xref: /llvm-project/libcxx/test/std/utilities/format/format.functions/escaped_output.unicode.pass.cpp (revision 3b00bae9a9f79bade0f4c67379d8f32d1d0271e3) 
1 //===----------------------------------------------------------------------===//
2 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
3 // See https://llvm.org/LICENSE.txt for license information.
4 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
5 //
6 //===----------------------------------------------------------------------===//
7 
8 // UNSUPPORTED: c++03, c++11, c++14, c++17, c++20
9 // UNSUPPORTED: libcpp-has-no-incomplete-format
10 
11 // This version runs the test when the platform has Unicode support.
12 // UNSUPPORTED: libcpp-has-no-unicode
13 
14 // TODO FMT Investigate Windows and AIX issues.
15 // UNSUPPORTED: msvc, target={{.+}}-windows-gnu
16 // UNSUPPORTED: LIBCXX-AIX-FIXME
17 
18 // TODO FMT Fix this test using GCC, it currently crashes.
19 // UNSUPPORTED: gcc-12
20 
21 // TODO FMT This test should not require std::to_chars(floating-point)
22 // This test requires std::to_chars(floating-point), which is in the dylib
23 // XFAIL: use_system_cxx_lib && target={{.+}}-apple-macosx{{10.9|10.10|10.11|10.12|10.13|10.14|10.15|11.0}}
24 
25 // <format>
26 
27 // This test the debug string type for the formatter specializations for char
28 // and string types. This tests Unicode strings.
29 
30 #include <format>
31 
32 #include <cassert>
33 #include <concepts>
34 #include <iterator>
35 #include <list>
36 #include <vector>
37 
38 #include "test_macros.h"
39 #include "make_string.h"
40 #include "test_format_string.h"
41 #include "assert_macros.h"
42 #include "concat_macros.h"
43 
44 #ifndef TEST_HAS_NO_LOCALIZATION
45 #  include <iostream>
46 #endif
47 
48 #define SV(S) MAKE_STRING_VIEW(CharT, S)
49 
50 auto test_format = []<class CharT, class... Args>(
51                        std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
52   {
53     std::basic_string<CharT> out = std::format(fmt, std::forward<Args>(args)...);
54     TEST_REQUIRE(out == expected,
55                  TEST_WRITE_CONCATENATED(
56                      "\nFormat string   ", fmt.get(), "\nExpected output ", expected, "\nActual output   ", out, '\n'));
57   }
58 #ifndef TEST_HAS_NO_LOCALIZATION
59   {
60     std::basic_string<CharT> out = std::format(std::locale(), fmt, std::forward<Args>(args)...);
61     assert(out == expected);
62   }
63 #endif // TEST_HAS_NO_LOCALIZATION
64 };
65 
66 auto test_format_to =
67     []<class CharT, class... Args>(
68         std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
69       {
70         std::basic_string<CharT> out(expected.size(), CharT(' '));
71         auto it = std::format_to(out.begin(), fmt, std::forward<Args>(args)...);
72         assert(it == out.end());
73         assert(out == expected);
74       }
75 #ifndef TEST_HAS_NO_LOCALIZATION
76       {
77         std::basic_string<CharT> out(expected.size(), CharT(' '));
78         auto it = std::format_to(out.begin(), std::locale(), fmt, std::forward<Args>(args)...);
79         assert(it == out.end());
80         assert(out == expected);
81       }
82 #endif // TEST_HAS_NO_LOCALIZATION
83       {
84         std::list<CharT> out;
85         std::format_to(std::back_inserter(out), fmt, std::forward<Args>(args)...);
86         assert(std::equal(out.begin(), out.end(), expected.begin(), expected.end()));
87       }
88       {
89         std::vector<CharT> out;
90         std::format_to(std::back_inserter(out), fmt, std::forward<Args>(args)...);
91         assert(std::equal(out.begin(), out.end(), expected.begin(), expected.end()));
92       }
93       {
94         assert(expected.size() < 4096 && "Update the size of the buffer.");
95         CharT out[4096];
96         CharT* it = std::format_to(out, fmt, std::forward<Args>(args)...);
97         assert(std::distance(out, it) == int(expected.size()));
98         // Convert to std::string since output contains '\0' for boolean tests.
