct.narrow(c: *wbb, dfault: 0) == '%') { |
| 4558 | result.push_back(c: L'%'); |
| 4559 | result.push_back(c: L'%'); |
| 4560 | ++wbb; |
| 4561 | continue; |
| 4562 | } |
| 4563 | result.push_back(c: *wbb); |
| 4564 | ++wbb; |
| 4565 | } |
| 4566 | return result; |
| 4567 | } |
| 4568 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 4569 | |
| 4570 | template <> |
| 4571 | void __time_get_storage<char>::init(const ctype<char>& ct) { |
| 4572 | tm t = {.tm_sec: 0}; |
| 4573 | char buf[100]; |
| 4574 | // __weeks_ |
| 4575 | for (int i = 0; i < 7; ++i) { |
| 4576 | t.tm_wday = i; |
| 4577 | __locale::__strftime(s: buf, max: std::size(buf), format: "%A", tm: &t, loc: __loc_); |
| 4578 | __weeks_[i] = buf; |
| 4579 | __locale::__strftime(s: buf, max: std::size(buf), format: "%a", tm: &t, loc: __loc_); |
| 4580 | __weeks_[i + 7] = buf; |
| 4581 | } |
| 4582 | // __months_ |
| 4583 | for (int i = 0; i < 12; ++i) { |
| 4584 | t.tm_mon = i; |
| 4585 | __locale::__strftime(s: buf, max: std::size(buf), format: "%B", tm: &t, loc: __loc_); |
| 4586 | __months_[i] = buf; |
| 4587 | __locale::__strftime(s: buf, max: std::size(buf), format: "%b", tm: &t, loc: __loc_); |
| 4588 | __months_[i + 12] = buf; |
| 4589 | } |
| 4590 | // __am_pm_ |
| 4591 | t.tm_hour = 1; |
| 4592 | __locale::__strftime(s: buf, max: std::size(buf), format: "%p", tm: &t, loc: __loc_); |
| 4593 | __am_pm_[0] = buf; |
| 4594 | t.tm_hour = 13; |
| 4595 | __locale::__strftime(s: buf, max: std::size(buf), format: "%p", tm: &t, loc: __loc_); |
| 4596 | __am_pm_[1] = buf; |
| 4597 | __c_ = __analyze(fmt: 'c', ct); |
| 4598 | __r_ = __analyze(fmt: 'r', ct); |
| 4599 | __x_ = __analyze(fmt: 'x', ct); |
| 4600 | __X_ = __analyze(fmt: 'X', ct); |
| 4601 | } |
| 4602 | |
| 4603 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4604 | template <> |
| 4605 | void __time_get_storage<wchar_t>::init(const ctype<wchar_t>& ct) { |
| 4606 | tm t = {.tm_sec: 0}; |
| 4607 | char buf[100]; |
| 4608 | wchar_t wbuf[100]; |
| 4609 | wchar_t* wbe; |
| 4610 | mbstate_t mb = {0}; |
| 4611 | // __weeks_ |
| 4612 | for (int i = 0; i < 7; ++i) { |
| 4613 | t.tm_wday = i; |
| 4614 | __locale::__strftime(s: buf, max: std::size(buf), format: "%A", tm: &t, loc: __loc_); |
| 4615 | mb = mbstate_t(); |
| 4616 | const char* bb = buf; |
| 4617 | size_t j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: __loc_); |
| 4618 | if (j == size_t(-1) || j == 0) |
| 4619 | std::__throw_runtime_error("locale not supported"); |
| 4620 | wbe = wbuf + j; |
| 4621 | __weeks_[i].assign(first: wbuf, last: wbe); |
| 4622 | __locale::__strftime(s: buf, max: std::size(buf), format: "%a", tm: &t, loc: __loc_); |
| 4623 | mb = mbstate_t(); |
| 4624 | bb = buf; |
| 4625 | j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: __loc_); |
| 4626 | if (j == size_t(-1) || j == 0) |
| 4627 | std::__throw_runtime_error("locale not supported"); |
| 4628 | wbe = wbuf + j; |
| 4629 | __weeks_[i + 7].assign(first: wbuf, last: wbe); |
| 4630 | } |
| 4631 | // __months_ |
| 4632 | for (int i = 0; i < 12; ++i) { |
| 4633 | t.tm_mon = i; |
| 4634 | __locale::__strftime(s: buf, max: std::size(buf), format: "%B", tm: &t, loc: __loc_); |
| 4635 | mb = mbstate_t(); |
| 4636 | const char* bb = buf; |
| 4637 | size_t j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: __loc_); |
| 4638 | if (j == size_t(-1) || j == 0) |
| 4639 | std::__throw_runtime_error("locale not supported"); |
| 4640 | wbe = wbuf + j; |
| 4641 | __months_[i].assign(first: wbuf, last: wbe); |
| 4642 | __locale::__strftime(s: buf, max: std::size(buf), format: "%b", tm: &t, loc: __loc_); |
| 4643 | mb = mbstate_t(); |
| 4644 | bb = buf; |
| 4645 | j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: __loc_); |
| 4646 | if (j == size_t(-1) || j == 0) |
| 4647 | std::__throw_runtime_error("locale not supported"); |
| 4648 | wbe = wbuf + j; |
| 4649 | __months_[i + 12].assign(first: wbuf, last: wbe); |
| 4650 | } |
| 4651 | // __am_pm_ |
| 4652 | t.tm_hour = 1; |
| 4653 | __locale::__strftime(s: buf, max: std::size(buf), format: "%p", tm: &t, loc: __loc_); |
| 4654 | mb = mbstate_t(); |
| 4655 | const char* bb = buf; |
| 4656 | size_t j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: __loc_); |
| 4657 | if (j == size_t(-1)) |
| 4658 | std::__throw_runtime_error("locale not supported"); |
| 4659 | wbe = wbuf + j; |
| 4660 | __am_pm_[0].assign(first: wbuf, last: wbe); |
| 4661 | t.tm_hour = 13; |
| 4662 | __locale::__strftime(s: buf, max: std::size(buf), format: "%p", tm: &t, loc: __loc_); |
| 4663 | mb = mbstate_t(); |
| 4664 | bb = buf; |
| 4665 | j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: __loc_); |
| 4666 | if (j == size_t(-1)) |
| 4667 | std::__throw_runtime_error("locale not supported"); |
| 4668 | wbe = wbuf + j; |
| 4669 | __am_pm_[1].assign(first: wbuf, last: wbe); |
| 4670 | __c_ = __analyze(fmt: 'c', ct); |
| 4671 | __r_ = __analyze(fmt: 'r', ct); |
| 4672 | __x_ = __analyze(fmt: 'x', ct); |
| 4673 | __X_ = __analyze(fmt: 'X', ct); |
| 4674 | } |
| 4675 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 4676 | |
| 4677 | template <class CharT> |
| 4678 | struct _LIBCPP_HIDDEN __time_get_temp : public ctype_byname<CharT> { |
| 4679 | explicit __time_get_temp(const char* nm) : ctype_byname<CharT>(nm, 1) {} |
| 4680 | explicit __time_get_temp(const string& nm) : ctype_byname<CharT>(nm, 1) {} |
| 4681 | }; |
| 4682 | |
| 4683 | template <> |
| 4684 | __time_get_storage<char>::__time_get_storage(const char* __nm) : __time_get(__nm) { |
| 4685 | const __time_get_temp<char> ct(__nm); |
| 4686 | init(ct); |
| 4687 | } |
| 4688 | |
