xref: /llvm-project/libcxx/test/std/utilities/format/format.functions/escaped_output.unicode.pass.cpp (revision dff62f5251f26db856124526ae1b335ddc1e892c) 
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 
10 // This version runs the test when the platform has Unicode support.
11 // UNSUPPORTED: libcpp-has-no-unicode
12 
13 // TODO FMT Investigate Windows and 32-bit AIX issues.
14 // UNSUPPORTED: msvc, target={{.+}}-windows-gnu, target=powerpc-ibm-aix{{.*}}
15 
16 // TODO FMT Fix this test using GCC, it currently crashes.
17 // UNSUPPORTED: gcc-12
18 
19 // TODO FMT This test should not require std::to_chars(floating-point)
20 // XFAIL: availability-fp_to_chars-missing
21 
22 // <format>
23 
24 // This test the debug string type for the formatter specializations for char
25 // and string types. This tests Unicode strings.
26 
27 #include <format>
28 
29 #include <cassert>
30 #include <concepts>
31 #include <iterator>
32 #include <list>
33 #include <vector>
34 
35 #include "test_macros.h"
36 #include "make_string.h"
37 #include "test_format_string.h"
38 #include "assert_macros.h"
39 #include "concat_macros.h"
40 
41 #ifndef TEST_HAS_NO_LOCALIZATION
42 #  include <iostream>
43 #endif
44 
45 #define SV(S) MAKE_STRING_VIEW(CharT, S)
46 
47 auto test_format = []<class CharT, class... Args>(
48                        std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
49   {
50     std::basic_string<CharT> out = std::format(fmt, std::forward<Args>(args)...);
51     TEST_REQUIRE(out == expected,
52                  TEST_WRITE_CONCATENATED(
53                      "\nFormat string   ", fmt.get(), "\nExpected output ", expected, "\nActual output   ", out, '\n'));
54   }
55 #ifndef TEST_HAS_NO_LOCALIZATION
56   {
57     std::basic_string<CharT> out = std::format(std::locale(), fmt, std::forward<Args>(args)...);
58     assert(out == expected);
59   }
60 #endif // TEST_HAS_NO_LOCALIZATION
61 };
62 
63 auto test_format_to =
64     []<class CharT, class... Args>(
65         std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
66       {
67         std::basic_string<CharT> out(expected.size(), CharT(' '));
68         auto it = std::format_to(out.begin(), fmt, std::forward<Args>(args)...);
69         assert(it == out.end());
70         assert(out == expected);
71       }
72 #ifndef TEST_HAS_NO_LOCALIZATION
73       {
74         std::basic_string<CharT> out(expected.size(), CharT(' '));
75         auto it = std::format_to(out.begin(), std::locale(), fmt, std::forward<Args>(args)...);
76         assert(it == out.end());
77         assert(out == expected);
78       }
79 #endif // TEST_HAS_NO_LOCALIZATION
80       {
81         std::list<CharT> out;
82         std::format_to(std::back_inserter(out), fmt, std::forward<Args>(args)...);
83         assert(std::equal(out.begin(), out.end(), expected.begin(), expected.end()));
84       }
85       {
86         std::vector<CharT> out;
87         std::format_to(std::back_inserter(out), fmt, std::forward<Args>(args)...);
88         assert(std::equal(out.begin(), out.end(), expected.begin(), expected.end()));
89       }
90       {
91         assert(expected.size() < 4096 && "Update the size of the buffer.");
92         CharT out[4096];
93         CharT* it = std::format_to(out, fmt, std::forward<Args>(args)...);
94         assert(std::distance(out, it) == int(expected.size()));
95         // Convert to std::string since output contains '\0' for boolean tests.