99         assert(std::basic_string<CharT>(out, it) == expected);
100       }
101     };
102 
103 auto test_formatted_size =
104     []<class CharT, class... Args>(
105         std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
106       {
107         std::size_t size = std::formatted_size(fmt, std::forward<Args>(args)...);
108         assert(size == expected.size());
109       }
110 #ifndef TEST_HAS_NO_LOCALIZATION
111       {
112         std::size_t size = std::formatted_size(std::locale(), fmt, std::forward<Args>(args)...);
113         assert(size == expected.size());
114       }
115 #endif // TEST_HAS_NO_LOCALIZATION
116     };
117 
118 auto test_format_to_n =
119     []<class CharT, class... Args>(
120         std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
121       {
122         std::size_t n = expected.size();
123         std::basic_string<CharT> out(n, CharT(' '));
124         std::format_to_n_result result = std::format_to_n(out.begin(), n, fmt, std::forward<Args>(args)...);
125         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
126         assert(result.out == out.end());
127         assert(out == expected);
128       }
129 #ifndef TEST_HAS_NO_LOCALIZATION
130       {
131         std::size_t n = expected.size();
132         std::basic_string<CharT> out(n, CharT(' '));
133         std::format_to_n_result result =
134             std::format_to_n(out.begin(), n, std::locale(), fmt, std::forward<Args>(args)...);
135         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
136         assert(result.out == out.end());
137         assert(out == expected);
138       }
139 #endif // TEST_HAS_NO_LOCALIZATION
140       {
141         std::ptrdiff_t n = 0;
142         std::basic_string<CharT> out;
143         std::format_to_n_result result = std::format_to_n(out.begin(), n, fmt, std::forward<Args>(args)...);
144         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
145         assert(result.out == out.end());
146         assert(out.empty());
147       }
148       {
149         std::ptrdiff_t n = expected.size() / 2;
150         std::basic_string<CharT> out(n, CharT(' '));
151         std::format_to_n_result result = std::format_to_n(out.begin(), n, fmt, std::forward<Args>(args)...);
152         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
153         assert(result.out == out.end());
154         assert(out == expected.substr(0, n));
155       }
156     };
157 
158 template <class CharT>
159 void test_char() {
160   // *** P2286 examples ***
161   test_format(SV("['\\'', '\"']"), SV("[{:?}, {:?}]"), CharT('\''), CharT('"'));
162 
163   // *** Specical cases ***
164   test_format(SV("'\\t'"), SV("{:?}"), CharT('\t'));
165   test_format(SV("'\\n'"), SV("{:?}"), CharT('\n'));
166   test_format(SV("'\\r'"), SV("{:?}"), CharT('\r'));
167   test_format(SV("'\\\\'"), SV("{:?}"), CharT('\\'));
168 
169   test_format(SV("'\\\''"), SV("{:?}"), CharT('\''));
