yaml-cpp

FORK: A YAML parser and emitter in C++
git clone https://git.neptards.moe/neptards/yaml-cpp.git
Log | Files | Refs | README | LICENSE

gtest-printers.cc (14711B)


      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 
     31 // Google Test - The Google C++ Testing and Mocking Framework
     32 //
     33 // This file implements a universal value printer that can print a
     34 // value of any type T:
     35 //
     36 //   void ::testing::internal::UniversalPrinter<T>::Print(value, ostream_ptr);
     37 //
     38 // It uses the << operator when possible, and prints the bytes in the
     39 // object otherwise.  A user can override its behavior for a class
     40 // type Foo by defining either operator<<(::std::ostream&, const Foo&)
     41 // or void PrintTo(const Foo&, ::std::ostream*) in the namespace that
     42 // defines Foo.
     43 
     44 #include "gtest/gtest-printers.h"
     45 #include <stdio.h>
     46 #include <cctype>
     47 #include <cwchar>
     48 #include <ostream>  // NOLINT
     49 #include <string>
     50 #include "gtest/internal/gtest-port.h"
     51 #include "src/gtest-internal-inl.h"
     52 
     53 namespace testing {
     54 
     55 namespace {
     56 
     57 using ::std::ostream;
     58 
     59 // Prints a segment of bytes in the given object.
     60 GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_
     61 GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_
     62 GTEST_ATTRIBUTE_NO_SANITIZE_HWADDRESS_
     63 GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_
     64 void PrintByteSegmentInObjectTo(const unsigned char* obj_bytes, size_t start,
     65                                 size_t count, ostream* os) {
     66   char text[5] = "";
     67   for (size_t i = 0; i != count; i++) {
     68     const size_t j = start + i;
     69     if (i != 0) {
     70       // Organizes the bytes into groups of 2 for easy parsing by
     71       // human.
     72       if ((j % 2) == 0)
     73         *os << ' ';
     74       else
     75         *os << '-';
     76     }
     77     GTEST_SNPRINTF_(text, sizeof(text), "%02X", obj_bytes[j]);
     78     *os << text;
     79   }
     80 }
     81 
     82 // Prints the bytes in the given value to the given ostream.
     83 void PrintBytesInObjectToImpl(const unsigned char* obj_bytes, size_t count,
     84                               ostream* os) {
     85   // Tells the user how big the object is.
     86   *os << count << "-byte object <";
     87 
     88   const size_t kThreshold = 132;
     89   const size_t kChunkSize = 64;
     90   // If the object size is bigger than kThreshold, we'll have to omit
     91   // some details by printing only the first and the last kChunkSize
     92   // bytes.
     93   if (count < kThreshold) {
     94     PrintByteSegmentInObjectTo(obj_bytes, 0, count, os);
     95   } else {
     96     PrintByteSegmentInObjectTo(obj_bytes, 0, kChunkSize, os);
     97     *os << " ... ";
     98     // Rounds up to 2-byte boundary.
     99     const size_t resume_pos = (count - kChunkSize + 1)/2*2;
    100     PrintByteSegmentInObjectTo(obj_bytes, resume_pos, count - resume_pos, os);
    101   }
    102   *os << ">";
    103 }
    104 
    105 }  // namespace
    106 
    107 namespace internal2 {
    108 
    109 // Delegates to PrintBytesInObjectToImpl() to print the bytes in the
    110 // given object.  The delegation simplifies the implementation, which
    111 // uses the << operator and thus is easier done outside of the
    112 // ::testing::internal namespace, which contains a << operator that
    113 // sometimes conflicts with the one in STL.
    114 void PrintBytesInObjectTo(const unsigned char* obj_bytes, size_t count,
    115                           ostream* os) {
    116   PrintBytesInObjectToImpl(obj_bytes, count, os);
    117 }
    118 
    119 }  // namespace internal2
    120 
    121 namespace internal {
    122 
    123 // Depending on the value of a char (or wchar_t), we print it in one
    124 // of three formats:
    125 //   - as is if it's a printable ASCII (e.g. 'a', '2', ' '),
    126 //   - as a hexadecimal escape sequence (e.g. '\x7F'), or
    127 //   - as a special escape sequence (e.g. '\r', '\n').