96         assert(std::basic_string<CharT>(out, it) == expected);
97       }
98     };
99 
100 auto test_formatted_size =
101     []<class CharT, class... Args>(
102         std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
103       {
104         std::size_t size = std::formatted_size(fmt, std::forward<Args>(args)...);
105         assert(size == expected.size());
106       }
107 #ifndef TEST_HAS_NO_LOCALIZATION
108       {
109         std::size_t size = std::formatted_size(std::locale(), fmt, std::forward<Args>(args)...);
110         assert(size == expected.size());
111       }
112 #endif // TEST_HAS_NO_LOCALIZATION
113     };
114 
115 auto test_format_to_n =
116     []<class CharT, class... Args>(
117         std::basic_string_view<CharT> expected, test_format_string<CharT, Args...> fmt, Args&&... args) {
118       {
119         std::size_t n = expected.size();
120         std::basic_string<CharT> out(n, CharT(' '));
121         std::format_to_n_result result = std::format_to_n(out.begin(), n, fmt, std::forward<Args>(args)...);
122         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
123         assert(result.out == out.end());
124         assert(out == expected);
125       }
126 #ifndef TEST_HAS_NO_LOCALIZATION
127       {
128         std::size_t n = expected.size();
129         std::basic_string<CharT> out(n, CharT(' '));
130         std::format_to_n_result result =
131             std::format_to_n(out.begin(), n, std::locale(), fmt, std::forward<Args>(args)...);
132         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
133         assert(result.out == out.end());
134         assert(out == expected);
135       }
136 #endif // TEST_HAS_NO_LOCALIZATION
137       {
138         std::ptrdiff_t n = 0;
139         std::basic_string<CharT> out;
140         std::format_to_n_result result = std::format_to_n(out.begin(), n, fmt, std::forward<Args>(args)...);
141         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
142         assert(result.out == out.end());
143         assert(out.empty());
144       }
145       {
146         std::ptrdiff_t n = expected.size() / 2;
147         std::basic_string<CharT> out(n, CharT(' '));
148         std::format_to_n_result result = std::format_to_n(out.begin(), n, fmt, std::forward<Args>(args)...);
149         assert(result.size == static_cast<std::ptrdiff_t>(expected.size()));
150         assert(result.out == out.end());
151         assert(out == expected.substr(0, n));
152       }
153     };
154 
155 template <class CharT>
156 void test_char() {
157   // *** P2286 examples ***
158   test_format(SV("['\\'', '\"']"), SV("[{:?}, {:?}]"), CharT('\''), CharT('"'));
159 
160   // *** Specical cases ***
161   test_format(SV("'\\t'"), SV("{:?}"), CharT('\t'));
162   test_format(SV("'\\n'"), SV("{:?}"), CharT('\n'));
163   test_format(SV("'\\r'"), SV("{:?}"), CharT('\r'));
164   test_format(SV("'\\\\'"), SV("{:?}"), CharT('\\'));
165 
166   test_format(SV("'\\\''"), SV("{:?}"), CharT('\''));