170   test_format(SV("'\"'"), SV("{:?}"), CharT('"')); // only special for string
171 
172   test_format(SV("' '"), SV("{:?}"), CharT(' '));
173 
174   // *** Printable ***
175   test_format(SV("'a'"), SV("{:?}"), CharT('a'));
176   test_format(SV("'b'"), SV("{:?}"), CharT('b'));
177   test_format(SV("'c'"), SV("{:?}"), CharT('c'));
178 
179   // *** Non-printable ***
180 
181   // Control
182   test_format(SV("'\\u{0}'"), SV("{:?}"), CharT('\0'));
183   test_format(SV("'\\u{1f}'"), SV("{:?}"), CharT('\x1f'));
184 
185   // Ill-formed
186   if constexpr (sizeof(CharT) == 1)
187     test_format(SV("'\\x{80}'"), SV("{:?}"), CharT('\x80'));
188 
189 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
190   if constexpr (sizeof(CharT) > 1) {
191     using V = std::basic_string_view<CharT>;
192 
193     // Unicode fitting in a 16-bit wchar_t
194 
195     // *** Non-printable ***
196 
197     // Space_Separator
198     test_format(V{L"'\\u{a0}'"}, L"{:?}", L'\xa0');     // NO-BREAK SPACE
199     test_format(V{L"'\\u{3000}'"}, L"{:?}", L'\x3000'); // IDEOGRAPHIC SPACE
200 
201     // Line_Separator
202     test_format(V{L"'\\u{2028}'"}, L"{:?}", L'\x2028'); // LINE SEPARATOR
203 
204     // Paragraph_Separator
205     test_format(V{L"'\\u{2029}'"}, L"{:?}", L'\x2029'); // PARAGRAPH SEPARATOR
206 
207     // Format
208     test_format(V{L"'\\u{ad}'"}, L"{:?}", L'\xad');     // SOFT HYPHEN
209     test_format(V{L"'\\u{600}'"}, L"{:?}", L'\x600');   // ARABIC NUMBER SIGN
210     test_format(V{L"'\\u{feff}'"}, L"{:?}", L'\xfeff'); // ZERO WIDTH NO-BREAK SPACE
211 
212     // Incomplete surrogate pair in UTF-16
213     test_format(V{L"'\\x{d800}'"}, L"{:?}", L'\xd800'); // <surrogate-D800>
214     test_format(V{L"'\\x{dfff}'"}, L"{:?}", L'\xdfff'); // <surrogate-DFFF>
215 
216     // Private_Use
217     test_format(V{L"'\\u{e000}'"}, L"{:?}", L'\xe000'); // <private-use-E000>
218     test_format(V{L"'\\u{f8ff}'"}, L"{:?}", L'\xf8ff'); // <private-use-F8FF>
219 
220     // Unassigned
221     test_format(V{L"'\\u{378}'"}, L"{:?}", L'\x378');   // <reserved-0378>
222     test_format(V{L"'\\u{1774}'"}, L"{:?}", L'\x1774'); // <reserved-1774>
223     test_format(V{L"'\\u{ffff}'"}, L"{:?}", L'\xffff'); // <noncharacter-FFFF>
224 
225     // Grapheme Extended
226     test_format(V{L"'\\u{300}'"}, L"{:?}", L'\x300');   // COMBINING GRAVE ACCENT
227     test_format(V{L"'\\u{fe20}'"}, L"{:?}", L'\xfe20'); // VARIATION SELECTOR-1
228   }
229 #  ifndef TEST_SHORT_WCHAR
230   if constexpr (sizeof(CharT) > 2) {
231     static_assert(sizeof(CharT) == 4, "add support for unexpected size");
232     // Unicode fitting in a 32-bit wchar_t
233 
234     constexpr wchar_t x  = 0x1ffff;
235     constexpr std::uint32_t y = 0x1ffff;
236     static_assert(x == y);
237 
238     using V = std::basic_string_view<CharT>;
239 
240     // *** Non-printable ***
241     // Format
242     test_format(V{L"'\\u{110bd}'"}, L"{:?}", L'\x110bd'); // KAITHI NUMBER SIGN