    128 enum CharFormat {
    129   kAsIs,
    130   kHexEscape,
    131   kSpecialEscape
    132 };
    133 
    134 // Returns true if c is a printable ASCII character.  We test the
    135 // value of c directly instead of calling isprint(), which is buggy on
    136 // Windows Mobile.
    137 inline bool IsPrintableAscii(wchar_t c) {
    138   return 0x20 <= c && c <= 0x7E;
    139 }
    140 
    141 // Prints a wide or narrow char c as a character literal without the
    142 // quotes, escaping it when necessary; returns how c was formatted.
    143 // The template argument UnsignedChar is the unsigned version of Char,
    144 // which is the type of c.
    145 template <typename UnsignedChar, typename Char>
    146 static CharFormat PrintAsCharLiteralTo(Char c, ostream* os) {
    147   wchar_t w_c = static_cast<wchar_t>(c);
    148   switch (w_c) {
    149     case L'\0':
    150       *os << "\\0";
    151       break;
    152     case L'\'':
    153       *os << "\\'";
    154       break;
    155     case L'\\':
    156       *os << "\\\\";
    157       break;
    158     case L'\a':
    159       *os << "\\a";
    160       break;
    161     case L'\b':
    162       *os << "\\b";
    163       break;
    164     case L'\f':
    165       *os << "\\f";
    166       break;
    167     case L'\n':
    168       *os << "\\n";
    169       break;
    170     case L'\r':
    171       *os << "\\r";
    172       break;
    173     case L'\t':
    174       *os << "\\t";
    175       break;
    176     case L'\v':
    177       *os << "\\v";
    178       break;
    179     default:
    180       if (IsPrintableAscii(w_c)) {
    181         *os << static_cast<char>(c);
    182         return kAsIs;
    183       } else {
    184         ostream::fmtflags flags = os->flags();
    185         *os << "\\x" << std::hex << std::uppercase
    186             << static_cast<int>(static_cast<UnsignedChar>(c));
    187         os->flags(flags);
    188         return kHexEscape;
    189       }
    190   }
    191   return kSpecialEscape;
    192 }
    193 
    194 // Prints a wchar_t c as if it's part of a string literal, escaping it when
    195 // necessary; returns how c was formatted.
    196 static CharFormat PrintAsStringLiteralTo(wchar_t c, ostream* os) {
    197   switch (c) {
    198     case L'\'':
    199       *os << "'";
    200       return kAsIs;
    201     case L'"':
    202       *os << "\\\"";
    203       return kSpecialEscape;
    204     default:
    205       return PrintAsCharLiteralTo<wchar_t>(c, os);
    206   }
    207 }
    208 
    209 // Prints a char c as if it's part of a string literal, escaping it when
    210 // necessary; returns how c was formatted.
    211 static CharFormat PrintAsStringLiteralTo(char c, ostream* os) {
    212   return PrintAsStringLiteralTo(
    213       static_cast<wchar_t>(static_cast<unsigned char>(c)), os);
    214 }
    215 
    216 // Prints a wide or narrow character c and its code.  '\0' is printed
    217 // as "'\\0'", other unprintable characters are also properly escaped
    218 // using the standard C++ escape sequence.  The template argument
    219 // UnsignedChar is the unsigned version of Char, which is the type of c.
    220 template <typename UnsignedChar, typename Char>
    221 void PrintCharAndCodeTo(Char c, ostream* os) {
    222   // First, print c as a literal in the most readable form we can find.
    223   *os << ((sizeof(c) > 1) ? "L'" : "'");
    224   const CharFormat format = PrintAsCharLiteralTo<UnsignedChar>(c, os);
    225   *os << "'";
    226 
    227   // To aid user debugging, we also print c's code in decimal, unless
    228   // it's 0 (in which case c was printed as '\\0', making the code
    229   // obvious).
    230   if (c == 0)
    231     return;
    232   *os << " (" << static_cast<int>(c);
    233 
    234   // For more convenience, we print c's code again in hexadecimal,
    235   // unless c was already printed in the form '\x##' or the code is in
    236   // [1, 9].
    237   if (format == kHexEscape || (1 <= c && c <= 9)) {
    238     // Do nothing.