167   test_format(SV("'\"'"), SV("{:?}"), CharT('"')); // only special for string
168 
169   test_format(SV("' '"), SV("{:?}"), CharT(' '));
170 
171   // *** Printable ***
172   test_format(SV("'a'"), SV("{:?}"), CharT('a'));
173   test_format(SV("'b'"), SV("{:?}"), CharT('b'));
174   test_format(SV("'c'"), SV("{:?}"), CharT('c'));
175 
176   // *** Non-printable ***
177 
178   // Control
179   test_format(SV("'\\u{0}'"), SV("{:?}"), CharT('\0'));
180   test_format(SV("'\\u{1f}'"), SV("{:?}"), CharT('\x1f'));
181 
182   // Ill-formed
183   if constexpr (sizeof(CharT) == 1)
184     test_format(SV("'\\x{80}'"), SV("{:?}"), CharT('\x80'));
185 
186 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
187   if constexpr (sizeof(CharT) > 1) {
188     using V = std::basic_string_view<CharT>;
189 
190     // Unicode fitting in a 16-bit wchar_t
191 
192     // *** Non-printable ***
193 
194     // Space_Separator
195     test_format(V{L"'\\u{a0}'"}, L"{:?}", L'\xa0');     // NO-BREAK SPACE
196     test_format(V{L"'\\u{3000}'"}, L"{:?}", L'\x3000'); // IDEOGRAPHIC SPACE
197 
198     // Line_Separator
199     test_format(V{L"'\\u{2028}'"}, L"{:?}", L'\x2028'); // LINE SEPARATOR
200 
201     // Paragraph_Separator
202     test_format(V{L"'\\u{2029}'"}, L"{:?}", L'\x2029'); // PARAGRAPH SEPARATOR
203 
204     // Format
205     test_format(V{L"'\\u{ad}'"}, L"{:?}", L'\xad');     // SOFT HYPHEN
206     test_format(V{L"'\\u{600}'"}, L"{:?}", L'\x600');   // ARABIC NUMBER SIGN
207     test_format(V{L"'\\u{feff}'"}, L"{:?}", L'\xfeff'); // ZERO WIDTH NO-BREAK SPACE
208 
209     // Incomplete surrogate pair in UTF-16
210     test_format(V{L"'\\x{d800}'"}, L"{:?}", L'\xd800'); // <surrogate-D800>
211     test_format(V{L"'\\x{dfff}'"}, L"{:?}", L'\xdfff'); // <surrogate-DFFF>
212 
213     // Private_Use
214     test_format(V{L"'\\u{e000}'"}, L"{:?}", L'\xe000'); // <private-use-E000>
215     test_format(V{L"'\\u{f8ff}'"}, L"{:?}", L'\xf8ff'); // <private-use-F8FF>
216 
217     // Unassigned
218     test_format(V{L"'\\u{378}'"}, L"{:?}", L'\x378');   // <reserved-0378>
219     test_format(V{L"'\\u{1774}'"}, L"{:?}", L'\x1774'); // <reserved-1774>
220     test_format(V{L"'\\u{ffff}'"}, L"{:?}", L'\xffff'); // <noncharacter-FFFF>
221 
222     // Grapheme Extended
223     test_format(V{L"'\\u{300}'"}, L"{:?}", L'\x300');   // COMBINING GRAVE ACCENT
224     test_format(V{L"'\\u{fe20}'"}, L"{:?}", L'\xfe20'); // VARIATION SELECTOR-1
225   }
226 #  ifndef TEST_SHORT_WCHAR
227   if constexpr (sizeof(CharT) > 2) {
228     static_assert(sizeof(CharT) == 4, "add support for unexpected size");
229     // Unicode fitting in a 32-bit wchar_t
230 
231     constexpr wchar_t x  = 0x1ffff;
232     constexpr std::uint32_t y = 0x1ffff;
233     static_assert(x == y);
234 
235     using V = std::basic_string_view<CharT>;
236 
237     // *** Non-printable ***
238     // Format
239     test_format(V{L"'\\u{110bd}'"}, L"{:?}", L'\x110bd'); // KAITHI NUMBER SIGN