243     test_format(V{L"'\\u{e007f}'"}, L"{:?}", L'\xe007f'); // CANCEL TAG
244 
245     // Private_Use
246     test_format(V{L"'\\u{f0000}'"}, L"{:?}", L'\xf0000'); // <private-use-F0000>
247     test_format(V{L"'\\u{ffffd}'"}, L"{:?}", L'\xffffd'); // <private-use-FFFFD>
248 
249     test_format(V{L"'\\u{100000}'"}, L"{:?}", L'\x100000'); // <private-use-100000>
250     test_format(V{L"'\\u{10fffd}'"}, L"{:?}", L'\x10fffd'); // <private-use-10FFFD>
251 
252     // Unassigned
253     test_format(V{L"'\\u{1000c}'"}, L"{:?}", L'\x1000c');   // <reserved-1000c>
254     test_format(V{L"'\\u{fffff}'"}, L"{:?}", L'\xfffff');   // <noncharacter-FFFFF>
255     test_format(V{L"'\\u{10fffe}'"}, L"{:?}", L'\x10fffe'); // <noncharacter-10FFFE>
256 
257     // Grapheme Extended
258     test_format(V{L"'\\u{101fd}'"}, L"{:?}", L'\x101fd'); // COMBINING OLD PERMIC LETTER AN
259     test_format(V{L"'\\u{e0100}'"}, L"{:?}", L'\xe0100'); // VARIATION SELECTOR-17
260 
261     // Ill-formed
262     test_format(V{L"'\\x{110000}'"}, L"{:?}", L'\x110000');
263     test_format(V{L"'\\x{ffffffff}'"}, L"{:?}", L'\xffffffff');
264   }
265 #  endif // TEST_SHORT_WCHAR
266 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
267 }
268 
269 template <class CharT>
270 void test_string() {
271   // *** P2286 examples ***
272   test_format(SV("[h\tllo]"), SV("[{}]"), SV("h\tllo"));
273   test_format(SV(R"(["h\tllo"])"), SV("[{:?}]"), SV("h\tllo"));
274   test_format(SV(R"(["Спасибо, Виктор ♥!"])"), SV("[{:?}]"), SV("Спасибо, Виктор ♥!"));
275 
276   test_format(SV(R"(["\u{0} \n \t \u{2} \u{1b}"])"), SV("[{:?}]"), SV("\0 \n \t \x02 \x1b"));
277 
278   if constexpr (sizeof(CharT) == 1) {
279     // Ill-formend UTF-8
280     test_format(SV(R"(["\x{c3}"])"), SV("[{:?}]"), "\xc3");
281     test_format(SV(R"(["\x{c3}("])"), SV("[{:?}]"), "\xc3\x28");
282 
283     /* U+0000..U+0007F 1 code unit range, encoded in 2 code units. */
284     test_format(SV(R"(["\x{c0}\x{80}"])"), SV("[{:?}]"), "\xc0\x80"); // U+0000
285     test_format(SV(R"(["\x{c1}\x{bf}"])"), SV("[{:?}]"), "\xc1\xbf"); // U+007F
286     test_format(SV(R"(["\u{80}"])"), SV("[{:?}]"), "\xc2\x80");       // U+0080 first valid (General_Category=Control)
287 
288     /* U+0000..U+07FFF 1 and 2 code unit range, encoded in 3 code units. */
289     test_format(SV(R"(["\x{e0}\x{80}\x{80}"])"), SV("[{:?}]"), "\xe0\x80\x80"); // U+0000
290     test_format(SV(R"(["\x{e0}\x{81}\x{bf}"])"), SV("[{:?}]"), "\xe0\x81\xbf"); // U+007F
291     test_format(SV(R"(["\x{e0}\x{82}\x{80}"])"), SV("[{:?}]"), "\xe0\x82\x80"); // U+0080
292     test_format(SV(R"(["\x{e0}\x{9f}\x{bf}"])"), SV("[{:?}]"), "\xe0\x9f\xbf"); // U+07FF
293     test_format(SV("[\"\u0800\"]"), SV("[{:?}]"), "\xe0\xa0\x80");              // U+0800 first valid
294 
295 #if 0
296 	// This code point is in the Hangul Jamo Extended-B block and at the time of writing
297 	// it's unassigned. When it comes defined, this branch might become true.