| 4689 | template <> |
| 4690 | __time_get_storage<char>::__time_get_storage(const string& __nm) : __time_get(__nm) { |
| 4691 | const __time_get_temp<char> ct(__nm); |
| 4692 | init(ct); |
| 4693 | } |
| 4694 | |
| 4695 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4696 | template <> |
| 4697 | __time_get_storage<wchar_t>::__time_get_storage(const char* __nm) : __time_get(__nm) { |
| 4698 | const __time_get_temp<wchar_t> ct(__nm); |
| 4699 | init(ct); |
| 4700 | } |
| 4701 | |
| 4702 | template <> |
| 4703 | __time_get_storage<wchar_t>::__time_get_storage(const string& __nm) : __time_get(__nm) { |
| 4704 | const __time_get_temp<wchar_t> ct(__nm); |
| 4705 | init(ct); |
| 4706 | } |
| 4707 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 4708 | |
| 4709 | template <> |
| 4710 | time_base::dateorder __time_get_storage<char>::__do_date_order() const { |
| 4711 | unsigned i; |
| 4712 | for (i = 0; i < __x_.size(); ++i) |
| 4713 | if (__x_[i] == '%') |
| 4714 | break; |
| 4715 | ++i; |
| 4716 | switch (__x_[i]) { |
| 4717 | case 'y': |
| 4718 | case 'Y': |
| 4719 | for (++i; i < __x_.size(); ++i) |
| 4720 | if (__x_[i] == '%') |
| 4721 | break; |
| 4722 | if (i == __x_.size()) |
| 4723 | break; |
| 4724 | ++i; |
| 4725 | switch (__x_[i]) { |
| 4726 | case 'm': |
| 4727 | for (++i; i < __x_.size(); ++i) |
| 4728 | if (__x_[i] == '%') |
| 4729 | break; |
| 4730 | if (i == __x_.size()) |
| 4731 | break; |
| 4732 | ++i; |
| 4733 | if (__x_[i] == 'd') |
| 4734 | return time_base::ymd; |
| 4735 | break; |
| 4736 | case 'd': |
| 4737 | for (++i; i < __x_.size(); ++i) |
| 4738 | if (__x_[i] == '%') |
| 4739 | break; |
| 4740 | if (i == __x_.size()) |
| 4741 | break; |
| 4742 | ++i; |
| 4743 | if (__x_[i] == 'm') |
| 4744 | return time_base::ydm; |
| 4745 | break; |
| 4746 | } |
| 4747 | break; |
| 4748 | case 'm': |
| 4749 | for (++i; i < __x_.size(); ++i) |
| 4750 | if (__x_[i] == '%') |
| 4751 | break; |
| 4752 | if (i == __x_.size()) |
| 4753 | break; |
| 4754 | ++i; |
| 4755 | if (__x_[i] == 'd') { |
| 4756 | for (++i; i < __x_.size(); ++i) |
| 4757 | if (__x_[i] == '%') |
| 4758 | break; |
| 4759 | if (i == __x_.size()) |
| 4760 | break; |
| 4761 | ++i; |
| 4762 | if (__x_[i] == 'y' || __x_[i] == 'Y') |
| 4763 | return time_base::mdy; |
| 4764 | break; |
| 4765 | } |
| 4766 | break; |
| 4767 | case 'd': |
| 4768 | for (++i; i < __x_.size(); ++i) |
| 4769 | if (__x_[i] == '%') |
| 4770 | break; |
| 4771 | if (i == __x_.size()) |
| 4772 | break; |
| 4773 | ++i; |
| 4774 | if (__x_[i] == 'm') { |
| 4775 | for (++i; i < __x_.size(); ++i) |
| 4776 | if (__x_[i] == '%') |
| 4777 | break; |
| 4778 | if (i == __x_.size()) |
| 4779 | break; |
| 4780 | ++i; |
| 4781 | if (__x_[i] == 'y' || __x_[i] == 'Y') |
| 4782 | return time_base::dmy; |
| 4783 | break; |
| 4784 | } |
| 4785 | break; |
| 4786 | } |
| 4787 | return time_base::no_order; |
| 4788 | } |
| 4789 | |
| 4790 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4791 | template <> |
| 4792 | time_base::dateorder __time_get_storage<wchar_t>::__do_date_order() const { |
| 4793 | unsigned i; |
| 4794 | for (i = 0; i < __x_.size(); ++i) |
| 4795 | if (__x_[i] == L'%') |
| 4796 | break; |
| 4797 | ++i; |
| 4798 | switch (__x_[i]) { |
| 4799 | case L'y': |
| 4800 | case L'Y': |
| 4801 | for (++i; i < __x_.size(); ++i) |
| 4802 | if (__x_[i] == L'%') |
| 4803 | break; |
| 4804 | if (i == __x_.size()) |
| 4805 | break; |
| 4806 | ++i; |
| 4807 | switch (__x_[i]) { |
| 4808 | case L'm': |
| 4809 | for (++i; i < __x_.size(); ++i) |
| 4810 | if (__x_[i] == L'%') |
| 4811 | break; |
| 4812 | if (i == __x_.size()) |
| 4813 | break; |
| 4814 | ++i; |
| 4815 | if (__x_[i] == L'd') |
| 4816 | return time_base::ymd; |
| 4817 | break; |
| 4818 | case L'd': |
| 4819 | for (++i; i < __x_.size(); ++i) |
| 4820 | if (__x_[i] == L'%') |
| 4821 | break; |
| 4822 | if (i == __x_.size()) |
| 4823 | break; |
| 4824 | ++i; |
| 4825 | if (__x_[i] == L'm') |
| 4826 | return time_base::ydm; |
| 4827 | break; |
| 4828 | } |
| 4829 | break; |
| 4830 | case L'm': |
| 4831 | for (++i; i < __x_.size(); ++i) |
| 4832 | if (__x_[i] == L'%') |
| 4833 | break; |
| 4834 | if (i == __x_.size()) |
| 4835 | break; |
| 4836 | ++i; |
| 4837 | if (__x_[i] == L'd') { |
| 4838 | for (++i; i < __x_.size(); ++i) |
| 4839 | if (__x_[i] == L'%') |
| 4840 | break; |
| 4841 | if (i == __x_.size()) |
| 4842 | break; |
| 4843 | ++i; |
| 4844 | if (__x_[i] == L'y' || __x_[i] == L'Y') |
| 4845 | return time_base::mdy; |
| 4846 | break; |
| 4847 | } |
| 4848 | break; |
| 4849 | case L'd': |
| 4850 | for (++i; i < __x_.size(); ++i) |
| 4851 | if (__x_[i] == L'%') |
| 4852 | break; |
| 4853 | if (i == __x_.size()) |
| 4854 | break; |
| 4855 | ++i; |
| 4856 | if (__x_[i] == L'm') { |
| 4857 | for (++i; i < __x_.size(); ++i) |
| 4858 | if (__x_[i] == L'%') |
| 4859 | break; |
| 4860 | if (i == __x_.size()) |
| 4861 | break; |
| 4862 | ++i; |
| 4863 | if (__x_[i] == L'y' || __x_[i] == L'Y') |
| 4864 | return time_base::dmy; |
| 4865 | break; |
| 4866 | } |
| 4867 | break; |
| 4868 | } |
| 4869 | return time_base::no_order; |
| 4870 | } |
| 4871 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 4872 | |
| 4873 | // time_put |
| 4874 | |
| 4875 | __time_put::__time_put(const char* nm) : __loc_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: nm, base: 0)) { |
| 4876 | if (__loc_ == 0) |
| 4877 | std::__throw_runtime_error(("time_put_byname failed to construct for "+ string(nm)).c_str()); |
| 4878 | } |
| 4879 | |
| 4880 | __time_put::__time_put(const string& nm) : __loc_(__locale::__newlocale(_LIBCPP_ALL_MASK, locale: nm.c_str(), base: 0)) { |