240     test_format(V{L"'\\u{e007f}'"}, L"{:?}", L'\xe007f'); // CANCEL TAG
241 
242     // Private_Use
243     test_format(V{L"'\\u{f0000}'"}, L"{:?}", L'\xf0000'); // <private-use-F0000>
244     test_format(V{L"'\\u{ffffd}'"}, L"{:?}", L'\xffffd'); // <private-use-FFFFD>
245 
246     test_format(V{L"'\\u{100000}'"}, L"{:?}", L'\x100000'); // <private-use-100000>
247     test_format(V{L"'\\u{10fffd}'"}, L"{:?}", L'\x10fffd'); // <private-use-10FFFD>
248 
249     // Unassigned
250     test_format(V{L"'\\u{1000c}'"}, L"{:?}", L'\x1000c');   // <reserved-1000c>
251     test_format(V{L"'\\u{fffff}'"}, L"{:?}", L'\xfffff');   // <noncharacter-FFFFF>
252     test_format(V{L"'\\u{10fffe}'"}, L"{:?}", L'\x10fffe'); // <noncharacter-10FFFE>
253 
254     // Grapheme Extended
255     test_format(V{L"'\\u{101fd}'"}, L"{:?}", L'\x101fd'); // COMBINING OLD PERMIC LETTER AN
256     test_format(V{L"'\\u{e0100}'"}, L"{:?}", L'\xe0100'); // VARIATION SELECTOR-17
257 
258     // Ill-formed
259     test_format(V{L"'\\x{110000}'"}, L"{:?}", L'\x110000');
260     test_format(V{L"'\\x{ffffffff}'"}, L"{:?}", L'\xffffffff');
261   }
262 #  endif // TEST_SHORT_WCHAR
263 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
264 }
265 
266 template <class CharT>
267 void test_string() {
268   // *** P2286 examples ***
269   test_format(SV("[h\tllo]"), SV("[{}]"), SV("h\tllo"));
270   test_format(SV(R"(["h\tllo"])"), SV("[{:?}]"), SV("h\tllo"));
271   test_format(SV(R"(["Спасибо, Виктор ♥!"])"), SV("[{:?}]"), SV("Спасибо, Виктор ♥!"));
272 
273   test_format(SV(R"(["\u{0} \n \t \u{2} \u{1b}"])"), SV("[{:?}]"), SV("\0 \n \t \x02 \x1b"));
274 
275   if constexpr (sizeof(CharT) == 1) {
276     // Ill-formend UTF-8
277     test_format(SV(R"(["\x{c3}"])"), SV("[{:?}]"), "\xc3");
278     test_format(SV(R"(["\x{c3}("])"), SV("[{:?}]"), "\xc3\x28");
279 
280     /* U+0000..U+0007F 1 code unit range, encoded in 2 code units. */
281     test_format(SV(R"(["\x{c0}\x{80}"])"), SV("[{:?}]"), "\xc0\x80"); // U+0000
282     test_format(SV(R"(["\x{c1}\x{bf}"])"), SV("[{:?}]"), "\xc1\xbf"); // U+007F
283     test_format(SV(R"(["\u{80}"])"), SV("[{:?}]"), "\xc2\x80");       // U+0080 first valid (General_Category=Control)
284 
285     /* U+0000..U+07FFF 1 and 2 code unit range, encoded in 3 code units. */
286     test_format(SV(R"(["\x{e0}\x{80}\x{80}"])"), SV("[{:?}]"), "\xe0\x80\x80"); // U+0000
287     test_format(SV(R"(["\x{e0}\x{81}\x{bf}"])"), SV("[{:?}]"), "\xe0\x81\xbf"); // U+007F
288     test_format(SV(R"(["\x{e0}\x{82}\x{80}"])"), SV("[{:?}]"), "\xe0\x82\x80"); // U+0080
289     test_format(SV(R"(["\x{e0}\x{9f}\x{bf}"])"), SV("[{:?}]"), "\xe0\x9f\xbf"); // U+07FF
290     test_format(SV("[\"\u0800\"]"), SV("[{:?}]"), "\xe0\xa0\x80");              // U+0800 first valid
291 
292 #if 0
293 	// This code point is in the Hangul Jamo Extended-B block and at the time of writing
294 	// it's unassigned. When it comes defined, this branch might become true.