298     test_format(SV("[\"\ud7ff\"]"), SV("[{:?}]"), "\xed\x9f\xbf");              // U+D7FF last valid
299 #else
300     /* U+D800..D+DFFFF surrogate range */
301     test_format(SV(R"(["\u{d7ff}"])"), SV("[{:?}]"), "\xed\x9f\xbf");           // U+D7FF last valid
302 #endif
303     test_format(SV(R"(["\x{ed}\x{a0}\x{80}"])"), SV("[{:?}]"), "\xed\xa0\x80"); // U+D800
304     test_format(SV(R"(["\x{ed}\x{af}\x{bf}"])"), SV("[{:?}]"), "\xed\xaf\xbf"); // U+DBFF
305     test_format(SV(R"(["\x{ed}\x{bf}\x{80}"])"), SV("[{:?}]"), "\xed\xbf\x80"); // U+DC00
306     test_format(SV(R"(["\x{ed}\x{bf}\x{bf}"])"), SV("[{:?}]"), "\xed\xbf\xbf"); // U+DFFF
307     test_format(SV(R"(["\u{e000}"])"), SV("[{:?}]"), "\xee\x80\x80");           // U+E000 first valid
308                                                                                 // (in the Private Use Area block)
309 
310     /* U+0000..U+FFFF 1, 2, and 3 code unit range */
311     test_format(SV(R"(["\x{f0}\x{80}\x{80}\x{80}"])"), SV("[{:?}]"), "\xf0\x80\x80\x80"); // U+0000
312     test_format(SV(R"(["\x{f0}\x{80}\x{81}\x{bf}"])"), SV("[{:?}]"), "\xf0\x80\x81\xbf"); // U+007F
313     test_format(SV(R"(["\x{f0}\x{80}\x{82}\x{80}"])"), SV("[{:?}]"), "\xf0\x80\x82\x80"); // U+0080
314     test_format(SV(R"(["\x{f0}\x{80}\x{9f}\x{bf}"])"), SV("[{:?}]"), "\xf0\x80\x9f\xbf"); // U+07FF
315     test_format(SV(R"(["\x{f0}\x{80}\x{a0}\x{80}"])"), SV("[{:?}]"), "\xf0\x80\xa0\x80"); // U+0800
316     test_format(SV(R"(["\x{f0}\x{8f}\x{bf}\x{bf}"])"), SV("[{:?}]"), "\xf0\x8f\xbf\xbf"); // U+FFFF
317     test_format(SV("[\"\U00010000\"]"), SV("[{:?}]"), "\xf0\x90\x80\x80");                // U+10000 first valid
318 
319     /* U+10FFFF..U+1FFFFF invalid range */
320     test_format(SV(R"(["\u{10ffff}"])"), SV("[{:?}]"), "\xf4\x8f\xbf\xbf"); // U+10FFFF last valid
321                                                                             // (in Supplementary Private Use Area-B)
322     test_format(SV(R"(["\x{f4}\x{90}\x{80}\x{80}"])"), SV("[{:?}]"), "\xf4\x90\x80\x80"); // U+110000
323     test_format(SV(R"(["\x{f4}\x{bf}\x{bf}\x{bf}"])"), SV("[{:?}]"), "\xf4\xbf\xbf\xbf"); // U+11FFFF
324   } else {
325     // Valid UTF-16 and UTF-32
326     test_format(SV("[\"\u00c3\"]"), SV("[{:?}]"), L"\xc3"); // LATIN CAPITAL LETTER A WITH TILDE
327     test_format(SV("[\"\u00c3(\"]"), SV("[{:?}]"), L"\xc3\x28");
328   }
329 
330   test_format(SV(R"(["����\u{200d}♂\u{fe0f}"])"), SV("[{:?}]"), SV("����‍♂️"));
331 
332   // *** Specical cases ***
333   test_format(SV(R"("\t\n\r\\'\" ")"), SV("{:?}"), SV("\t\n\r\\'\" "));
334 
335   // *** Printable ***
336   test_format(SV(R"("abcdefg")"), SV("{:?}"), SV("abcdefg"));
337 
338   // *** Non-printable ***
339 
340   // Control
341   test_format(SV(R"("\u{0}\u{1f}")"), SV("{:?}"), SV("\0\x1f"));
342 
343   // Ill-formed
344   if constexpr (sizeof(CharT) == 1)
345     test_format(SV(R"("\x{80}")"), SV("{:?}"), SV("\x80"));
346 
347 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
348   if constexpr (sizeof(CharT) > 1) {
