1// -*- C++ -*- 2//===----------------------------------------------------------------------===// 3// 4// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5// See https://llvm.org/LICENSE.txt for license information. 6// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7// 8//===----------------------------------------------------------------------===// 9 10#ifndef _LIBCPP_PRINT 11#define _LIBCPP_PRINT 12 13/* 14namespace std { 15 // [print.fun], print functions 16 template<class... Args> 17 void print(format_string<Args...> fmt, Args&&... args); 18 template<class... Args> 19 void print(FILE* stream, format_string<Args...> fmt, Args&&... args); 20 21 template<class... Args> 22 void println(format_string<Args...> fmt, Args&&... args); 23 template<class... Args> 24 void println(FILE* stream, format_string<Args...> fmt, Args&&... args); 25 26 void vprint_unicode(string_view fmt, format_args args); 27 void vprint_unicode(FILE* stream, string_view fmt, format_args args); 28 29 void vprint_nonunicode(string_view fmt, format_args args); 30 void vprint_nonunicode(FILE* stream, string_view fmt, format_args args); 31} 32*/ 33 34#include <__assert> // all public C++ headers provide the assertion handler 35#include <__availability> 36#include <__concepts/same_as.h> 37#include <__config> 38#include <__system_error/system_error.h> 39#include <__utility/forward.h> 40#include <cerrno> 41#include <cstdio> 42#include <format> 43#include <string> 44#include <string_view> 45#include <version> 46 47#if !defined(_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER) 48# pragma GCC system_header 49#endif 50 51_LIBCPP_BEGIN_NAMESPACE_STD 52 53#ifdef _LIBCPP_WIN32API 54_LIBCPP_EXPORTED_FROM_ABI bool __is_windows_terminal(FILE* __stream); 55 56# ifndef _LIBCPP_HAS_NO_WIDE_CHARACTERS 57// A wrapper for WriteConsoleW which is used to write to the Windows 58// console. This function is in the dylib to avoid pulling in windows.h 59// in the library headers. The function itself uses some private parts 60// of the dylib too. 61// 62// The function does not depend on the language standard used. Guarding 63// it with C++23 would fail since the dylib is currently built using C++20. 64// 65// Note the function is only implemented on the Windows platform. 66_LIBCPP_EXPORTED_FROM_ABI void __write_to_windows_console(FILE* __stream, wstring_view __view); 67# endif // _LIBCPP_HAS_NO_WIDE_CHARACTERS 68#elif __has_include(<unistd.h>) 69_LIBCPP_EXPORTED_FROM_ABI bool __is_posix_terminal(FILE* __stream); 70#endif // _LIBCPP_WIN32API 71 72#if _LIBCPP_STD_VER >= 23 73 74# ifndef _LIBCPP_HAS_NO_UNICODE 75// This is the code to transcode UTF-8 to UTF-16. This is used on 76// Windows for the native Unicode API. The code is modeled to make it 77// easier to extend to 78// 79// P2728R0 Unicode in the Library, Part 1: UTF Transcoding 80// 81// This paper is still under heavy development so it makes no sense yet 82// to strictly follow the paper. 83namespace __unicode { 84 85// The names of these concepts are modelled after P2728R0, but the 86// implementation is not. char16_t may contain 32-bits so depending on the 87// number of bits is an issue. 88# ifdef _LIBCPP_SHORT_WCHAR 89template <class _Tp> 90concept __utf16_code_unit = 91 same_as<_Tp, char16_t> 92# ifndef _LIBCPP_HAS_NO_WIDE_CHARACTERS 93 || same_as<_Tp, wchar_t> 94# endif 95 ; 96template <class _Tp> 97concept __utf32_code_unit = same_as<_Tp, char32_t>; 98# else // _LIBCPP_SHORT_WCHAR 99template <class _Tp> 100concept __utf16_code_unit = same_as<_Tp, char16_t>; 101template <class _Tp> 102concept __utf32_code_unit = 103 same_as<_Tp, char32_t> 104# ifndef _LIBCPP_HAS_NO_WIDE_CHARACTERS 105 || same_as<_Tp, wchar_t> 106# endif 107 ; 108# endif // _LIBCPP_SHORT_WCHAR 109 110// Pass by reference since an output_iterator may not be copyable. 111template <class _OutIt> 112_LIBCPP_HIDE_FROM_ABI constexpr void __encode(_OutIt&, char32_t) = delete; 113 114template <class _OutIt> 115 requires __utf16_code_unit<iter_value_t<_OutIt>> 116_LIBCPP_HIDE_FROM_ABI constexpr void __encode(_OutIt& __out_it, char32_t __value) { 117 // [print.fun]/7 : "if `out` contains invalid code units, the behavior is undefined and implementations are encouraged 118 // to diagnose it". 