295     test_format(SV("[\"\ud7ff\"]"), SV("[{:?}]"), "\xed\x9f\xbf");              // U+D7FF last valid
296 #else
297     /* U+D800..D+DFFFF surrogate range */
298     test_format(SV(R"(["\u{d7ff}"])"), SV("[{:?}]"), "\xed\x9f\xbf");           // U+D7FF last valid
299 #endif
300     test_format(SV(R"(["\x{ed}\x{a0}\x{80}"])"), SV("[{:?}]"), "\xed\xa0\x80"); // U+D800
301     test_format(SV(R"(["\x{ed}\x{af}\x{bf}"])"), SV("[{:?}]"), "\xed\xaf\xbf"); // U+DBFF
302     test_format(SV(R"(["\x{ed}\x{bf}\x{80}"])"), SV("[{:?}]"), "\xed\xbf\x80"); // U+DC00
303     test_format(SV(R"(["\x{ed}\x{bf}\x{bf}"])"), SV("[{:?}]"), "\xed\xbf\xbf"); // U+DFFF
304     test_format(SV(R"(["\u{e000}"])"), SV("[{:?}]"), "\xee\x80\x80");           // U+E000 first valid
305                                                                                 // (in the Private Use Area block)
306 
307     /* U+0000..U+FFFF 1, 2, and 3 code unit range */
308     test_format(SV(R"(["\x{f0}\x{80}\x{80}\x{80}"])"), SV("[{:?}]"), "\xf0\x80\x80\x80"); // U+0000
309     test_format(SV(R"(["\x{f0}\x{80}\x{81}\x{bf}"])"), SV("[{:?}]"), "\xf0\x80\x81\xbf"); // U+007F
310     test_format(SV(R"(["\x{f0}\x{80}\x{82}\x{80}"])"), SV("[{:?}]"), "\xf0\x80\x82\x80"); // U+0080
311     test_format(SV(R"(["\x{f0}\x{80}\x{9f}\x{bf}"])"), SV("[{:?}]"), "\xf0\x80\x9f\xbf"); // U+07FF
312     test_format(SV(R"(["\x{f0}\x{80}\x{a0}\x{80}"])"), SV("[{:?}]"), "\xf0\x80\xa0\x80"); // U+0800
313     test_format(SV(R"(["\x{f0}\x{8f}\x{bf}\x{bf}"])"), SV("[{:?}]"), "\xf0\x8f\xbf\xbf"); // U+FFFF
314     test_format(SV("[\"\U00010000\"]"), SV("[{:?}]"), "\xf0\x90\x80\x80");                // U+10000 first valid
315 
316     /* U+10FFFF..U+1FFFFF invalid range */
317     test_format(SV(R"(["\u{10ffff}"])"), SV("[{:?}]"), "\xf4\x8f\xbf\xbf"); // U+10FFFF last valid
318                                                                             // (in Supplementary Private Use Area-B)
319     test_format(SV(R"(["\x{f4}\x{90}\x{80}\x{80}"])"), SV("[{:?}]"), "\xf4\x90\x80\x80"); // U+110000
320     test_format(SV(R"(["\x{f4}\x{bf}\x{bf}\x{bf}"])"), SV("[{:?}]"), "\xf4\xbf\xbf\xbf"); // U+11FFFF
321   } else {
322     // Valid UTF-16 and UTF-32
323     test_format(SV("[\"\u00c3\"]"), SV("[{:?}]"), L"\xc3"); // LATIN CAPITAL LETTER A WITH TILDE
324     test_format(SV("[\"\u00c3(\"]"), SV("[{:?}]"), L"\xc3\x28");
325   }
326 
327   test_format(SV(R"(["����\u{200d}♂\u{fe0f}"])"), SV("[{:?}]"), SV("����‍♂️"));
328 
329   // *** Specical cases ***
330   test_format(SV(R"("\t\n\r\\'\" ")"), SV("{:?}"), SV("\t\n\r\\'\" "));
331 
332   // *** Printable ***
333   test_format(SV(R"("abcdefg")"), SV("{:?}"), SV("abcdefg"));
334 
335   // *** Non-printable ***
336 
337   // Control
338   test_format(SV(R"("\u{0}\u{1f}")"), SV("{:?}"), SV("\0\x1f"));
339 
340   // Ill-formed
341   if constexpr (sizeof(CharT) == 1)
342     test_format(SV(R"("\x{80}")"), SV("{:?}"), SV("\x80"));
343 
344 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
345   if constexpr (sizeof(CharT) > 1) {