| 4881 | if (__loc_ == 0) |
| 4882 | std::__throw_runtime_error(("time_put_byname failed to construct for "+ nm).c_str()); |
| 4883 | } |
| 4884 | |
| 4885 | __time_put::~__time_put() { |
| 4886 | if (__loc_ != __locale::__get_c_locale()) |
| 4887 | __locale::__freelocale(loc: __loc_); |
| 4888 | } |
| 4889 | |
| 4890 | void __time_put::__do_put(char* __nb, char*& __ne, const tm* __tm, char __fmt, char __mod) const { |
| 4891 | char fmt[] = {'%', __fmt, __mod, 0}; |
| 4892 | if (__mod != 0) |
| 4893 | swap(x&: fmt[1], y&: fmt[2]); |
| 4894 | size_t n = __locale::__strftime(s: __nb, max: std::distance(first: __nb, last: __ne), format: fmt, __tm, loc: __loc_); |
| 4895 | __ne = __nb + n; |
| 4896 | } |
| 4897 | |
| 4898 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 4899 | void __time_put::__do_put(wchar_t* __wb, wchar_t*& __we, const tm* __tm, char __fmt, char __mod) const { |
| 4900 | char __nar[100]; |
| 4901 | char* __ne = __nar + 100; |
| 4902 | __do_put(nb: __nar, __ne, __tm, __fmt, __mod); |
| 4903 | mbstate_t mb = {0}; |
| 4904 | const char* __nb = __nar; |
| 4905 | size_t j = __locale::__mbsrtowcs(dest: __wb, src: &__nb, len: std::distance(first: __wb, last: __we), ps: &mb, loc: __loc_); |
| 4906 | if (j == size_t(-1)) |
| 4907 | std::__throw_runtime_error("locale not supported"); |
| 4908 | __we = __wb + j; |
| 4909 | } |
| 4910 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 4911 | |
| 4912 | // moneypunct_byname |
| 4913 | |
| 4914 | template <class charT> |
| 4915 | static void __init_pat( |
| 4916 | money_base::pattern& pat, |
| 4917 | basic_string<charT>& __curr_symbol_, |
| 4918 | bool intl, |
| 4919 | char cs_precedes, |
| 4920 | char sep_by_space, |
| 4921 | char sign_posn, |
| 4922 | charT space_char) { |
| 4923 | const char sign = static_cast<char>(money_base::sign); |
| 4924 | const char space = static_cast<char>(money_base::space); |
| 4925 | const char none = static_cast<char>(money_base::none); |
| 4926 | const char symbol = static_cast<char>(money_base::symbol); |
| 4927 | const char value = static_cast<char>(money_base::value); |
| 4928 | const bool symbol_contains_sep = intl && __curr_symbol_.size() == 4; |
| 4929 | |
| 4930 | // Comments on case branches reflect 'C11 7.11.2.1 The localeconv |
| 4931 | // function'. "Space between sign and symbol or value" means that |
| 4932 | // if the sign is adjacent to the symbol, there's a space between |
| 4933 | // them, and otherwise there's a space between the sign and value. |
| 4934 | // |
| 4935 | // C11's localeconv specifies that the fourth character of an |
| 4936 | // international curr_symbol is used to separate the sign and |
| 4937 | // value when sep_by_space says to do so. C++ can't represent |
| 4938 | // that, so we just use a space. When sep_by_space says to |
| 4939 | // separate the symbol and value-or-sign with a space, we rearrange the |
| 4940 | // curr_symbol to put its spacing character on the correct side of |
| 4941 | // the symbol. |
| 4942 | // |
| 4943 | // We also need to avoid adding an extra space between the sign |
| 4944 | // and value when the currency symbol is suppressed (by not |
| 4945 | // setting showbase). We match glibc's strfmon by interpreting |
| 4946 | // sep_by_space==1 as "omit the space when the currency symbol is |
| 4947 | // absent". |
| 4948 | // |
| 4949 | // Users who want to get this right should use ICU instead. |
| 4950 | |
| 4951 | switch (cs_precedes) { |
| 4952 | case 0: // value before curr_symbol |
| 4953 | if (symbol_contains_sep) { |
| 4954 | // Move the separator to before the symbol, to place it |
| 4955 | // between the value and symbol. |
| 4956 | rotate(__curr_symbol_.begin(), __curr_symbol_.begin() + 3, __curr_symbol_.end()); |
| 4957 | } |
| 4958 | switch (sign_posn) { |
| 4959 | case 0: // Parentheses surround the quantity and currency symbol. |
| 4960 | pat.field[0] = sign; |
| 4961 | pat.field[1] = value; |
| 4962 | pat.field[2] = none; // Any space appears in the symbol. |
| 4963 | pat.field[3] = symbol; |
| 4964 | switch (sep_by_space) { |
| 4965 | case 0: // No space separates the currency symbol and value. |
| 4966 | // This case may have changed between C99 and C11; |
| 4967 | // assume the currency symbol matches the intention. |
| 4968 | case 2: // Space between sign and currency or value. |
| 4969 | // The "sign" is two parentheses, so no space here either. |
| 4970 | return; |
| 4971 | case 1: // Space between currency-and-sign or currency and value. |
| 4972 | if (!symbol_contains_sep) { |
| 4973 | // We insert the space into the symbol instead of |
| 4974 | // setting pat.field[2]=space so that when |
| 4975 | // showbase is not set, the space goes away too. |
| 4976 | __curr_symbol_.insert(0, 1, space_char); |
| 4977 | } |
| 4978 | return; |
| 4979 | default: |
| 4980 | break; |
| 4981 | } |
| 4982 | break; |
| 4983 | case 1: // The sign string precedes the quantity and currency symbol. |
| 4984 | pat.field[0] = sign; |
| 4985 | pat.field[3] = symbol; |
| 4986 | switch (sep_by_space) { |
| 4987 | case 0: // No space separates the currency symbol and value. |
| 4988 | pat.field[1] = value; |
| 4989 | pat.field[2] = none; |
| 4990 | return; |
| 4991 | case 1: // Space between currency-and-sign or currency and value. |
| 4992 | pat.field[1] = value; |
| 4993 | pat.field[2] = none; |
| 4994 | if (!symbol_contains_sep) { |
| 4995 | // We insert the space into the symbol instead of |
| 4996 | // setting pat.field[2]=space so that when |
| 4997 | // showbase is not set, the space goes away too. |
| 4998 | __curr_symbol_.insert(0, 1, space_char); |
| 4999 | } |
