| 1 | // Copyright 2007, Google Inc. |
| 2 | // All rights reserved. |
| 3 | // |
| 4 | // Redistribution and use in source and binary forms, with or without |
| 5 | // modification, are permitted provided that the following conditions are |
| 6 | // met: |
| 7 | // |
| 8 | // * Redistributions of source code must retain the above copyright |
| 9 | // notice, this list of conditions and the following disclaimer. |
| 10 | // * Redistributions in binary form must reproduce the above |
| 11 | // copyright notice, this list of conditions and the following disclaimer |
| 12 | // in the documentation and/or other materials provided with the |
| 13 | // distribution. |
| 14 | // * Neither the name of Google Inc. nor the names of its |
| 15 | // contributors may be used to endorse or promote products derived from |
| 16 | // this software without specific prior written permission. |
| 17 | // |
| 18 | // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| 19 | // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| 20 | // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
| 21 | // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
| 22 | // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
| 23 | // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
| 24 | // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
| 25 | // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
| 26 | // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 27 | // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 28 | // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 29 | |
| 30 | // Google Test - The Google C++ Testing and Mocking Framework |
| 31 | // |
| 32 | // This file implements a universal value printer that can print a |
| 33 | // value of any type T: |
| 34 | // |
| 35 | // void ::testing::internal::UniversalPrinter<T>::Print(value, ostream_ptr); |
| 36 | // |
| 37 | // It uses the << operator when possible, and prints the bytes in the |
| 38 | // object otherwise. A user can override its behavior for a class |
| 39 | // type Foo by defining either operator<<(::std::ostream&, const Foo&) |
| 40 | // or void PrintTo(const Foo&, ::std::ostream*) in the namespace that |
| 41 | // defines Foo. |
| 42 | |
| 43 | #include "gtest/gtest-printers.h" |
| 44 | |
| 45 | #include <stdio.h> |
| 46 | |
| 47 | #include <cctype> |
| 48 | #include <cstdint> |
| 49 | #include <cwchar> |
| 50 | #include <iomanip> |
| 51 | #include <ios> |
| 52 | #include <ostream> // NOLINT |
| 53 | #include <string> |
| 54 | #include <type_traits> |
| 55 | |
| 56 | #include "gtest/internal/gtest-port.h" |
| 57 | #include "src/gtest-internal-inl.h" |
| 58 | |
| 59 | namespace testing { |
| 60 | |
| 61 | namespace { |
| 62 | |
| 63 | using ::std::ostream; |
| 64 | |
| 65 | // Prints a segment of bytes in the given object. |
| 66 | GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_ |
| 67 | GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_ |
| 68 | GTEST_ATTRIBUTE_NO_SANITIZE_HWADDRESS_ |
| 69 | GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_ |
| 70 | void PrintByteSegmentInObjectTo(const unsigned char* obj_bytes, size_t start, |
| 71 | size_t count, ostream* os) { |
| 72 | char text[5] = "" ; |
| 73 | for (size_t i = 0; i != count; i++) { |