346     using V = std::basic_string_view<CharT>;
347 
348     // Unicode fitting in a 16-bit wchar_t
349 
350     // *** Non-printable ***
351 
352     // Space_Separator
353     test_format(V{LR"("\u{a0}\u{3000}")"}, L"{:?}", L"\xa0\x3000");
354 
355     // Line_Separator
356     test_format(V{LR"("\u{2028}")"}, L"{:?}", L"\x2028"); // LINE SEPARATOR
357 
358     // Paragraph_Separator
359     test_format(V{LR"("\u{2029}")"}, L"{:?}", L"\x2029"); // PARAGRAPH SEPARATOR
360 
361     // Format
362     test_format(V{LR"("\u{ad}\u{600}\u{feff}")"}, L"{:?}", L"\xad\x600\xfeff");
363 
364     // Incomplete surrogate pair in UTF-16
365     test_format(V{LR"("\x{d800}")"}, L"{:?}", L"\xd800");
366 
367     // Private_Use
368     test_format(V{LR"("\u{e000}\u{f8ff}")"}, L"{:?}", L"\xe000\xf8ff");
369 
370     // Unassigned
371     test_format(V{LR"("\u{378}\u{1774}\u{ffff}")"}, L"{:?}", L"\x378\x1774\xffff");
372 
373     // Grapheme Extended
374     test_format(V{LR"("\u{300}\u{fe20}")"}, L"{:?}", L"\x300\xfe20");
375   }
376 #  ifndef TEST_SHORT_WCHAR
377   if constexpr (sizeof(CharT) > 2) {
378     static_assert(sizeof(CharT) == 4, "add support for unexpected size");
379     // Unicode fitting in a 32-bit wchar_t
380 
381     constexpr wchar_t x  = 0x1ffff;
382     constexpr std::uint32_t y = 0x1ffff;
383     static_assert(x == y);
384 
385     using V = std::basic_string_view<CharT>;
386 
387     // *** Non-printable ***
388     // Format
389     test_format(V{LR"("\u{110bd}\u{e007f}")"}, L"{:?}", L"\x110bd\xe007f");
390 
391     // Private_Use
392     test_format(V{LR"("\u{f0000}\u{ffffd}\u{100000}\u{10fffd}")"}, L"{:?}", L"\xf0000\xffffd\x100000\x10fffd");
393 
394     // Unassigned
395     test_format(V{LR"("\u{1000c}\u{fffff}\u{10fffe}")"}, L"{:?}", L"\x1000c\xfffff\x10fffe");
396 
397     // Grapheme Extended
398     test_format(V{LR"("\u{101fd}\u{e0100}")"}, L"{:?}", L"\x101fd\xe0100");
399 
400     // Ill-formed
401     test_format(V{LR"("\x{110000}\x{ffffffff}")"}, L"{:?}", L"\x110000\xffffffff");
402   }
403 #  endif // TEST_SHORT_WCHAR
404 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
405 }
406 
407 template <class CharT, class TestFunction>
408 void test_format_functions(TestFunction check) {
409   // *** align-fill & width ***
410   check(SV(R"(***"hellö")"), SV("{:*>10?}"), SV("hellö")); // ö is LATIN SMALL LETTER O WITH DIAERESIS
411   check(SV(R"(*"hellö"**)"), SV("{:*^10?}"), SV("hellö"));
412   check(SV(R"("hellö"***)"), SV("{:*<10?}"), SV("hellö"));
413 
414   check(SV(R"("hello\u{308}")"), SV("{:*>10?}"), SV("hello\u0308"));
415   check(SV(R"(***"hello\u{308}")"), SV("{:*>17?}"), SV("hello\u0308"));
416   check(SV(R"(*"hello\u{308}"**)"), SV("{:*^17?}"), SV("hello\u0308"));
417   check(SV(R"("hello\u{308}"***)"), SV("{:*<17?}"), SV("hello\u0308"));
418 
419   check(SV(R"("hello ����\u{200d}♂\u{fe0f}")"), SV("{:*>10?}"), SV("hello ����‍♂️"));
420   check(SV(R"(***"hello ����\u{200d}♂\u{fe0f}")"), SV("{:*>30?}"), SV("hello ����‍♂️"));
421   check(SV(R"(*"hello ����\u{200d}♂\u{fe0f}"**)"), SV("{:*^30?}"), SV("hello ����‍♂️"));
