1 // Copyright (C) 2017-2021 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
20 #include <locale>
21 #include <codecvt>
22 #include <testsuite_hooks.h>
23
24 // PR libstdc++/79980
25
mode(std::codecvt_mode m)26 constexpr std::codecvt_mode mode(std::codecvt_mode m)
27 { return static_cast<std::codecvt_mode>(m | std::consume_header); }
28
29 template<typename WCh, unsigned long Max = 0x10FFFF,
30 std::codecvt_mode Mode = std::consume_header>
31 using Conv
32 = std::wstring_convert<std::codecvt_utf16<WCh, Max, mode(Mode)>, WCh>;
33
34 void
test01()35 test01()
36 {
37 const char src[] = "\xFE\xFF\xAB\xCD";
38 Conv<char16_t> conv;
39 auto dst = conv.from_bytes(src, src+4);
40 VERIFY( dst[0] == 0xabcd );
41 }
42
43 void
test02()44 test02()
45 {
46 const char src[] = "\xFF\xFE\xAB\xCD";
47 Conv<char16_t> conv;
48 auto dst = conv.from_bytes(src, src+4);
49 VERIFY( dst[0] == 0xcdab );
50 }
51
52 void
test03()53 test03()
54 {
55 const char src[] = "\xFE\xFF\xAB\xCD";
56 Conv<char16_t, 0x10FFFF, std::little_endian> conv;
57 auto dst = conv.from_bytes(src, src+4);
58 VERIFY( dst[0] == 0xabcd );
59 }
60
61 void
test04()62 test04()
63 {
64 const char src[] = "\xFF\xFE\xAB\xCD";
65 Conv<char16_t, 0x10FFFF, std::little_endian> conv;
66 auto dst = conv.from_bytes(src, src+4);
67 VERIFY( dst[0] == 0xcdab );
68 }
69
70 void
test05()71 test05()
72 {
73 const char src[] = "\0\x61\xAB\xCD"; // character greater than 0x00FF
74 Conv<char16_t, 0xFF> conv("to_bytes failed", u"from_bytes failed");
75 std::u16string result = conv.from_bytes(src, src+4);
76 VERIFY( result == u"from_bytes failed" );
77 VERIFY( conv.converted() == 2 );
78 }
79
80 void
test06()81 test06()
82 {
83 const char src[] = "\0\x61\xAB\xCD";
84 Conv<char16_t> conv("to_bytes failed", u"from_bytes failed");
85 std::u16string result = conv.from_bytes(src, src+3); // incomplete character
86 VERIFY( result == u"from_bytes failed" );
87 VERIFY( conv.converted() == 2 );
88 }
89
90 void
test07()91 test07()
92 {
93 Conv<char16_t> conv("to_bytes failed", u"from_bytes failed");
94 // ucs2 to utf-16 conversion should fail on invalid ucs2 input:
95 std::u16string utf16 = u"1234\U00001111\U0001ffff";
96 auto out = conv.to_bytes(utf16);
97 VERIFY( out == "to_bytes failed" );
98 VERIFY( conv.converted() == 5 );
99
100 // And should also fail on incomplete surrogate pair (not return partial):
101 out = conv.to_bytes(utf16.substr(0, utf16.size()-1));
102 VERIFY( out == "to_bytes failed" );
103 VERIFY( conv.converted() == 5 );
104 }
105
106 void
test08()107 test08()
108 {
109 // Read/write UTF-16 code units from data not correctly aligned for char16_t
110 Conv<char16_t, 0x10FFFF, std::generate_header> conv;
111 const char src[] = "-\xFE\xFF\0\x61\xAB\xCD";
112 auto out = conv.from_bytes(src + 1, src + 7);
113 VERIFY( out[0] == 0x0061 );
114 VERIFY( out[1] == 0xabcd );
115 auto bytes = conv.to_bytes(out);
116 VERIFY( bytes == std::string(src + 1, 6) );
117 }
118
119 void
test09()120 test09()
121 {
122 // Read/write UTF-16 code units from data not correctly aligned for char16_t
123 Conv<char32_t, 0x10FFFF, std::generate_header> conv;
124 const char src[] = "-\xFE\xFF\xD8\x08\xDF\x45";
125 auto out = conv.from_bytes(src + 1, src + 7);
126 VERIFY( out == U"\U00012345" );
127 auto bytes = conv.to_bytes(out);
128 VERIFY( bytes == std::string(src + 1, 6) );
129 }
130
main()131 int main()
132 {
133 test01();
134 test02();
135 test03();
136 test04();
137 test05();
138 test06();
139 test07();
140 test08();
141 test09();
142 }
143