119 _LIBCPP_ASSERT_UNCATEGORIZED(__is_scalar_value(__value), "an invalid unicode scalar value results in invalid UTF-16"); 120 121 if (__value < 0x10000) { 122 *__out_it++ = __value; 123 return; 124 } 125 126 __value -= 0x10000; 127 *__out_it++ = 0xd800 + (__value >> 10); 128 *__out_it++ = 0xdc00 + (__value & 0x3FF); 129} 130 131template <class _OutIt> 132 requires __utf32_code_unit<iter_value_t<_OutIt>> 133_LIBCPP_HIDE_FROM_ABI constexpr void __encode(_OutIt& __out_it, char32_t __value) { 134 // [print.fun]/7 : "if `out` contains invalid code units, the behavior is undefined and implementations are encouraged 135 // to diagnose it". 136 _LIBCPP_ASSERT_UNCATEGORIZED(__is_scalar_value(__value), "an invalid unicode scalar value results in invalid UTF-32"); 137 *__out_it++ = __value; 138} 139 140template <class _OutIt, input_iterator _InIt> 141 requires output_iterator<_OutIt, const iter_value_t<_OutIt>&> && (!same_as<iter_value_t<_OutIt>, iter_value_t<_InIt>>) 142_LIBCPP_HIDE_FROM_ABI constexpr _OutIt __transcode(_InIt __first, _InIt __last, _OutIt __out_it) { 143 // The __code_point_view has a basic_string_view interface. 144 // When transcoding becomes part of the standard we probably want to 145 // look at smarter algorithms. 146 // For example, when processing a code point that is encoded in 147 // 1 to 3 code units in UTF-8, the result will always be encoded 148 // in 1 code unit in UTF-16 (code points that require 4 code 149 // units in UTF-8 will require 2 code units in UTF-16). 150 // 151 // Note if P2728 is accepted types like int may become valid. In that case 152 // the __code_point_view should use a span. Libc++ will remove support for 153 // char_traits<int>. 154 155 // TODO PRINT Validate with clang-tidy 156 // NOLINTNEXTLINE(bugprone-dangling-handle) 157 basic_string_view<iter_value_t<_InIt>> __data{__first, __last}; 158 __code_point_view<iter_value_t<_InIt>> __view{__data.begin(), __data.end()}; 159 while (!__view.__at_end()) 160 __unicode::__encode(__out_it, __view.__consume().__code_point); 161 return __out_it; 162} 163 164} // namespace __unicode 165 166# endif // _LIBCPP_HAS_NO_UNICODE 167 168namespace __print { 169 170// [print.fun]/2 171// Effects: If the ordinary literal encoding ([lex.charset]) is UTF-8, equivalent to: 172// vprint_unicode(stream, fmt.str, make_format_args(args...)); 173// Otherwise, equivalent to: 174// vprint_nonunicode(stream, fmt.str, make_format_args(args...)); 175// 176// Based on the compiler and its compilation flags this value is or is 177// not true. As mentioned in P2093R14 this only affects Windows. The 178// test below could also be done for 179// - GCC using __GNUC_EXECUTION_CHARSET_NAME 180// https://gcc.gnu.org/onlinedocs/cpp/Common-Predefined-Macros.html 181// - Clang using __clang_literal_encoding__ 182// https://clang.llvm.org/docs/LanguageExtensions.html#builtin-macros 183// (note at the time of writing Clang is hard-coded to UTF-8.) 184// 185 186# ifdef _LIBCPP_HAS_NO_UNICODE 187inline constexpr bool __use_unicode_execution_charset = false; 188# elif defined(_MSVC_EXECUTION_CHARACTER_SET) 189// This is the same test MSVC STL uses in their implementation of <print> 190// See: https://learn.microsoft.com/en-us/windows/win32/intl/code-page-identifiers 191inline constexpr bool __use_unicode_execution_charset = _MSVC_EXECUTION_CHARACTER_SET == 65001; 192# else 193inline constexpr bool __use_unicode_execution_charset = true; 194# endif 195 196_LIBCPP_HIDE_FROM_ABI inline bool __is_terminal([[maybe_unused]] FILE* __stream) { 197 // The macro _LIBCPP_TESTING_PRINT_IS_TERMINAL is used to change 198 // the behavior in the test. This is not part of the public API. 199# ifdef _LIBCPP_TESTING_PRINT_IS_TERMINAL 200 return _LIBCPP_TESTING_PRINT_IS_TERMINAL(__stream); 201# elif _LIBCPP_AVAILABILITY_HAS_PRINT == 0 202 return false; 203# elif defined(_LIBCPP_WIN32API) 204 return