| 5000 | return; |
| 5001 | case 2: // Space between sign and currency or value. |
| 5002 | pat.field[1] = space; |
| 5003 | pat.field[2] = value; |
| 5004 | if (symbol_contains_sep) { |
| 5005 | // Remove the separator from the symbol, since it |
| 5006 | // has already appeared after the sign. |
| 5007 | __curr_symbol_.erase(__curr_symbol_.begin()); |
| 5008 | } |
| 5009 | return; |
| 5010 | default: |
| 5011 | break; |
| 5012 | } |
| 5013 | break; |
| 5014 | case 2: // The sign string succeeds the quantity and currency symbol. |
| 5015 | pat.field[0] = value; |
| 5016 | pat.field[3] = sign; |
| 5017 | switch (sep_by_space) { |
| 5018 | case 0: // No space separates the currency symbol and value. |
| 5019 | pat.field[1] = none; |
| 5020 | pat.field[2] = symbol; |
| 5021 | return; |
| 5022 | case 1: // Space between currency-and-sign or currency and value. |
| 5023 | if (!symbol_contains_sep) { |
| 5024 | // We insert the space into the symbol instead of |
| 5025 | // setting pat.field[1]=space so that when |
| 5026 | // showbase is not set, the space goes away too. |
| 5027 | __curr_symbol_.insert(0, 1, space_char); |
| 5028 | } |
| 5029 | pat.field[1] = none; |
| 5030 | pat.field[2] = symbol; |
| 5031 | return; |
| 5032 | case 2: // Space between sign and currency or value. |
| 5033 | pat.field[1] = symbol; |
| 5034 | pat.field[2] = space; |
| 5035 | if (symbol_contains_sep) { |
| 5036 | // Remove the separator from the symbol, since it |
| 5037 | // should not be removed if showbase is absent. |
| 5038 | __curr_symbol_.erase(__curr_symbol_.begin()); |
| 5039 | } |
| 5040 | return; |
| 5041 | default: |
| 5042 | break; |
| 5043 | } |
| 5044 | break; |
| 5045 | case 3: // The sign string immediately precedes the currency symbol. |
| 5046 | pat.field[0] = value; |
| 5047 | pat.field[3] = symbol; |
| 5048 | switch (sep_by_space) { |
| 5049 | case 0: // No space separates the currency symbol and value. |
| 5050 | pat.field[1] = none; |
| 5051 | pat.field[2] = sign; |
| 5052 | return; |
| 5053 | case 1: // Space between currency-and-sign or currency and value. |
| 5054 | pat.field[1] = space; |
| 5055 | pat.field[2] = sign; |
| 5056 | if (symbol_contains_sep) { |
| 5057 | // Remove the separator from the symbol, since it |
| 5058 | // has already appeared before the sign. |
| 5059 | __curr_symbol_.erase(__curr_symbol_.begin()); |
| 5060 | } |
| 5061 | return; |
| 5062 | case 2: // Space between sign and currency or value. |
| 5063 | pat.field[1] = sign; |
| 5064 | pat.field[2] = none; |
| 5065 | if (!symbol_contains_sep) { |
| 5066 | // We insert the space into the symbol instead of |
| 5067 | // setting pat.field[2]=space so that when |
| 5068 | // showbase is not set, the space goes away too. |
| 5069 | __curr_symbol_.insert(0, 1, space_char); |
| 5070 | } |
| 5071 | return; |
| 5072 | default: |
| 5073 | break; |
| 5074 | } |
| 5075 | break; |
| 5076 | case 4: // The sign string immediately succeeds the currency symbol. |
| 5077 | pat.field[0] = value; |
| 5078 | pat.field[3] = sign; |
| 5079 | switch (sep_by_space) { |
| 5080 | case 0: // No space separates the currency symbol and value. |
| 5081 | pat.field[1] = none; |
| 5082 | pat.field[2] = symbol; |
| 5083 | return; |
| 5084 | case 1: // Space between currency-and-sign or currency and value. |
| 5085 | pat.field[1] = none; |
| 5086 | pat.field[2] = symbol; |
| 5087 | if (!symbol_contains_sep) { |
| 5088 | // We insert the space into the symbol instead of |
| 5089 | // setting pat.field[1]=space so that when |
| 5090 | // showbase is not set, the space goes away too. |
| 5091 | __curr_symbol_.insert(0, 1, space_char); |
| 5092 | } |
| 5093 | return; |
| 5094 | case 2: // Space between sign and currency or value. |
| 5095 | pat.field[1] = symbol; |
| 5096 | pat.field[2] = space; |
| 5097 | if (symbol_contains_sep) { |
| 5098 | // Remove the separator from the symbol, since it |
| 5099 | // should not disappear when showbase is absent. |
| 5100 | __curr_symbol_.erase(__curr_symbol_.begin()); |
| 5101 | } |
| 5102 | return; |
| 5103 | default: |
| 5104 | break; |
| 5105 | } |
| 5106 | break; |
| 5107 | default: |
| 5108 | break; |
| 5109 | } |
| 5110 | break; |
| 5111 | case 1: // curr_symbol before value |
| 5112 | switch (sign_posn) { |
| 5113 | case 0: // Parentheses surround the quantity and currency symbol. |
| 5114 | pat.field[0] = sign; |
| 5115 | pat.field[1] = symbol; |
| 5116 | pat.field[2] = none; // Any space appears in the symbol. |
| 5117 | pat.field[3] = value; |
| 5118 | switch (sep_by_space) { |
| 5119 | case 0: // No space separates the currency symbol and value. |
| 5120 | // This case may have changed between C99 and C11; |
| 5121 | // assume the currency symbol matches the intention. |
| 5122 | case 2: // Space between sign and currency or value. |
| 5123 | // The "sign" is two parentheses, so no space here either. |
| 5124 | return; |
| 5125 | case 1: // Space between currency-and-sign or currency and value. |
| 5126 | if (!symbol_contains_sep) { |
| 5127 | // We insert the space into the symbol instead of |
| 5128 | // setting pat.field[2]=space so that when |
| 5129 | // showbase is not set, the space goes away too. |
| 5130 | __curr_symbol_.insert(0, 1, space_char); |
| 5131 | } |
| 5132 | return; |
| 5133 | default: |
| 5134 | break; |
| 5135 | } |
| 5136 | break; |
| 5137 | case 1: // The sign string precedes the quantity and currency symbol. |
| 5138 | pat.field[0] = sign; |
| 5139 | pat.field[3] = value; |
| 5140 | switch (sep_by_space) { |
| 5141 | case 0: // No space separates the currency symbol and value. |
| 5142 | pat.field[1] = symbol; |