349     using V = std::basic_string_view<CharT>;
350 
351     // Unicode fitting in a 16-bit wchar_t
352 
353     // *** Non-printable ***
354 
355     // Space_Separator
356     test_format(V{LR"("\u{a0}\u{3000}")"}, L"{:?}", L"\xa0\x3000");
357 
358     // Line_Separator
359     test_format(V{LR"("\u{2028}")"}, L"{:?}", L"\x2028"); // LINE SEPARATOR
360 
361     // Paragraph_Separator
362     test_format(V{LR"("\u{2029}")"}, L"{:?}", L"\x2029"); // PARAGRAPH SEPARATOR
363 
364     // Format
365     test_format(V{LR"("\u{ad}\u{600}\u{feff}")"}, L"{:?}", L"\xad\x600\xfeff");
366 
367     // Incomplete surrogate pair in UTF-16
368     test_format(V{LR"("\x{d800}")"}, L"{:?}", L"\xd800");
369 
370     // Private_Use
371     test_format(V{LR"("\u{e000}\u{f8ff}")"}, L"{:?}", L"\xe000\xf8ff");
372 
373     // Unassigned
374     test_format(V{LR"("\u{378}\u{1774}\u{ffff}")"}, L"{:?}", L"\x378\x1774\xffff");
375 
376     // Grapheme Extended
377     test_format(V{LR"("\u{300}\u{fe20}")"}, L"{:?}", L"\x300\xfe20");
378   }
379 #  ifndef TEST_SHORT_WCHAR
380   if constexpr (sizeof(CharT) > 2) {
381     static_assert(sizeof(CharT) == 4, "add support for unexpected size");
382     // Unicode fitting in a 32-bit wchar_t
383 
384     constexpr wchar_t x  = 0x1ffff;
385     constexpr std::uint32_t y = 0x1ffff;
386     static_assert(x == y);
387 
388     using V = std::basic_string_view<CharT>;
389 
390     // *** Non-printable ***
391     // Format
392     test_format(V{LR"("\u{110bd}\u{e007f}")"}, L"{:?}", L"\x110bd\xe007f");
393 
394     // Private_Use
395     test_format(V{LR"("\u{f0000}\u{ffffd}\u{100000}\u{10fffd}")"}, L"{:?}", L"\xf0000\xffffd\x100000\x10fffd");
396 
397     // Unassigned
398     test_format(V{LR"("\u{1000c}\u{fffff}\u{10fffe}")"}, L"{:?}", L"\x1000c\xfffff\x10fffe");
399 
400     // Grapheme Extended
401     test_format(V{LR"("\u{101fd}\u{e0100}")"}, L"{:?}", L"\x101fd\xe0100");
402 
403     // Ill-formed
404     test_format(V{LR"("\x{110000}\x{ffffffff}")"}, L"{:?}", L"\x110000\xffffffff");
405   }
406 #  endif // TEST_SHORT_WCHAR
407 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
408 }
409 
410 template <class CharT, class TestFunction>
411 void test_format_functions(TestFunction check) {
412   // *** align-fill & width ***
413   check(SV(R"(***"hellö")"), SV("{:*>10?}"), SV("hellö")); // ö is LATIN SMALL LETTER O WITH DIAERESIS
414   check(SV(R"(*"hellö"**)"), SV("{:*^10?}"), SV("hellö"));
415   check(SV(R"("hellö"***)"), SV("{:*<10?}"), SV("hellö"));
416 
417   check(SV(R"("hello\u{308}")"), SV("{:*>10?}"), SV("hello\u0308"));
418   check(SV(R"(***"hello\u{308}")"), SV("{:*>17?}"), SV("hello\u0308"));
419   check(SV(R"(*"hello\u{308}"**)"), SV("{:*^17?}"), SV("hello\u0308"));
420   check(SV(R"("hello\u{308}"***)"), SV("{:*<17?}"), SV("hello\u0308"));
421 
422   check(SV(R"("hello ����\u{200d}♂\u{fe0f}")"), SV("{:*>10?}"), SV("hello ����‍♂️"));
423   check(SV(R"(***"hello ����\u{200d}♂\u{fe0f}")"), SV("{:*>30?}"), SV("hello ����‍♂️"));
424   check(SV(R"(*"hello ����\u{200d}♂\u{fe0f}"**)"), SV("{:*^30?}"), SV("hello ����‍♂️"));