std::__is_windows_terminal(__stream); 205# elif __has_include(<unistd.h>) 206 return std::__is_posix_terminal(__stream); 207# else 208# error "Provide a way to determine whether a FILE* is a terminal" 209# endif 210} 211 212template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 213_LIBCPP_HIDE_FROM_ABI inline void 214__vprint_nonunicode(FILE* __stream, string_view __fmt, format_args __args, bool __write_nl) { 215 _LIBCPP_ASSERT_NON_NULL(__stream, "__stream must be a valid pointer to an output C stream"); 216 string __str = std::vformat(__fmt, __args); 217 if (__write_nl) 218 __str.push_back('\n'); 219 220 size_t __size = fwrite(__str.data(), 1, __str.size(), __stream); 221 if (__size < __str.size()) { 222 if (std::feof(__stream)) 223 std::__throw_system_error(EIO, "EOF while writing the formatted output"); 224 std::__throw_system_error(std::ferror(__stream), "failed to write formatted output"); 225 } 226} 227 228# ifndef _LIBCPP_HAS_NO_UNICODE 229 230// Note these helper functions are mainly used to aid testing. 231// On POSIX systems and Windows the output is no longer considered a 232// terminal when the output is redirected. Typically during testing the 233// output is redirected to be able to capture it. This makes it hard to 234// test this code path. 235template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 236_LIBCPP_HIDE_FROM_ABI inline void 237__vprint_unicode_posix(FILE* __stream, string_view __fmt, format_args __args, bool __write_nl, bool __is_terminal) { 238 // TODO PRINT Should flush errors throw too? 239 if (__is_terminal) 240 std::fflush(__stream); 241 242 __print::__vprint_nonunicode(__stream, __fmt, __args, __write_nl); 243} 244 245# ifndef _LIBCPP_HAS_NO_WIDE_CHARACTERS 246template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 247_LIBCPP_HIDE_FROM_ABI inline void 248__vprint_unicode_windows(FILE* __stream, string_view __fmt, format_args __args, bool __write_nl, bool __is_terminal) { 249 if (!__is_terminal) 250 return __print::__vprint_nonunicode(__stream, __fmt, __args, __write_nl); 251 252 // TODO PRINT Should flush errors throw too? 253 std::fflush(__stream); 254 255 string __str = std::vformat(__fmt, __args); 256 // UTF-16 uses the same number or less code units than UTF-8. 257 // However the size of the code unit is 16 bits instead of 8 bits. 258 // 259 // The buffer uses the worst-case estimate and should never resize. 260 // However when the string is large this could lead to OOM. Using a 261 // smaller size might work, but since the buffer uses a grow factor 262 // the final size might be larger when the estimate is wrong. 263 // 264 // TODO PRINT profile and improve the speed of this code. 265 __format::__retarget_buffer<wchar_t> __buffer{__str.size()}; 266 __unicode::__transcode(__str.begin(), __str.end(), __buffer.__make_output_iterator()); 267 if (__write_nl) 268 __buffer.push_back(L'\n'); 269 270 [[maybe_unused]] wstring_view __view = __buffer.__view(); 271 272 // The macro _LIBCPP_TESTING_PRINT_WRITE_TO_WINDOWS_CONSOLE_FUNCTION is used to change 273 // the behavior in the test. This is not part of the public API. 274# ifdef _LIBCPP_TESTING_PRINT_WRITE_TO_WINDOWS_CONSOLE_FUNCTION 275 _LIBCPP_TESTING_PRINT_WRITE_TO_WINDOWS_CONSOLE_FUNCTION(__stream, __view); 276# elif defined(_LIBCPP_WIN32API) 277 std::__write_to_windows_console(__stream, __view); 278# else 279 std::__throw_runtime_error("No defintion of _LIBCPP_TESTING_PRINT_WRITE_TO_WINDOWS_CONSOLE_FUNCTION and " 280 "__write_to_windows_console is not available."); 281# endif 282} 283# endif // _LIBCPP_HAS_NO_WIDE_CHARACTERS 284 285template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 286_LIBCPP_HIDE_FROM_ABI inline void 287__vprint_unicode([[maybe_unused]] FILE* __stream, 288 [[maybe_unused]] string_view __fmt, 289 [[maybe_unused]] format_args __args, 290 [[maybe_unused]] bool __write_nl) { 291 _LIBCPP_ASSERT_NON_NULL(__stream, "__stream must be a valid pointer to an output C stream"); 292 293 // [print.fun] 294 // 7 - Effects: If stream refers to a terminal capable of displaying 295 // Unicode, writes out to the terminal using the native Unicode 296 // API; if out contains invalid code units, the behavior is 297 // undefined and implementations are encouraged to diagnose it. 298 // Otherwise writes out to stream unchanged. If the native 299 // Unicode API is used, the function flushes stream before 300 // writing out. 301 // 8 - Throws: Any exception thrown by the call to vformat 302 // ([format.err.report]). system_error if writing to the terminal 303 // or stream fails. May throw bad_alloc. 