| 5143 | pat.field[2] = none; |
| 5144 | return; |
| 5145 | case 1: // Space between currency-and-sign or currency and value. |
| 5146 | pat.field[1] = symbol; |
| 5147 | pat.field[2] = none; |
| 5148 | if (!symbol_contains_sep) { |
| 5149 | // We insert the space into the symbol instead of |
| 5150 | // setting pat.field[2]=space so that when |
| 5151 | // showbase is not set, the space goes away too. |
| 5152 | __curr_symbol_.push_back(space_char); |
| 5153 | } |
| 5154 | return; |
| 5155 | case 2: // Space between sign and currency or value. |
| 5156 | pat.field[1] = space; |
| 5157 | pat.field[2] = symbol; |
| 5158 | if (symbol_contains_sep) { |
| 5159 | // Remove the separator from the symbol, since it |
| 5160 | // has already appeared after the sign. |
| 5161 | __curr_symbol_.pop_back(); |
| 5162 | } |
| 5163 | return; |
| 5164 | default: |
| 5165 | break; |
| 5166 | } |
| 5167 | break; |
| 5168 | case 2: // The sign string succeeds the quantity and currency symbol. |
| 5169 | pat.field[0] = symbol; |
| 5170 | pat.field[3] = sign; |
| 5171 | switch (sep_by_space) { |
| 5172 | case 0: // No space separates the currency symbol and value. |
| 5173 | pat.field[1] = none; |
| 5174 | pat.field[2] = value; |
| 5175 | return; |
| 5176 | case 1: // Space between currency-and-sign or currency and value. |
| 5177 | pat.field[1] = none; |
| 5178 | pat.field[2] = value; |
| 5179 | if (!symbol_contains_sep) { |
| 5180 | // We insert the space into the symbol instead of |
| 5181 | // setting pat.field[1]=space so that when |
| 5182 | // showbase is not set, the space goes away too. |
| 5183 | __curr_symbol_.push_back(space_char); |
| 5184 | } |
| 5185 | return; |
| 5186 | case 2: // Space between sign and currency or value. |
| 5187 | pat.field[1] = value; |
| 5188 | pat.field[2] = space; |
| 5189 | if (symbol_contains_sep) { |
| 5190 | // Remove the separator from the symbol, since it |
| 5191 | // will appear before the sign. |
| 5192 | __curr_symbol_.pop_back(); |
| 5193 | } |
| 5194 | return; |
| 5195 | default: |
| 5196 | break; |
| 5197 | } |
| 5198 | break; |
| 5199 | case 3: // The sign string immediately precedes the currency symbol. |
| 5200 | pat.field[0] = sign; |
| 5201 | pat.field[3] = value; |
| 5202 | switch (sep_by_space) { |
| 5203 | case 0: // No space separates the currency symbol and value. |
| 5204 | pat.field[1] = symbol; |
| 5205 | pat.field[2] = none; |
| 5206 | return; |
| 5207 | case 1: // Space between currency-and-sign or currency and value. |
| 5208 | pat.field[1] = symbol; |
| 5209 | pat.field[2] = none; |
| 5210 | if (!symbol_contains_sep) { |
| 5211 | // We insert the space into the symbol instead of |
| 5212 | // setting pat.field[2]=space so that when |
| 5213 | // showbase is not set, the space goes away too. |
| 5214 | __curr_symbol_.push_back(space_char); |
| 5215 | } |
| 5216 | return; |
| 5217 | case 2: // Space between sign and currency or value. |
| 5218 | pat.field[1] = space; |
| 5219 | pat.field[2] = symbol; |
| 5220 | if (symbol_contains_sep) { |
| 5221 | // Remove the separator from the symbol, since it |
| 5222 | // has already appeared after the sign. |
| 5223 | __curr_symbol_.pop_back(); |
| 5224 | } |
| 5225 | return; |
| 5226 | default: |
| 5227 | break; |
| 5228 | } |
| 5229 | break; |
| 5230 | case 4: // The sign string immediately succeeds the currency symbol. |
| 5231 | pat.field[0] = symbol; |
| 5232 | pat.field[3] = value; |
| 5233 | switch (sep_by_space) { |
| 5234 | case 0: // No space separates the currency symbol and value. |
| 5235 | pat.field[1] = sign; |
| 5236 | pat.field[2] = none; |
| 5237 | return; |
| 5238 | case 1: // Space between currency-and-sign or currency and value. |
| 5239 | pat.field[1] = sign; |
| 5240 | pat.field[2] = space; |
| 5241 | if (symbol_contains_sep) { |
| 5242 | // Remove the separator from the symbol, since it |
| 5243 | // should not disappear when showbase is absent. |
| 5244 | __curr_symbol_.pop_back(); |
| 5245 | } |
| 5246 | return; |
| 5247 | case 2: // Space between sign and currency or value. |
| 5248 | pat.field[1] = none; |
| 5249 | pat.field[2] = sign; |
| 5250 | if (!symbol_contains_sep) { |
| 5251 | // We insert the space into the symbol instead of |
| 5252 | // setting pat.field[1]=space so that when |
| 5253 | // showbase is not set, the space goes away too. |
| 5254 | __curr_symbol_.push_back(space_char); |
| 5255 | } |
| 5256 | return; |
| 5257 | default: |
| 5258 | break; |
| 5259 | } |
| 5260 | break; |
| 5261 | default: |
| 5262 | break; |
| 5263 | } |
| 5264 | break; |
| 5265 | default: |
| 5266 | break; |
| 5267 | } |
| 5268 | pat.field[0] = symbol; |
| 5269 | pat.field[1] = sign; |
| 5270 | pat.field[2] = none; |
| 5271 | pat.field[3] = value; |
| 5272 | } |
| 5273 | |
| 5274 | template <> |
| 5275 | void moneypunct_byname<char, false>::init(const char* nm) { |
| 5276 | typedef moneypunct<char, false> base; |
| 5277 | __libcpp_unique_locale loc(nm); |
| 5278 | if (!loc) |
| 5279 | std::__throw_runtime_error(("moneypunct_byname failed to construct for "+ string(nm)).c_str()); |
| 5280 | |
| 5281 | __locale::__lconv_t* lc = __locale::__localeconv(loc&: loc.get()); |
| 5282 | if (!checked_string_to_char_convert(dest&: __decimal_point_, ptr: lc->mon_decimal_point, loc: loc.get())) |
| 5283 | __decimal_point_ = base::do_decimal_point(); |
| 5284 | if (!checked_string_to_char_convert(dest&: __thousands_sep_, ptr: lc->mon_thousands_sep, loc: loc.get())) |
| 5285 | __thousands_sep_ = base::do_thousands_sep(); |
| 5286 | |
| 5287 | __grouping_ = lc->mon_grouping; |
| 5288 | __curr_symbol_ = lc->currency_symbol; |
| 5289 | if (lc->frac_digits != CHAR_MAX) |