425   check(SV(R"("hello ����\u{200d}♂\u{fe0f}"***)"), SV("{:*<30?}"), SV("hello ����‍♂️"));
426 
427   // *** width ***
428   check(SV(R"("hellö"   )"), SV("{:10?}"), SV("hellö"));
429   check(SV(R"("hello\u{308}"   )"), SV("{:17?}"), SV("hello\u0308"));
430   check(SV(R"("hello ����\u{200d}♂\u{fe0f}"   )"), SV("{:30?}"), SV("hello ����‍♂️"));
431 
432   // *** precision ***
433   check(SV(R"("hell)"), SV("{:.5?}"), SV("hellö"));
434   check(SV(R"("hellö)"), SV("{:.6?}"), SV("hellö"));
435   check(SV(R"("hellö")"), SV("{:.7?}"), SV("hellö"));
436 
437   check(SV(R"("hello )"), SV("{:.7?}"), SV("hello ����‍♂️"));
438   check(SV(R"("hello )"), SV("{:.8?}"), SV("hello ����‍♂️")); // shrug is two columns
439   check(SV(R"("hello ����)"), SV("{:.9?}"), SV("hello ����‍♂️"));
440   check(SV(R"("hello ����\)"), SV("{:.10?}"), SV("hello ����‍♂️"));
441   check(SV(R"("hello ����\u{200d})"), SV("{:.17?}"), SV("hello ����‍♂️"));
442   check(SV(R"("hello ����\u{200d}♂)"), SV("{:.18?}"), SV("hello ����‍♂️"));
443   check(SV(R"("hello ����\u{200d}♂\)"), SV("{:.19?}"), SV("hello ����‍♂️"));
444   check(SV(R"("hello ����\u{200d}♂\u{fe0f}")"), SV("{:.28?}"), SV("hello ����‍♂️"));
445 
446   // *** width & precision ***
447   check(SV(R"("hell#########################)"), SV("{:#<30.5?}"), SV("hellö"));
448   check(SV(R"("hellö########################)"), SV("{:#<30.6?}"), SV("hellö"));
449   check(SV(R"("hellö"#######################)"), SV("{:#<30.7?}"), SV("hellö"));
450 
451   check(SV(R"("hello #######################)"), SV("{:#<30.7?}"), SV("hello ����‍♂️"));
452   check(SV(R"("hello #######################)"), SV("{:#<30.8?}"), SV("hello ����‍♂️"));
453   check(SV(R"("hello ����#####################)"), SV("{:#<30.9?}"), SV("hello ����‍♂️"));
454   check(SV(R"("hello ����\####################)"), SV("{:#<30.10?}"), SV("hello ����‍♂️"));
455   check(SV(R"("hello ����\u{200d}#############)"), SV("{:#<30.17?}"), SV("hello ����‍♂️"));
456   check(SV(R"("hello ����\u{200d}♂############)"), SV("{:#<30.18?}"), SV("hello ����‍♂️"));
457   check(SV(R"("hello ����\u{200d}♂\###########)"), SV("{:#<30.19?}"), SV("hello ����‍♂️"));
458   check(SV(R"("hello ����\u{200d}♂\u{fe0f}"###)"), SV("{:#<30.28?}"), SV("hello ����‍♂️"));
459 }
460 
461 template <class CharT>
462 void test() {
463   test_char<CharT>();
464   test_string<CharT>();
465 
466   test_format_functions<CharT>(test_format);
467   test_format_functions<CharT>(test_format_to);
468   test_format_functions<CharT>(test_formatted_size);
469   test_format_functions<CharT>(test_format_to_n);
470 }
471 
472 static void test_ill_formed_utf8() {
473   using namespace std::literals;
474 
475   // Too few code units
476   test_format(R"("\x{df}")"sv, "{:?}", "\xdf");
477   test_format(R"("\x{ef}")"sv, "{:?}", "\xef");
478   test_format(R"("\x{ef}\x{bf}")"sv, "{:?}", "\xef\xbf");
479   test_format(R"("\x{f7}")"sv, "{:?}", "\xf7");
480   test_format(R"("\x{f7}\x{bf}")"sv, "{:?}", "\xf7\xbf");
481   test_format(R"("\x{f7}\x{bf}\x{bf}")"sv, "{:?}", "\xf7\xbf\xbf");
482 
483   // Invalid continuation byte