| 74 | const size_t j = start + i; |
| 75 | if (i != 0) { |
| 76 | // Organizes the bytes into groups of 2 for easy parsing by |
| 77 | // human. |
| 78 | if ((j % 2) == 0) |
| 79 | *os << ' '; |
| 80 | else |
| 81 | *os << '-'; |
| 82 | } |
| 83 | GTEST_SNPRINTF_(s: text, maxlen: sizeof(text), format: "%02X" , obj_bytes[j]); |
| 84 | *os << text; |
| 85 | } |
| 86 | } |
| 87 | |
| 88 | // Prints the bytes in the given value to the given ostream. |
| 89 | void PrintBytesInObjectToImpl(const unsigned char* obj_bytes, size_t count, |
| 90 | ostream* os) { |
| 91 | // Tells the user how big the object is. |
| 92 | *os << count << "-byte object <" ; |
| 93 | |
| 94 | const size_t kThreshold = 132; |
| 95 | const size_t kChunkSize = 64; |
| 96 | // If the object size is bigger than kThreshold, we'll have to omit |
| 97 | // some details by printing only the first and the last kChunkSize |
| 98 | // bytes. |
| 99 | if (count < kThreshold) { |
| 100 | PrintByteSegmentInObjectTo(obj_bytes, start: 0, count, os); |
| 101 | } else { |
| 102 | PrintByteSegmentInObjectTo(obj_bytes, start: 0, count: kChunkSize, os); |
| 103 | *os << " ... " ; |
| 104 | // Rounds up to 2-byte boundary. |
| 105 | const size_t resume_pos = (count - kChunkSize + 1) / 2 * 2; |
| 106 | PrintByteSegmentInObjectTo(obj_bytes, start: resume_pos, count: count - resume_pos, os); |
| 107 | } |
| 108 | *os << ">" ; |
| 109 | } |
| 110 | |
| 111 | // Helpers for widening a character to char32_t. Since the standard does not |
| 112 | // specify if char / wchar_t is signed or unsigned, it is important to first |
| 113 | // convert it to the unsigned type of the same width before widening it to |
| 114 | // char32_t. |
| 115 | template <typename CharType> |
| 116 | char32_t ToChar32(CharType in) { |
| 117 | return static_cast<char32_t>( |
| 118 | static_cast<typename std::make_unsigned<CharType>::type>(in)); |
| 119 | } |
| 120 | |
| 121 | } // namespace |
| 122 | |
| 123 | namespace internal { |
| 124 | |
| 125 | // Delegates to PrintBytesInObjectToImpl() to print the bytes in the |
| 126 | // given object. The delegation simplifies the implementation, which |
| 127 | // uses the << operator and thus is easier done outside of the |
| 128 | // ::testing::internal namespace, which contains a << operator that |
| 129 | // sometimes conflicts with the one in STL. |
| 130 | void PrintBytesInObjectTo(const unsigned char* obj_bytes, size_t count, |
| 131 | ostream* os) { |
| 132 | PrintBytesInObjectToImpl(obj_bytes, count, os); |
| 133 | } |
| 134 | |
| 135 | // Depending on the value of a char (or wchar_t), we print it in one |
| 136 | // of three formats: |
| 137 | // - as is if it's a printable ASCII (e.g. 'a', '2', ' '), |
| 138 | // - as a hexadecimal escape sequence (e.g. '\x7F'), or |
| 139 | // - as a special escape sequence (e.g. '\r', '\n'). |
| 140 | enum CharFormat { kAsIs, kHexEscape, kSpecialEscape }; |
| 141 | |
| 142 | // Returns true if c is a printable ASCII character. We test the |
| 143 | // value of c directly instead of calling isprint(), which is buggy on |
| 144 | // Windows Mobile. |
| 145 | inline bool IsPrintableAscii(char32_t c) { return 0x20 <= c && c <= 0x7E; } |
| 146 | |
| 147 | // Prints c (of type char, char8_t, char16_t, char32_t, or wchar_t) as a |