    239   } else {
    240     *os << ", 0x" << String::FormatHexInt(static_cast<int>(c));
    241   }
    242   *os << ")";
    243 }
    244 
    245 void PrintTo(unsigned char c, ::std::ostream* os) {
    246   PrintCharAndCodeTo<unsigned char>(c, os);
    247 }
    248 void PrintTo(signed char c, ::std::ostream* os) {
    249   PrintCharAndCodeTo<unsigned char>(c, os);
    250 }
    251 
    252 // Prints a wchar_t as a symbol if it is printable or as its internal
    253 // code otherwise and also as its code.  L'\0' is printed as "L'\\0'".
    254 void PrintTo(wchar_t wc, ostream* os) {
    255   PrintCharAndCodeTo<wchar_t>(wc, os);
    256 }
    257 
    258 // Prints the given array of characters to the ostream.  CharType must be either
    259 // char or wchar_t.
    260 // The array starts at begin, the length is len, it may include '\0' characters
    261 // and may not be NUL-terminated.
    262 template <typename CharType>
    263 GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_
    264 GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_
    265 GTEST_ATTRIBUTE_NO_SANITIZE_HWADDRESS_
    266 GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_
    267 static CharFormat PrintCharsAsStringTo(
    268     const CharType* begin, size_t len, ostream* os) {
    269   const char* const kQuoteBegin = sizeof(CharType) == 1 ? "\"" : "L\"";
    270   *os << kQuoteBegin;
    271   bool is_previous_hex = false;
    272   CharFormat print_format = kAsIs;
    273   for (size_t index = 0; index < len; ++index) {
    274     const CharType cur = begin[index];
    275     if (is_previous_hex && IsXDigit(cur)) {
    276       // Previous character is of '\x..' form and this character can be
    277       // interpreted as another hexadecimal digit in its number. Break string to
    278       // disambiguate.
    279       *os << "\" " << kQuoteBegin;
    280     }
    281     is_previous_hex = PrintAsStringLiteralTo(cur, os) == kHexEscape;
    282     // Remember if any characters required hex escaping.
    283     if (is_previous_hex) {
    284       print_format = kHexEscape;
    285     }
    286   }
    287   *os << "\"";
    288   return print_format;
    289 }
    290 
    291 // Prints a (const) char/wchar_t array of 'len' elements, starting at address
    292 // 'begin'.  CharType must be either char or wchar_t.
    293 template <typename CharType>
    294 GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_
    295 GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_
    296 GTEST_ATTRIBUTE_NO_SANITIZE_HWADDRESS_
    297 GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_
    298 static void UniversalPrintCharArray(
    299     const CharType* begin, size_t len, ostream* os) {
    300   // The code
    301   //   const char kFoo[] = "foo";
    302   // generates an array of 4, not 3, elements, with the last one being '\0'.
    303   //
    304   // Therefore when printing a char array, we don't print the last element if
    305   // it's '\0', such that the output matches the string literal as it's
    306   // written in the source code.
    307   if (len > 0 && begin[len - 1] == '\0') {
    308     PrintCharsAsStringTo(begin, len - 1, os);
    309     return;
    310   }
    311 
    312   // If, however, the last element in the array is not '\0', e.g.
    313   //    const char kFoo[] = { 'f', 'o', 'o' };
    314   // we must print the entire array.  We also print a message to indicate
    315   // that the array is not NUL-terminated.
    316   PrintCharsAsStringTo(begin, len, os);
    317   *os << " (no terminating NUL)";
    318 }
    319 
    320 // Prints a (const) char array of 'len' elements, starting at address 'begin'.
    321 void UniversalPrintArray(const char* begin, size_t len, ostream* os) {
    322   UniversalPrintCharArray(begin, len, os);
    323 }
    324 
    325 // Prints a (const) wchar_t array of 'len' elements, starting at address
    326 // 'begin'.
    327 void UniversalPrintArray(const wchar_t* begin, size_t len, ostream* os) {
    328   UniversalPrintCharArray(begin, len, os);
    329 }
    330 
    331 // Prints the given C string to the ostream.
    332 void PrintTo(const char* s, ostream* os) {
    333   if (s == nullptr) {
    334     *os << "NULL";
    335   } else {
    336     *os << ImplicitCast_<const void*>(s) << " pointing to ";
    337     PrintCharsAsStringTo(s, strlen(s), os);
    338   }
    339 }
    340 
    341 // MSVC compiler can be configured to define whar_t as a typedef
    342 // of unsigned short. Defining an overload for const wchar_t* in that case
    343 // would cause pointers to unsigned shorts be printed as wide strings,
    344 // possibly accessing more memory than intended and causing invalid
    345 // memory accesses. MSVC defines _NATIVE_WCHAR_T_DEFINED symbol when
    346 // wchar_t is implemented as a native type.