484   test_format(R"("\x{df}a")"sv,
485               "{:?}",
486               "\xdf"
487               "a");
488   test_format(R"("\x{ef}a")"sv,
489               "{:?}",
490               "\xef"
491               "a");
492   test_format(R"("\x{ef}\x{bf}a")"sv,
493               "{:?}",
494               "\xef\xbf"
495               "a");
496   test_format(R"("\x{f7}a")"sv,
497               "{:?}",
498               "\xf7"
499               "a");
500   test_format(R"("\x{f7}\x{bf}a")"sv,
501               "{:?}",
502               "\xf7\xbf"
503               "a");
504   test_format(R"("\x{f7}\x{bf}\x{bf}a")"sv,
505               "{:?}",
506               "\xf7\xbf\xbf"
507               "a");
508 
509   // Code unit out of range
510   test_format(R"("\u{10ffff}")"sv, "{:?}", "\xf4\x8f\xbf\xbf");               // last valid code point
511   test_format(R"("\x{f4}\x{90}\x{80}\x{80}")"sv, "{:?}", "\xf4\x90\x80\x80"); // first invalid code point
512   test_format(R"("\x{f5}\x{b1}\x{b2}\x{b3}")"sv, "{:?}", "\xf5\xb1\xb2\xb3");
513   test_format(R"("\x{f7}\x{bf}\x{bf}\x{bf}")"sv, "{:?}", "\xf7\xbf\xbf\xbf"); // largest encoded code point
514 }
515 
516 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
517 #  ifdef _LIBCPP_SHORT_WCHAR
518 static void test_ill_formed_utf16() {
519   using namespace std::literals;
520 
521   // Too few code units
522   test_format(LR"("\x{d800}")"sv, L"{:?}", L"\xd800");
523   test_format(LR"("\x{dbff}")"sv, L"{:?}", L"\xdbff");
524 
525   // Start with low surrogate pair
526   test_format(LR"("\x{dc00}a")"sv,
527               L"{:?}",
528               L"\xdc00"
529               "a");
530   test_format(LR"("\x{dfff}a")"sv,
531               L"{:?}",
532               L"\xdfff"
533               "a");
534 
535   // Only high surrogate pair
536   test_format(LR"("\x{d800}a")"sv,
537               L"{:?}",
538               L"\xd800"
539               "a");
540   test_format(LR"("\x{dbff}a")"sv,
541               L"{:?}",
542               L"\xdbff"
543               "a");
544 }
545 #  else // _LIBCPP_SHORT_WCHAR
546 static void test_ill_formed_utf32() {
547   using namespace std::literals;
548 
549   test_format(LR"("\u{10ffff}")"sv, L"{:?}", L"\x10ffff");     // last valid code point
550   test_format(LR"("\x{110000}")"sv, L"{:?}", L"\x110000");     // first invalid code point
551   test_format(LR"("\x{ffffffff}")"sv, L"{:?}", L"\xffffffff"); // largest encoded code point
552 }
553 
554 #  endif // _LIBCPP_SHORT_WCHAR
555 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
556 
557 int main(int, char**) {
558   test<char>();
559 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
560   test<wchar_t>();
561 #endif
562 
563   test_ill_formed_utf8();
564 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
565 #  ifdef _LIBCPP_SHORT_WCHAR
566   test_ill_formed_utf16();
567   assert(false);
568 #  else  // _LIBCPP_SHORT_WCHAR
569   test_ill_formed_utf32();
570 #  endif // _LIBCPP_SHORT_WCHAR
571 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
572 
573   return 0;
574 }
575