| 148 | // character literal without the quotes, escaping it when necessary; returns how |
| 149 | // c was formatted. |
| 150 | template <typename Char> |
| 151 | static CharFormat PrintAsCharLiteralTo(Char c, ostream* os) { |
| 152 | const char32_t u_c = ToChar32(c); |
| 153 | switch (u_c) { |
| 154 | case L'\0': |
| 155 | *os << "\\0" ; |
| 156 | break; |
| 157 | case L'\'': |
| 158 | *os << "\\'" ; |
| 159 | break; |
| 160 | case L'\\': |
| 161 | *os << "\\\\" ; |
| 162 | break; |
| 163 | case L'\a': |
| 164 | *os << "\\a" ; |
| 165 | break; |
| 166 | case L'\b': |
| 167 | *os << "\\b" ; |
| 168 | break; |
| 169 | case L'\f': |
| 170 | *os << "\\f" ; |
| 171 | break; |
| 172 | case L'\n': |
| 173 | *os << "\\n" ; |
| 174 | break; |
| 175 | case L'\r': |
| 176 | *os << "\\r" ; |
| 177 | break; |
| 178 | case L'\t': |
| 179 | *os << "\\t" ; |
| 180 | break; |
| 181 | case L'\v': |
| 182 | *os << "\\v" ; |
| 183 | break; |
| 184 | default: |
| 185 | if (IsPrintableAscii(c: u_c)) { |
| 186 | *os << static_cast<char>(c); |
| 187 | return kAsIs; |
| 188 | } else { |
| 189 | ostream::fmtflags flags = os->flags(); |
| 190 | *os << "\\x" << std::hex << std::uppercase << static_cast<int>(u_c); |
| 191 | os->flags(fmtfl: flags); |
| 192 | return kHexEscape; |
| 193 | } |
| 194 | } |
| 195 | return kSpecialEscape; |
| 196 | } |
| 197 | |
| 198 | // Prints a char32_t c as if it's part of a string literal, escaping it when |
| 199 | // necessary; returns how c was formatted. |
| 200 | static CharFormat PrintAsStringLiteralTo(char32_t c, ostream* os) { |
| 201 | switch (c) { |
| 202 | case L'\'': |
| 203 | *os << "'" ; |
| 204 | return kAsIs; |
| 205 | case L'"': |
| 206 | *os << "\\\"" ; |
| 207 | return kSpecialEscape; |
| 208 | default: |
| 209 | return PrintAsCharLiteralTo(c, os); |
| 210 | } |
| 211 | } |
| 212 | |
| 213 | static const char* GetCharWidthPrefix(char) { return "" ; } |
| 214 | |
| 215 | static const char* GetCharWidthPrefix(signed char) { return "" ; } |
| 216 | |
| 217 | static const char* GetCharWidthPrefix(unsigned char) { return "" ; } |
| 218 | |
| 219 | #ifdef __cpp_lib_char8_t |
| 220 | static const char* GetCharWidthPrefix(char8_t) { return "u8" ; } |
| 221 | #endif |
| 222 | |
| 223 | static const char* GetCharWidthPrefix(char16_t) { return "u" ; } |
| 224 | |
| 225 | static const char* GetCharWidthPrefix(char32_t) { return "U" ; } |
| 226 | |
| 227 | static const char* GetCharWidthPrefix(wchar_t) { return "L" ; } |
| 228 | |
| 229 | // Prints a char c as if it's part of a string literal, escaping it when |
| 230 | // necessary; returns how c was formatted. |
| 231 | static CharFormat PrintAsStringLiteralTo(char c, ostream* os) { |
| 232 | return PrintAsStringLiteralTo(c: ToChar32(in: c), os); |
| 233 | } |
| 234 | |
| 235 | #ifdef __cpp_lib_char8_t |
| 236 | static CharFormat PrintAsStringLiteralTo(char8_t c, ostream* os) { |
| 237 | return PrintAsStringLiteralTo(ToChar32(c), os); |
| 238 | } |
| 239 | #endif |
| 240 | |
| 241 | static CharFormat PrintAsStringLiteralTo(char16_t c, ostream* os) { |
| 242 | return PrintAsStringLiteralTo(c: ToChar32(in: c), os); |
| 243 | } |
| 244 | |
| 245 | static CharFormat PrintAsStringLiteralTo(wchar_t c, ostream* os) { |