    347 #if !defined(_MSC_VER) || defined(_NATIVE_WCHAR_T_DEFINED)
    348 // Prints the given wide C string to the ostream.
    349 void PrintTo(const wchar_t* s, ostream* os) {
    350   if (s == nullptr) {
    351     *os << "NULL";
    352   } else {
    353     *os << ImplicitCast_<const void*>(s) << " pointing to ";
    354     PrintCharsAsStringTo(s, wcslen(s), os);
    355   }
    356 }
    357 #endif  // wchar_t is native
    358 
    359 namespace {
    360 
    361 bool ContainsUnprintableControlCodes(const char* str, size_t length) {
    362   const unsigned char *s = reinterpret_cast<const unsigned char *>(str);
    363 
    364   for (size_t i = 0; i < length; i++) {
    365     unsigned char ch = *s++;
    366     if (std::iscntrl(ch)) {
    367         switch (ch) {
    368         case '\t':
    369         case '\n':
    370         case '\r':
    371           break;
    372         default:
    373           return true;
    374         }
    375       }
    376   }
    377   return false;
    378 }
    379 
    380 bool IsUTF8TrailByte(unsigned char t) { return 0x80 <= t && t<= 0xbf; }
    381 
    382 bool IsValidUTF8(const char* str, size_t length) {
    383   const unsigned char *s = reinterpret_cast<const unsigned char *>(str);
    384 
    385   for (size_t i = 0; i < length;) {
    386     unsigned char lead = s[i++];
    387 
    388     if (lead <= 0x7f) {
    389       continue;  // single-byte character (ASCII) 0..7F
    390     }
    391     if (lead < 0xc2) {
    392       return false;  // trail byte or non-shortest form
    393     } else if (lead <= 0xdf && (i + 1) <= length && IsUTF8TrailByte(s[i])) {
    394       ++i;  // 2-byte character
    395     } else if (0xe0 <= lead && lead <= 0xef && (i + 2) <= length &&
    396                IsUTF8TrailByte(s[i]) &&
    397                IsUTF8TrailByte(s[i + 1]) &&
    398                // check for non-shortest form and surrogate
    399                (lead != 0xe0 || s[i] >= 0xa0) &&
    400                (lead != 0xed || s[i] < 0xa0)) {
    401       i += 2;  // 3-byte character
    402     } else if (0xf0 <= lead && lead <= 0xf4 && (i + 3) <= length &&
    403                IsUTF8TrailByte(s[i]) &&
    404                IsUTF8TrailByte(s[i + 1]) &&
    405                IsUTF8TrailByte(s[i + 2]) &&
    406                // check for non-shortest form
    407                (lead != 0xf0 || s[i] >= 0x90) &&
    408                (lead != 0xf4 || s[i] < 0x90)) {
    409       i += 3;  // 4-byte character
    410     } else {
    411       return false;
    412     }
    413   }
    414   return true;
    415 }
    416 
    417 void ConditionalPrintAsText(const char* str, size_t length, ostream* os) {
    418   if (!ContainsUnprintableControlCodes(str, length) &&
    419       IsValidUTF8(str, length)) {
    420     *os << "\n    As Text: \"" << str << "\"";
    421   }
    422 }
    423 
    424 }  // anonymous namespace
    425 
    426 void PrintStringTo(const ::std::string& s, ostream* os) {
    427   if (PrintCharsAsStringTo(s.data(), s.size(), os) == kHexEscape) {
    428     if (GTEST_FLAG(print_utf8)) {
    429       ConditionalPrintAsText(s.data(), s.size(), os);
    430     }
    431   }
    432 }
    433 
    434 #if GTEST_HAS_STD_WSTRING
    435 void PrintWideStringTo(const ::std::wstring& s, ostream* os) {
    436   PrintCharsAsStringTo(s.data(), s.size(), os);
    437 }
    438 #endif  // GTEST_HAS_STD_WSTRING
    439 
    440 }  // namespace internal
    441 
    442 }  // namespace testing