422   check(SV(R"("hello ����\u{200d}♂\u{fe0f}"***)"), SV("{:*<30?}"), SV("hello ����‍♂️"));
423 
424   // *** width ***
425   check(SV(R"("hellö"   )"), SV("{:10?}"), SV("hellö"));
426   check(SV(R"("hello\u{308}"   )"), SV("{:17?}"), SV("hello\u0308"));
427   check(SV(R"("hello ����\u{200d}♂\u{fe0f}"   )"), SV("{:30?}"), SV("hello ����‍♂️"));
428 
429   // *** precision ***
430   check(SV(R"("hell)"), SV("{:.5?}"), SV("hellö"));
431   check(SV(R"("hellö)"), SV("{:.6?}"), SV("hellö"));
432   check(SV(R"("hellö")"), SV("{:.7?}"), SV("hellö"));
433 
434   check(SV(R"("hello )"), SV("{:.7?}"), SV("hello ����‍♂️"));
435   check(SV(R"("hello )"), SV("{:.8?}"), SV("hello ����‍♂️")); // shrug is two columns
436   check(SV(R"("hello ����)"), SV("{:.9?}"), SV("hello ����‍♂️"));
437   check(SV(R"("hello ����\)"), SV("{:.10?}"), SV("hello ����‍♂️"));
438   check(SV(R"("hello ����\u{200d})"), SV("{:.17?}"), SV("hello ����‍♂️"));
439   check(SV(R"("hello ����\u{200d}♂)"), SV("{:.18?}"), SV("hello ����‍♂️"));
440   check(SV(R"("hello ����\u{200d}♂\)"), SV("{:.19?}"), SV("hello ����‍♂️"));
441   check(SV(R"("hello ����\u{200d}♂\u{fe0f}")"), SV("{:.28?}"), SV("hello ����‍♂️"));
442 
443   // *** width & precision ***
444   check(SV(R"("hell#########################)"), SV("{:#<30.5?}"), SV("hellö"));
445   check(SV(R"("hellö########################)"), SV("{:#<30.6?}"), SV("hellö"));
446   check(SV(R"("hellö"#######################)"), SV("{:#<30.7?}"), SV("hellö"));
447 
448   check(SV(R"("hello #######################)"), SV("{:#<30.7?}"), SV("hello ����‍♂️"));
449   check(SV(R"("hello #######################)"), SV("{:#<30.8?}"), SV("hello ����‍♂️"));
450   check(SV(R"("hello ����#####################)"), SV("{:#<30.9?}"), SV("hello ����‍♂️"));
451   check(SV(R"("hello ����\####################)"), SV("{:#<30.10?}"), SV("hello ����‍♂️"));
452   check(SV(R"("hello ����\u{200d}#############)"), SV("{:#<30.17?}"), SV("hello ����‍♂️"));
453   check(SV(R"("hello ����\u{200d}♂############)"), SV("{:#<30.18?}"), SV("hello ����‍♂️"));
454   check(SV(R"("hello ����\u{200d}♂\###########)"), SV("{:#<30.19?}"), SV("hello ����‍♂️"));
455   check(SV(R"("hello ����\u{200d}♂\u{fe0f}"###)"), SV("{:#<30.28?}"), SV("hello ����‍♂️"));
456 }
457 
458 template <class CharT>
459 void test() {
460   test_char<CharT>();
461   test_string<CharT>();
462 
463   test_format_functions<CharT>(test_format);
464   test_format_functions<CharT>(test_format_to);
465   test_format_functions<CharT>(test_formatted_size);
466   test_format_functions<CharT>(test_format_to_n);
467 }
468 
469 static void test_ill_formed_utf8() {
470   using namespace std::literals;
471 
472   // Too few code units
473   test_format(R"("\x{df}")"sv, "{:?}", "\xdf");
474   test_format(R"("\x{ef}")"sv, "{:?}", "\xef");
475   test_format(R"("\x{ef}\x{bf}")"sv, "{:?}", "\xef\xbf");
476   test_format(R"("\x{f7}")"sv, "{:?}", "\xf7");
477   test_format(R"("\x{f7}\x{bf}")"sv, "{:?}", "\xf7\xbf");
478   test_format(R"("\x{f7}\x{bf}\x{bf}")"sv, "{:?}", "\xf7\xbf\xbf");
479 