| 246 | return PrintAsStringLiteralTo(c: ToChar32(in: c), os); |
| 247 | } |
| 248 | |
| 249 | // Prints a character c (of type char, char8_t, char16_t, char32_t, or wchar_t) |
| 250 | // and its code. '\0' is printed as "'\\0'", other unprintable characters are |
| 251 | // also properly escaped using the standard C++ escape sequence. |
| 252 | template <typename Char> |
| 253 | void PrintCharAndCodeTo(Char c, ostream* os) { |
| 254 | // First, print c as a literal in the most readable form we can find. |
| 255 | *os << GetCharWidthPrefix(c) << "'" ; |
| 256 | const CharFormat format = PrintAsCharLiteralTo(c, os); |
| 257 | *os << "'" ; |
| 258 | |
| 259 | // To aid user debugging, we also print c's code in decimal, unless |
| 260 | // it's 0 (in which case c was printed as '\\0', making the code |
| 261 | // obvious). |
| 262 | if (c == 0) return; |
| 263 | *os << " (" << static_cast<int>(c); |
| 264 | |
| 265 | // For more convenience, we print c's code again in hexadecimal, |
| 266 | // unless c was already printed in the form '\x##' or the code is in |
| 267 | // [1, 9]. |
| 268 | if (format == kHexEscape || (1 <= c && c <= 9)) { |
| 269 | // Do nothing. |
| 270 | } else { |
| 271 | *os << ", 0x" << String::FormatHexInt(value: static_cast<int>(c)); |
| 272 | } |
| 273 | *os << ")" ; |
| 274 | } |
| 275 | |
| 276 | void PrintTo(unsigned char c, ::std::ostream* os) { PrintCharAndCodeTo(c, os); } |
| 277 | void PrintTo(signed char c, ::std::ostream* os) { PrintCharAndCodeTo(c, os); } |
| 278 | |
| 279 | // Prints a wchar_t as a symbol if it is printable or as its internal |
| 280 | // code otherwise and also as its code. L'\0' is printed as "L'\\0'". |
| 281 | void PrintTo(wchar_t wc, ostream* os) { PrintCharAndCodeTo(c: wc, os); } |
| 282 | |
| 283 | // TODO(dcheng): Consider making this delegate to PrintCharAndCodeTo() as well. |
| 284 | void PrintTo(char32_t c, ::std::ostream* os) { |
| 285 | *os << std::hex << "U+" << std::uppercase << std::setfill('0') << std::setw(4) |
| 286 | << static_cast<uint32_t>(c); |
| 287 | } |
| 288 | |
| 289 | // gcc/clang __{u,}int128_t |
| 290 | #if defined(__SIZEOF_INT128__) |
| 291 | void PrintTo(__uint128_t v, ::std::ostream* os) { |
| 292 | if (v == 0) { |
| 293 | *os << "0" ; |
| 294 | return; |
| 295 | } |
| 296 | |
| 297 | // Buffer large enough for ceil(log10(2^128))==39 and the null terminator |
| 298 | char buf[40]; |
| 299 | char* p = buf + sizeof(buf); |
| 300 | |
| 301 | // Some configurations have a __uint128_t, but no support for built in |
| 302 | // division. Do manual long division instead. |
| 303 | |
| 304 | uint64_t high = static_cast<uint64_t>(v >> 64); |
| 305 | uint64_t low = static_cast<uint64_t>(v); |
| 306 | |
| 307 | *--p = 0; |
| 308 | while (high != 0 || low != 0) { |
| 309 | uint64_t high_mod = high % 10; |
| 310 | high = high / 10; |
| 311 | // This is the long division algorithm specialized for a divisor of 10 and |
| 312 | // only two elements. |
| 313 | // Notable values: |
| 314 | // 2^64 / 10 == 1844674407370955161 |
| 315 | // 2^64 % 10 == 6 |
| 316 | const uint64_t carry = 6 * high_mod + low % 10; |
| 317 | low = low / 10 + high_mod * 1844674407370955161 + carry / 10; |
| 318 | |
| 319 | char digit = static_cast<char>(carry % 10); |