| 5290 | __frac_digits_ = lc->frac_digits; |
| 5291 | else |
| 5292 | __frac_digits_ = base::do_frac_digits(); |
| 5293 | if (lc->p_sign_posn == 0) |
| 5294 | __positive_sign_ = "()"; |
| 5295 | else |
| 5296 | __positive_sign_ = lc->positive_sign; |
| 5297 | if (lc->n_sign_posn == 0) |
| 5298 | __negative_sign_ = "()"; |
| 5299 | else |
| 5300 | __negative_sign_ = lc->negative_sign; |
| 5301 | // Assume the positive and negative formats will want spaces in |
| 5302 | // the same places in curr_symbol since there's no way to |
| 5303 | // represent anything else. |
| 5304 | string_type __dummy_curr_symbol = __curr_symbol_; |
| 5305 | __init_pat(pat&: __pos_format_, curr_symbol_&: __dummy_curr_symbol, intl: false, cs_precedes: lc->p_cs_precedes, sep_by_space: lc->p_sep_by_space, sign_posn: lc->p_sign_posn, space_char: ' '); |
| 5306 | __init_pat(pat&: __neg_format_, __curr_symbol_, intl: false, cs_precedes: lc->n_cs_precedes, sep_by_space: lc->n_sep_by_space, sign_posn: lc->n_sign_posn, space_char: ' '); |
| 5307 | } |
| 5308 | |
| 5309 | template <> |
| 5310 | void moneypunct_byname<char, true>::init(const char* nm) { |
| 5311 | typedef moneypunct<char, true> base; |
| 5312 | __libcpp_unique_locale loc(nm); |
| 5313 | if (!loc) |
| 5314 | std::__throw_runtime_error(("moneypunct_byname failed to construct for "+ string(nm)).c_str()); |
| 5315 | |
| 5316 | __locale::__lconv_t* lc = __locale::__localeconv(loc&: loc.get()); |
| 5317 | if (!checked_string_to_char_convert(dest&: __decimal_point_, ptr: lc->mon_decimal_point, loc: loc.get())) |
| 5318 | __decimal_point_ = base::do_decimal_point(); |
| 5319 | if (!checked_string_to_char_convert(dest&: __thousands_sep_, ptr: lc->mon_thousands_sep, loc: loc.get())) |
| 5320 | __thousands_sep_ = base::do_thousands_sep(); |
| 5321 | __grouping_ = lc->mon_grouping; |
| 5322 | __curr_symbol_ = lc->int_curr_symbol; |
| 5323 | if (lc->int_frac_digits != CHAR_MAX) |
| 5324 | __frac_digits_ = lc->int_frac_digits; |
| 5325 | else |
| 5326 | __frac_digits_ = base::do_frac_digits(); |
| 5327 | #ifdef _WIN32 |
| 5328 | if (lc->p_sign_posn == 0) |
| 5329 | #else |
| 5330 | if (lc->int_p_sign_posn == 0) |
| 5331 | #endif |
| 5332 | __positive_sign_ = "()"; |
| 5333 | else |
| 5334 | __positive_sign_ = lc->positive_sign; |
| 5335 | #ifdef _WIN32 |
| 5336 | if (lc->n_sign_posn == 0) |
| 5337 | #else |
| 5338 | if (lc->int_n_sign_posn == 0) |
| 5339 | #endif |
| 5340 | __negative_sign_ = "()"; |
| 5341 | else |
| 5342 | __negative_sign_ = lc->negative_sign; |
| 5343 | // Assume the positive and negative formats will want spaces in |
| 5344 | // the same places in curr_symbol since there's no way to |
| 5345 | // represent anything else. |
| 5346 | string_type __dummy_curr_symbol = __curr_symbol_; |
| 5347 | #ifdef _WIN32 |
| 5348 | __init_pat(__pos_format_, __dummy_curr_symbol, true, lc->p_cs_precedes, lc->p_sep_by_space, lc->p_sign_posn, ' '); |
| 5349 | __init_pat(__neg_format_, __curr_symbol_, true, lc->n_cs_precedes, lc->n_sep_by_space, lc->n_sign_posn, ' '); |
| 5350 | #else // _WIN32 |
| 5351 | __init_pat( |
| 5352 | pat&: __pos_format_, |
| 5353 | curr_symbol_&: __dummy_curr_symbol, |
| 5354 | intl: true, |
| 5355 | cs_precedes: lc->int_p_cs_precedes, |
| 5356 | sep_by_space: lc->int_p_sep_by_space, |
| 5357 | sign_posn: lc->int_p_sign_posn, |
| 5358 | space_char: ' '); |
| 5359 | __init_pat( |
| 5360 | pat&: __neg_format_, __curr_symbol_, intl: true, cs_precedes: lc->int_n_cs_precedes, sep_by_space: lc->int_n_sep_by_space, sign_posn: lc->int_n_sign_posn, space_char: ' '); |
| 5361 | #endif // _WIN32 |
| 5362 | } |
| 5363 | |
| 5364 | #if _LIBCPP_HAS_WIDE_CHARACTERS |
| 5365 | template <> |
| 5366 | void moneypunct_byname<wchar_t, false>::init(const char* nm) { |
| 5367 | typedef moneypunct<wchar_t, false> base; |
| 5368 | __libcpp_unique_locale loc(nm); |
| 5369 | if (!loc) |
| 5370 | std::__throw_runtime_error(("moneypunct_byname failed to construct for "+ string(nm)).c_str()); |
| 5371 | __locale::__lconv_t* lc = __locale::__localeconv(loc&: loc.get()); |
| 5372 | if (!checked_string_to_wchar_convert(dest&: __decimal_point_, ptr: lc->mon_decimal_point, loc: loc.get())) |
| 5373 | __decimal_point_ = base::do_decimal_point(); |
| 5374 | if (!checked_string_to_wchar_convert(dest&: __thousands_sep_, ptr: lc->mon_thousands_sep, loc: loc.get())) |
| 5375 | __thousands_sep_ = base::do_thousands_sep(); |
| 5376 | __grouping_ = lc->mon_grouping; |
| 5377 | wchar_t wbuf[100]; |
| 5378 | mbstate_t mb = {0}; |
| 5379 | const char* bb = lc->currency_symbol; |
| 5380 | size_t j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: loc.get()); |
| 5381 | if (j == size_t(-1)) |
| 5382 | std::__throw_runtime_error("locale not supported"); |
| 5383 | wchar_t* wbe = wbuf + j; |
| 5384 | __curr_symbol_.assign(first: wbuf, last: wbe); |
| 5385 | if (lc->frac_digits != CHAR_MAX) |
| 5386 | __frac_digits_ = lc->frac_digits; |
| 5387 | else |
| 5388 | __frac_digits_ = base::do_frac_digits(); |
| 5389 | if (lc->p_sign_posn == 0) |
| 5390 | __positive_sign_ = L"()"; |
| 5391 | else { |
| 5392 | mb = mbstate_t(); |
| 5393 | bb = lc->positive_sign; |
| 5394 | j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: loc.get()); |
| 5395 | if (j == size_t(-1)) |
| 5396 | std::__throw_runtime_error("locale not supported"); |
| 5397 | wbe = wbuf + j; |
| 5398 | __positive_sign_.assign(first: wbuf, last: wbe); |
| 5399 | } |
| 5400 | if (lc->n_sign_posn == 0) |
| 5401 | __negative_sign_ = L"()"; |
| 5402 | else { |
| 5403 | mb = mbstate_t(); |
| 5404 | bb = lc->negative_sign; |