480   // Invalid continuation byte
481   test_format(R"("\x{df}a")"sv,
482               "{:?}",
483               "\xdf"
484               "a");
485   test_format(R"("\x{ef}a")"sv,
486               "{:?}",
487               "\xef"
488               "a");
489   test_format(R"("\x{ef}\x{bf}a")"sv,
490               "{:?}",
491               "\xef\xbf"
492               "a");
493   test_format(R"("\x{f7}a")"sv,
494               "{:?}",
495               "\xf7"
496               "a");
497   test_format(R"("\x{f7}\x{bf}a")"sv,
498               "{:?}",
499               "\xf7\xbf"
500               "a");
501   test_format(R"("\x{f7}\x{bf}\x{bf}a")"sv,
502               "{:?}",
503               "\xf7\xbf\xbf"
504               "a");
505 
506   // Code unit out of range
507   test_format(R"("\u{10ffff}")"sv, "{:?}", "\xf4\x8f\xbf\xbf");               // last valid code point
508   test_format(R"("\x{f4}\x{90}\x{80}\x{80}")"sv, "{:?}", "\xf4\x90\x80\x80"); // first invalid code point
509   test_format(R"("\x{f5}\x{b1}\x{b2}\x{b3}")"sv, "{:?}", "\xf5\xb1\xb2\xb3");
510   test_format(R"("\x{f7}\x{bf}\x{bf}\x{bf}")"sv, "{:?}", "\xf7\xbf\xbf\xbf"); // largest encoded code point
511 }
512 
513 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
514 #  ifdef _LIBCPP_SHORT_WCHAR
515 static void test_ill_formed_utf16() {
516   using namespace std::literals;
517 
518   // Too few code units
519   test_format(LR"("\x{d800}")"sv, L"{:?}", L"\xd800");
520   test_format(LR"("\x{dbff}")"sv, L"{:?}", L"\xdbff");
521 
522   // Start with low surrogate pair
523   test_format(LR"("\x{dc00}a")"sv,
524               L"{:?}",
525               L"\xdc00"
526               "a");
527   test_format(LR"("\x{dfff}a")"sv,
528               L"{:?}",
529               L"\xdfff"
530               "a");
531 
532   // Only high surrogate pair
533   test_format(LR"("\x{d800}a")"sv,
534               L"{:?}",
535               L"\xd800"
536               "a");
537   test_format(LR"("\x{dbff}a")"sv,
538               L"{:?}",
539               L"\xdbff"
540               "a");
541 }
542 #  else // _LIBCPP_SHORT_WCHAR
543 static void test_ill_formed_utf32() {
544   using namespace std::literals;
545 
546   test_format(LR"("\u{10ffff}")"sv, L"{:?}", L"\x10ffff");     // last valid code point
547   test_format(LR"("\x{110000}")"sv, L"{:?}", L"\x110000");     // first invalid code point
548   test_format(LR"("\x{ffffffff}")"sv, L"{:?}", L"\xffffffff"); // largest encoded code point
549 }
550 
551 #  endif // _LIBCPP_SHORT_WCHAR
552 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
553 
554 int main(int, char**) {
555   test<char>();
556 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
557   test<wchar_t>();
558 #endif
559 
560   test_ill_formed_utf8();
561 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
562 #  ifdef _LIBCPP_SHORT_WCHAR
563   test_ill_formed_utf16();
564   assert(false);
565 #  else  // _LIBCPP_SHORT_WCHAR
566   test_ill_formed_utf32();
567 #  endif // _LIBCPP_SHORT_WCHAR
568 #endif   // TEST_HAS_NO_WIDE_CHARACTERS
569 
570   return 0;
571 }
572