| 320 | *--p = static_cast<char>('0' + digit); |
| 321 | } |
| 322 | *os << p; |
| 323 | } |
| 324 | void PrintTo(__int128_t v, ::std::ostream* os) { |
| 325 | __uint128_t uv = static_cast<__uint128_t>(v); |
| 326 | if (v < 0) { |
| 327 | *os << "-" ; |
| 328 | uv = -uv; |
| 329 | } |
| 330 | PrintTo(v: uv, os); |
| 331 | } |
| 332 | #endif // __SIZEOF_INT128__ |
| 333 | |
| 334 | // Prints the given array of characters to the ostream. CharType must be either |
| 335 | // char, char8_t, char16_t, char32_t, or wchar_t. |
| 336 | // The array starts at begin, the length is len, it may include '\0' characters |
| 337 | // and may not be NUL-terminated. |
| 338 | template <typename CharType> |
| 339 | GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_ GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_ |
| 340 | GTEST_ATTRIBUTE_NO_SANITIZE_HWADDRESS_ |
| 341 | GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_ static CharFormat |
| 342 | PrintCharsAsStringTo(const CharType* begin, size_t len, ostream* os) { |
| 343 | const char* const quote_prefix = GetCharWidthPrefix(*begin); |
| 344 | *os << quote_prefix << "\"" ; |
| 345 | bool is_previous_hex = false; |
| 346 | CharFormat print_format = kAsIs; |
| 347 | for (size_t index = 0; index < len; ++index) { |
| 348 | const CharType cur = begin[index]; |
| 349 | if (is_previous_hex && IsXDigit(cur)) { |
| 350 | // Previous character is of '\x..' form and this character can be |
| 351 | // interpreted as another hexadecimal digit in its number. Break string to |
| 352 | // disambiguate. |
| 353 | *os << "\" " << quote_prefix << "\"" ; |
| 354 | } |
| 355 | is_previous_hex = PrintAsStringLiteralTo(cur, os) == kHexEscape; |
| 356 | // Remember if any characters required hex escaping. |
| 357 | if (is_previous_hex) { |
| 358 | print_format = kHexEscape; |
| 359 | } |
| 360 | } |
| 361 | *os << "\"" ; |
| 362 | return print_format; |
| 363 | } |
| 364 | |
| 365 | // Prints a (const) char/wchar_t array of 'len' elements, starting at address |
| 366 | // 'begin'. CharType must be either char or wchar_t. |
| 367 | template <typename CharType> |
| 368 | GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_ GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_ |
| 369 | GTEST_ATTRIBUTE_NO_SANITIZE_HWADDRESS_ |
| 370 | GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_ static void |
| 371 | UniversalPrintCharArray(const CharType* begin, size_t len, |
| 372 | ostream* os) { |
| 373 | // The code |
| 374 | // const char kFoo[] = "foo"; |
| 375 | // generates an array of 4, not 3, elements, with the last one being '\0'. |
| 376 | // |
| 377 | // Therefore when printing a char array, we don't print the last element if |
| 378 | // it's '\0', such that the output matches the string literal as it's |
| 379 | // written in the source code. |
| 380 | if (len > 0 && begin[len - 1] == '\0') { |
| 381 | PrintCharsAsStringTo(begin, len - 1, os); |
| 382 | return; |
| 383 | } |
| 384 | |
| 385 | // If, however, the last element in the array is not '\0', e.g. |
| 386 | // const char kFoo[] = { 'f', 'o', 'o' }; |
| 387 | // we must print the entire array. We also print a message to indicate |
| 388 | // that the array is not NUL-terminated. |