| 5405 | j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: loc.get()); |
| 5406 | if (j == size_t(-1)) |
| 5407 | std::__throw_runtime_error("locale not supported"); |
| 5408 | wbe = wbuf + j; |
| 5409 | __negative_sign_.assign(first: wbuf, last: wbe); |
| 5410 | } |
| 5411 | // Assume the positive and negative formats will want spaces in |
| 5412 | // the same places in curr_symbol since there's no way to |
| 5413 | // represent anything else. |
| 5414 | string_type __dummy_curr_symbol = __curr_symbol_; |
| 5415 | __init_pat(pat&: __pos_format_, curr_symbol_&: __dummy_curr_symbol, intl: false, cs_precedes: lc->p_cs_precedes, sep_by_space: lc->p_sep_by_space, sign_posn: lc->p_sign_posn, space_char: L' '); |
| 5416 | __init_pat(pat&: __neg_format_, __curr_symbol_, intl: false, cs_precedes: lc->n_cs_precedes, sep_by_space: lc->n_sep_by_space, sign_posn: lc->n_sign_posn, space_char: L' '); |
| 5417 | } |
| 5418 | |
| 5419 | template <> |
| 5420 | void moneypunct_byname<wchar_t, true>::init(const char* nm) { |
| 5421 | typedef moneypunct<wchar_t, true> base; |
| 5422 | __libcpp_unique_locale loc(nm); |
| 5423 | if (!loc) |
| 5424 | std::__throw_runtime_error(("moneypunct_byname failed to construct for "+ string(nm)).c_str()); |
| 5425 | |
| 5426 | __locale::__lconv_t* lc = __locale::__localeconv(loc&: loc.get()); |
| 5427 | if (!checked_string_to_wchar_convert(dest&: __decimal_point_, ptr: lc->mon_decimal_point, loc: loc.get())) |
| 5428 | __decimal_point_ = base::do_decimal_point(); |
| 5429 | if (!checked_string_to_wchar_convert(dest&: __thousands_sep_, ptr: lc->mon_thousands_sep, loc: loc.get())) |
| 5430 | __thousands_sep_ = base::do_thousands_sep(); |
| 5431 | __grouping_ = lc->mon_grouping; |
| 5432 | wchar_t wbuf[100]; |
| 5433 | mbstate_t mb = {0}; |
| 5434 | const char* bb = lc->int_curr_symbol; |
| 5435 | size_t j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: loc.get()); |
| 5436 | if (j == size_t(-1)) |
| 5437 | std::__throw_runtime_error("locale not supported"); |
| 5438 | wchar_t* wbe = wbuf + j; |
| 5439 | __curr_symbol_.assign(first: wbuf, last: wbe); |
| 5440 | if (lc->int_frac_digits != CHAR_MAX) |
| 5441 | __frac_digits_ = lc->int_frac_digits; |
| 5442 | else |
| 5443 | __frac_digits_ = base::do_frac_digits(); |
| 5444 | # ifdef _WIN32 |
| 5445 | if (lc->p_sign_posn == 0) |
| 5446 | # else |
| 5447 | if (lc->int_p_sign_posn == 0) |
| 5448 | # endif |
| 5449 | __positive_sign_ = L"()"; |
| 5450 | else { |
| 5451 | mb = mbstate_t(); |
| 5452 | bb = lc->positive_sign; |
| 5453 | j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: loc.get()); |
| 5454 | if (j == size_t(-1)) |
| 5455 | std::__throw_runtime_error("locale not supported"); |
| 5456 | wbe = wbuf + j; |
| 5457 | __positive_sign_.assign(first: wbuf, last: wbe); |
| 5458 | } |
| 5459 | # ifdef _WIN32 |
| 5460 | if (lc->n_sign_posn == 0) |
| 5461 | # else |
| 5462 | if (lc->int_n_sign_posn == 0) |
| 5463 | # endif |
| 5464 | __negative_sign_ = L"()"; |
| 5465 | else { |
| 5466 | mb = mbstate_t(); |
| 5467 | bb = lc->negative_sign; |
| 5468 | j = __locale::__mbsrtowcs(dest: wbuf, src: &bb, len: std::size(wbuf), ps: &mb, loc: loc.get()); |
| 5469 | if (j == size_t(-1)) |
| 5470 | std::__throw_runtime_error("locale not supported"); |
| 5471 | wbe = wbuf + j; |
| 5472 | __negative_sign_.assign(first: wbuf, last: wbe); |
| 5473 | } |
| 5474 | // Assume the positive and negative formats will want spaces in |
| 5475 | // the same places in curr_symbol since there's no way to |
| 5476 | // represent anything else. |
| 5477 | string_type __dummy_curr_symbol = __curr_symbol_; |
| 5478 | # ifdef _WIN32 |
| 5479 | __init_pat(__pos_format_, __dummy_curr_symbol, true, lc->p_cs_precedes, lc->p_sep_by_space, lc->p_sign_posn, L' '); |
| 5480 | __init_pat(__neg_format_, __curr_symbol_, true, lc->n_cs_precedes, lc->n_sep_by_space, lc->n_sign_posn, L' '); |
| 5481 | # else // _WIN32 |
| 5482 | __init_pat( |
| 5483 | pat&: __pos_format_, |
| 5484 | curr_symbol_&: __dummy_curr_symbol, |
| 5485 | intl: true, |
| 5486 | cs_precedes: lc->int_p_cs_precedes, |
| 5487 | sep_by_space: lc->int_p_sep_by_space, |
| 5488 | sign_posn: lc->int_p_sign_posn, |
| 5489 | space_char: L' '); |
| 5490 | __init_pat( |
| 5491 | pat&: __neg_format_, __curr_symbol_, intl: true, cs_precedes: lc->int_n_cs_precedes, sep_by_space: lc->int_n_sep_by_space, sign_posn: lc->int_n_sign_posn, space_char: L' '); |
| 5492 | # endif // _WIN32 |
| 5493 | } |
| 5494 | #endif // _LIBCPP_HAS_WIDE_CHARACTERS |
| 5495 | |
| 5496 | void __do_nothing(void*) {} |
| 5497 | |
| 5498 | // Legacy ABI __num_get functions - the new ones are _LIBCPP_HIDE_FROM_ABI |
| 5499 | template <class _CharT> |
| 5500 | string __num_get<_CharT>::__stage2_int_prep(ios_base& __iob, _CharT* __atoms, _CharT& __thousands_sep) { |
| 5501 | locale __loc = __iob.getloc(); |
| 5502 | std::use_facet<ctype<_CharT> >(__loc).widen(__src, __src + __int_chr_cnt, __atoms); |
| 5503 | const numpunct<_CharT>& __np = std::use_facet<numpunct<_CharT> >(__loc); |
| 5504 | __thousands_sep = __np.thousands_sep(); |
| 5505 | return __np.grouping(); |
| 5506 | } |
| 5507 | |
| 5508 | template <class _CharT> |
| 5509 | int __num_get<_CharT>::__stage2_int_loop( |
| 5510 | _CharT __ct, |
| 5511 | int __base, |
| 5512 | char* __a, |
| 5513 | char*& __a_end, |
| 5514 | unsigned& __dc, |
| 5515 | _CharT __thousands_sep, |
| 5516 | const string& __grouping, |
| 5517 | unsigned* __g, |
| 5518 | unsigned*& __g_end, |
| 5519 | _CharT* __atoms) { |
| 5520 | if (__a_end == __a && (__ct == __atoms[24] || __ct == __atoms[25])) { |
| 5521 | *__a_end++ = __ct == __atoms[24] ? '+' : '-'; |
| 5522 | __dc = 0; |
