libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf16/79980.cc

   1 // Copyright (C) 2017-2025 Free Software Foundation, Inc.
   2 //
   3 // This file is part of the GNU ISO C++ Library.  This library is free
   4 // software; you can redistribute it and/or modify it under the
   5 // terms of the GNU General Public License as published by the
   6 // Free Software Foundation; either version 3, or (at your option)
   7 // any later version.
   8
   9 // This library is distributed in the hope that it will be useful,
  10 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 // GNU General Public License for more details.
  13
  14 // You should have received a copy of the GNU General Public License along
  15 // with this library; see the file COPYING3.  If not see
  16 // <http://www.gnu.org/licenses/>.
  17
  18 // { dg-do run { target c++11 } }
  19 // { dg-additional-options "-Wno-deprecated-declarations" { target c++17 } }
  20
  21 #include <locale>
  22 #include <codecvt>
  23 #include <testsuite_hooks.h>
  24
  25 // PR libstdc++/79980
  26
  27 constexpr std::codecvt_mode mode(std::codecvt_mode m)
  28 { return static_cast<std::codecvt_mode>(m | std::consume_header); }
  29
  30 template<typename WCh, unsigned long Max = 0x10FFFF,
  31          std::codecvt_mode Mode = std::consume_header>
  32   using Conv
  33     = std::wstring_convert<std::codecvt_utf16<WCh, Max, mode(Mode)>, WCh>;
  34
  35 void
  36 test01()
  37 {
  38   const char src[] = "\xFE\xFF\xAB\xCD";
  39   Conv<char16_t> conv;
  40   auto dst = conv.from_bytes(src, src+4);
  41   VERIFY( dst[0] == 0xabcd );
  42 }
  43
  44 void
  45 test02()
  46 {
  47   const char src[] = "\xFF\xFE\xAB\xCD";
  48   Conv<char16_t> conv;
  49   auto dst = conv.from_bytes(src, src+4);
  50   VERIFY( dst[0] == 0xcdab );
  51 }
  52
  53 void
  54 test03()
  55 {
  56   const char src[] = "\xFE\xFF\xAB\xCD";
  57   Conv<char16_t, 0x10FFFF, std::little_endian> conv;
  58   auto dst = conv.from_bytes(src, src+4);
  59   VERIFY( dst[0] == 0xabcd );
  60 }
  61
  62 void
  63 test04()
  64 {
  65   const char src[] = "\xFF\xFE\xAB\xCD";
  66   Conv<char16_t, 0x10FFFF, std::little_endian> conv;
  67   auto dst = conv.from_bytes(src, src+4);
  68   VERIFY( dst[0] == 0xcdab );
  69 }
  70
  71 void
  72 test05()
  73 {
  74   const char src[] = "\0\x61\xAB\xCD"; // character greater than 0x00FF
  75   Conv<char16_t, 0xFF> conv("to_bytes failed", u"from_bytes failed");
  76   std::u16string result = conv.from_bytes(src, src+4);
  77   VERIFY( result == u"from_bytes failed" );
  78   VERIFY( conv.converted() == 2 );
  79 }
  80
  81 void
  82 test06()
  83 {
  84   const char src[] = "\0\x61\xAB\xCD";
  85   Conv<char16_t> conv("to_bytes failed", u"from_bytes failed");
  86   std::u16string result = conv.from_bytes(src, src+3); // incomplete character
  87   VERIFY( result == u"\u0061" );
  88   VERIFY( conv.converted() == 2 );
  89 }
  90
  91 void
  92 test07()
  93 {
  94   Conv<char16_t> conv("to_bytes failed", u"from_bytes failed");
  95   // ucs2 to utf-16 conversion should fail on invalid ucs2 input:
  96   std::u16string utf16 = u"1234\U00001111\U0001ffff";
  97   auto out = conv.to_bytes(utf16);
  98   VERIFY( out == "to_bytes failed" );
  99   VERIFY( conv.converted() == 5 );
 100
 101   // And should also fail on incomplete surrogate pair (not return partial):
 102   out = conv.to_bytes(utf16.substr(0, utf16.size()-1));
 103   VERIFY( out == "to_bytes failed" );
 104   VERIFY( conv.converted() == 5 );
 105 }
 106
 107 void
 108 test08()
 109 {
 110   // Read/write UTF-16 code units from data not correctly aligned for char16_t
 111   Conv<char16_t, 0x10FFFF, std::generate_header> conv;
 112   const char src[] = "-\xFE\xFF\0\x61\xAB\xCD";
 113   auto out = conv.from_bytes(src + 1, src + 7);
 114   VERIFY( out[0] == 0x0061 );
 115   VERIFY( out[1] == 0xabcd );
 116   auto bytes = conv.to_bytes(out);
 117   VERIFY( bytes == std::string(src + 1, 6) );
 118 }
 119
 120 void
 121 test09()
 122 {
 123   // Read/write UTF-16 code units from data not correctly aligned for char16_t
 124   Conv<char32_t, 0x10FFFF, std::generate_header> conv;
 125   const char src[] = "-\xFE\xFF\xD8\x08\xDF\x45";
 126   auto out = conv.from_bytes(src + 1, src + 7);
 127   VERIFY( out == U"\U00012345" );
 128   auto bytes = conv.to_bytes(out);
 129   VERIFY( bytes == std::string(src + 1, 6) );
 130 }
 131
 132 int main()
 133 {
 134   test01();
 135   test02();
 136   test03();
 137   test04();
 138   test05();
 139   test06();
 140   test07();
 141   test08();
 142   test09();
 143 }