| 389 | PrintCharsAsStringTo(begin, len, os); |
| 390 | *os << " (no terminating NUL)" ; |
| 391 | } |
| 392 | |
| 393 | // Prints a (const) char array of 'len' elements, starting at address 'begin'. |
| 394 | void UniversalPrintArray(const char* begin, size_t len, ostream* os) { |
| 395 | UniversalPrintCharArray(begin, len, os); |
| 396 | } |
| 397 | |
| 398 | #ifdef __cpp_lib_char8_t |
| 399 | // Prints a (const) char8_t array of 'len' elements, starting at address |
| 400 | // 'begin'. |
| 401 | void UniversalPrintArray(const char8_t* begin, size_t len, ostream* os) { |
| 402 | UniversalPrintCharArray(begin, len, os); |
| 403 | } |
| 404 | #endif |
| 405 | |
| 406 | // Prints a (const) char16_t array of 'len' elements, starting at address |
| 407 | // 'begin'. |
| 408 | void UniversalPrintArray(const char16_t* begin, size_t len, ostream* os) { |
| 409 | UniversalPrintCharArray(begin, len, os); |
| 410 | } |
| 411 | |
| 412 | // Prints a (const) char32_t array of 'len' elements, starting at address |
| 413 | // 'begin'. |
| 414 | void UniversalPrintArray(const char32_t* begin, size_t len, ostream* os) { |
| 415 | UniversalPrintCharArray(begin, len, os); |
| 416 | } |
| 417 | |
| 418 | // Prints a (const) wchar_t array of 'len' elements, starting at address |
| 419 | // 'begin'. |
| 420 | void UniversalPrintArray(const wchar_t* begin, size_t len, ostream* os) { |
| 421 | UniversalPrintCharArray(begin, len, os); |
| 422 | } |
| 423 | |
| 424 | namespace { |
| 425 | |
| 426 | // Prints a null-terminated C-style string to the ostream. |
| 427 | template <typename Char> |
| 428 | void PrintCStringTo(const Char* s, ostream* os) { |
| 429 | if (s == nullptr) { |
| 430 | *os << "NULL" ; |
| 431 | } else { |
| 432 | *os << ImplicitCast_<const void*>(s) << " pointing to " ; |
| 433 | PrintCharsAsStringTo(s, std::char_traits<Char>::length(s), os); |
| 434 | } |
| 435 | } |
| 436 | |
| 437 | } // anonymous namespace |
| 438 | |
| 439 | void PrintTo(const char* s, ostream* os) { PrintCStringTo(s, os); } |
| 440 | |
| 441 | #ifdef __cpp_lib_char8_t |
| 442 | void PrintTo(const char8_t* s, ostream* os) { PrintCStringTo(s, os); } |
| 443 | #endif |
| 444 | |
| 445 | void PrintTo(const char16_t* s, ostream* os) { PrintCStringTo(s, os); } |
| 446 | |
| 447 | void PrintTo(const char32_t* s, ostream* os) { PrintCStringTo(s, os); } |
| 448 | |
| 449 | // MSVC compiler can be configured to define whar_t as a typedef |
| 450 | // of unsigned short. Defining an overload for const wchar_t* in that case |
| 451 | // would cause pointers to unsigned shorts be printed as wide strings, |
| 452 | // possibly accessing more memory than intended and causing invalid |
| 453 | // memory accesses. MSVC defines _NATIVE_WCHAR_T_DEFINED symbol when |
| 454 | // wchar_t is implemented as a native type. |
| 455 | #if !defined(_MSC_VER) || defined(_NATIVE_WCHAR_T_DEFINED) |
| 456 | // Prints the given wide C string to the ostream. |
| 457 | void PrintTo(const wchar_t* s, ostream* os) { PrintCStringTo(s, os); } |
| 458 | #endif // wchar_t is native |
| 459 | |
| 460 | namespace { |
| 461 | |
| 462 | bool ContainsUnprintableControlCodes(const char* str, size_t length) { |
| 463 | const unsigned char* s = reinterpret_cast<const unsigned char*>(str); |
| 464 | |