| 5523 | return 0; |
| 5524 | } |
| 5525 | if (__grouping.size() != 0 && __ct == __thousands_sep) { |
| 5526 | if (__g_end - __g < __num_get_buf_sz) { |
| 5527 | *__g_end++ = __dc; |
| 5528 | __dc = 0; |
| 5529 | } |
| 5530 | return 0; |
| 5531 | } |
| 5532 | ptrdiff_t __f = __atoms_offset(__atoms, val: __ct); |
| 5533 | if (__f >= 24) |
| 5534 | return -1; |
| 5535 | switch (__base) { |
| 5536 | case 8: |
| 5537 | case 10: |
| 5538 | if (__f >= __base) |
| 5539 | return -1; |
| 5540 | break; |
| 5541 | case 16: |
| 5542 | if (__f < 22) |
| 5543 | break; |
| 5544 | if (__a_end != __a && __a_end - __a <= 2 && __a_end[-1] == '0') { |
| 5545 | __dc = 0; |
| 5546 | *__a_end++ = __src[__f]; |
| 5547 | return 0; |
| 5548 | } |
| 5549 | return -1; |
| 5550 | } |
| 5551 | *__a_end++ = __src[__f]; |
| 5552 | ++__dc; |
| 5553 | return 0; |
| 5554 | } |
| 5555 | |
| 5556 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS collate<char>; |
| 5557 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS collate<wchar_t>;) |
| 5558 | |
| 5559 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS num_get<char>; |
| 5560 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS num_get<wchar_t>;) |
| 5561 | |
| 5562 | template struct _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __num_get<char>; |
| 5563 | _LIBCPP_IF_WIDE_CHARACTERS(template struct _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __num_get<wchar_t>;) |
| 5564 | |
| 5565 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS num_put<char>; |
| 5566 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS num_put<wchar_t>;) |
| 5567 | |
| 5568 | template struct _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __num_put<char>; |
| 5569 | _LIBCPP_IF_WIDE_CHARACTERS(template struct _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __num_put<wchar_t>;) |
| 5570 | |
| 5571 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_get<char>; |
| 5572 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_get<wchar_t>;) |
| 5573 | |
| 5574 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_get_byname<char>; |
| 5575 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_get_byname<wchar_t>;) |
| 5576 | |
| 5577 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_put<char>; |
| 5578 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_put<wchar_t>;) |
| 5579 | |
| 5580 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_put_byname<char>; |
| 5581 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS time_put_byname<wchar_t>;) |
| 5582 | |
| 5583 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct<char, false>; |
| 5584 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct<char, true>; |
| 5585 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct<wchar_t, false>;) |
| 5586 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct<wchar_t, true>;) |
| 5587 | |
| 5588 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct_byname<char, false>; |
| 5589 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct_byname<char, true>; |
| 5590 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct_byname<wchar_t, false>;) |
| 5591 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS moneypunct_byname<wchar_t, true>;) |
| 5592 | |
| 5593 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS money_get<char>; |
| 5594 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS money_get<wchar_t>;) |
| 5595 | |
| 5596 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __money_get<char>; |
| 5597 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __money_get<wchar_t>;) |
| 5598 | |
| 5599 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS money_put<char>; |
| 5600 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS money_put<wchar_t>;) |
| 5601 | |
| 5602 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __money_put<char>; |
| 5603 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __money_put<wchar_t>;) |
| 5604 | |
| 5605 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS messages<char>; |
| 5606 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS messages<wchar_t>;) |
| 5607 | |
| 5608 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS messages_byname<char>; |
| 5609 | _LIBCPP_IF_WIDE_CHARACTERS(template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS messages_byname<wchar_t>;) |
| 5610 | |
| 5611 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS codecvt_byname<char, char, mbstate_t>; |
| 5612 | _LIBCPP_IF_WIDE_CHARACTERS( |
| 5613 | template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS codecvt_byname<wchar_t, char, mbstate_t>;) |
| 5614 | template class _LIBCPP_DEPRECATED_IN_CXX20 _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS |
| 5615 | codecvt_byname<char16_t, char, mbstate_t>; |
| 5616 | template class _LIBCPP_DEPRECATED_IN_CXX20 _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS |
| 5617 | codecvt_byname<char32_t, char, mbstate_t>; |
| 5618 | #if _LIBCPP_HAS_CHAR8_T |
| 5619 | template class _LIBCPP_DEPRECATED_IN_CXX20 _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS |
| 5620 | codecvt_byname<char16_t, char8_t, mbstate_t>; |
| 5621 | template class _LIBCPP_DEPRECATED_IN_CXX20 _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS |
| 5622 | codecvt_byname<char32_t, char8_t, mbstate_t>; |
| 5623 | #endif |
| 5624 | |
| 5625 | _LIBCPP_END_EXPLICIT_ABI_ANNOTATIONS |
| 5626 | _LIBCPP_END_NAMESPACE_STD |
| 5627 | |
| 5628 | _LIBCPP_POP_MACROS |
| 5629 |
Generated on 2026-Oct-20 from project llvm_runtimes revision 7be7bc433
Powered by 
Code Browser 2.1
Generator usage only permitted with license.