| 465 | for (size_t i = 0; i < length; i++) { |
| 466 | unsigned char ch = *s++; |
| 467 | if (std::iscntrl(ch)) { |
| 468 | switch (ch) { |
| 469 | case '\t': |
| 470 | case '\n': |
| 471 | case '\r': |
| 472 | break; |
| 473 | default: |
| 474 | return true; |
| 475 | } |
| 476 | } |
| 477 | } |
| 478 | return false; |
| 479 | } |
| 480 | |
| 481 | bool IsUTF8TrailByte(unsigned char t) { return 0x80 <= t && t <= 0xbf; } |
| 482 | |
| 483 | bool IsValidUTF8(const char* str, size_t length) { |
| 484 | const unsigned char* s = reinterpret_cast<const unsigned char*>(str); |
| 485 | |
| 486 | for (size_t i = 0; i < length;) { |
| 487 | unsigned char lead = s[i++]; |
| 488 | |
| 489 | if (lead <= 0x7f) { |
| 490 | continue; // single-byte character (ASCII) 0..7F |
| 491 | } |
| 492 | if (lead < 0xc2) { |
| 493 | return false; // trail byte or non-shortest form |
| 494 | } else if (lead <= 0xdf && (i + 1) <= length && IsUTF8TrailByte(t: s[i])) { |
| 495 | ++i; // 2-byte character |
| 496 | } else if (0xe0 <= lead && lead <= 0xef && (i + 2) <= length && |
| 497 | IsUTF8TrailByte(t: s[i]) && IsUTF8TrailByte(t: s[i + 1]) && |
| 498 | // check for non-shortest form and surrogate |
| 499 | (lead != 0xe0 || s[i] >= 0xa0) && |
| 500 | (lead != 0xed || s[i] < 0xa0)) { |
| 501 | i += 2; // 3-byte character |
| 502 | } else if (0xf0 <= lead && lead <= 0xf4 && (i + 3) <= length && |
| 503 | IsUTF8TrailByte(t: s[i]) && IsUTF8TrailByte(t: s[i + 1]) && |
| 504 | IsUTF8TrailByte(t: s[i + 2]) && |
| 505 | // check for non-shortest form |
| 506 | (lead != 0xf0 || s[i] >= 0x90) && |
| 507 | (lead != 0xf4 || s[i] < 0x90)) { |
| 508 | i += 3; // 4-byte character |
| 509 | } else { |
| 510 | return false; |
| 511 | } |
| 512 | } |
| 513 | return true; |
| 514 | } |
| 515 | |
| 516 | void ConditionalPrintAsText(const char* str, size_t length, ostream* os) { |
| 517 | if (!ContainsUnprintableControlCodes(str, length) && |
| 518 | IsValidUTF8(str, length)) { |
| 519 | *os << "\n As Text: \"" << str << "\"" ; |
| 520 | } |
| 521 | } |
| 522 | |
| 523 | } // anonymous namespace |
| 524 | |
| 525 | void PrintStringTo(const ::std::string& s, ostream* os) { |
| 526 | if (PrintCharsAsStringTo(begin: s.data(), len: s.size(), os) == kHexEscape) { |
| 527 | if (GTEST_FLAG_GET(print_utf8)) { |
| 528 | ConditionalPrintAsText(str: s.data(), length: s.size(), os); |
| 529 | } |
| 530 | } |
| 531 | } |
| 532 | |
| 533 | #ifdef __cpp_lib_char8_t |
| 534 | void PrintU8StringTo(const ::std::u8string& s, ostream* os) { |
| 535 | PrintCharsAsStringTo(s.data(), s.size(), os); |
| 536 | } |
| 537 | #endif |
| 538 | |
| 539 | void PrintU16StringTo(const ::std::u16string& s, ostream* os) { |
| 540 | PrintCharsAsStringTo(begin: s.data(), len: s.size(), os); |
| 541 | } |
| 542 | |
| 543 | void PrintU32StringTo(const ::std::u32string& s, ostream* os) { |
| 544 | PrintCharsAsStringTo(begin: s.data(), len: s.size(), os); |
| 545 | } |
| 546 | |
| 547 | #if GTEST_HAS_STD_WSTRING |
| 548 | void PrintWideStringTo(const ::std::wstring& s, ostream* os) { |
| 549 | PrintCharsAsStringTo(begin: s.data(), len: s.size(), os); |
| 550 | } |
| 551 | #endif // GTEST_HAS_STD_WSTRING |
| 552 | |
| 553 | } // namespace internal |
| 554 | |
| 555 | } // namespace testing |
| 556 | |