304 // 9 - Recommended practice: If invoking the native Unicode API 305 // requires transcoding, implementations should substitute 306 // invalid code units with U+FFFD replacement character per the 307 // Unicode Standard, Chapter 3.9 U+FFFD Substitution in 308 // Conversion. 309 310 // On non-Windows platforms the Unicode API is the normal file I/O API 311 // so there the call can be forwarded to the non_unicode API. On 312 // Windows there is a different API. This API requires transcoding. 313 314# ifndef _LIBCPP_WIN32API 315 __print::__vprint_unicode_posix(__stream, __fmt, __args, __write_nl, __print::__is_terminal(__stream)); 316# elif !defined(_LIBCPP_HAS_NO_WIDE_CHARACTERS) 317 __print::__vprint_unicode_windows(__stream, __fmt, __args, __write_nl, __print::__is_terminal(__stream)); 318# else 319# error "Windows builds with wchar_t disabled are not supported." 320# endif 321} 322 323# endif // _LIBCPP_HAS_NO_UNICODE 324 325} // namespace __print 326 327template <class... _Args> 328_LIBCPP_HIDE_FROM_ABI void print(FILE* __stream, format_string<_Args...> __fmt, _Args&&... __args) { 329# ifndef _LIBCPP_HAS_NO_UNICODE 330 if constexpr (__print::__use_unicode_execution_charset) 331 __print::__vprint_unicode(__stream, __fmt.get(), std::make_format_args(__args...), false); 332 else 333 __print::__vprint_nonunicode(__stream, __fmt.get(), std::make_format_args(__args...), false); 334# else // _LIBCPP_HAS_NO_UNICODE 335 __print::__vprint_nonunicode(__stream, __fmt.get(), std::make_format_args(__args...), false); 336# endif // _LIBCPP_HAS_NO_UNICODE 337} 338 339template <class... _Args> 340_LIBCPP_HIDE_FROM_ABI void print(format_string<_Args...> __fmt, _Args&&... __args) { 341 std::print(stdout, __fmt, std::forward<_Args>(__args)...); 342} 343 344template <class... _Args> 345_LIBCPP_HIDE_FROM_ABI void println(FILE* __stream, format_string<_Args...> __fmt, _Args&&... __args) { 346# ifndef _LIBCPP_HAS_NO_UNICODE 347 // Note the wording in the Standard is inefficient. The output of 348 // std::format is a std::string which is then copied. This solution 349 // just appends a newline at the end of the output. 350 if constexpr (__print::__use_unicode_execution_charset) 351 __print::__vprint_unicode(__stream, __fmt.get(), std::make_format_args(__args...), true); 352 else 353 __print::__vprint_nonunicode(__stream, __fmt.get(), std::make_format_args(__args...), true); 354# else // _LIBCPP_HAS_NO_UNICODE 355 __print::__vprint_nonunicode(__stream, __fmt.get(), std::make_format_args(__args...), true); 356# endif // _LIBCPP_HAS_NO_UNICODE 357} 358 359template <class... _Args> 360_LIBCPP_HIDE_FROM_ABI void println(format_string<_Args...> __fmt, _Args&&... __args) { 361 std::println(stdout, __fmt, std::forward<_Args>(__args)...); 362} 363 364# ifndef _LIBCPP_HAS_NO_UNICODE 365template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 366_LIBCPP_HIDE_FROM_ABI inline void vprint_unicode(FILE* __stream, string_view __fmt, format_args __args) { 367 __print::__vprint_unicode(__stream, __fmt, __args, false); 368} 369 370template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 371_LIBCPP_HIDE_FROM_ABI inline void vprint_unicode(string_view __fmt, format_args __args) { 372 std::vprint_unicode(stdout, __fmt, __args); 373} 374 375# endif // _LIBCPP_HAS_NO_UNICODE 376 377template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 378_LIBCPP_HIDE_FROM_ABI inline void vprint_nonunicode(FILE* __stream, string_view __fmt, format_args __args) { 379 __print::__vprint_nonunicode(__stream, __fmt, __args, false); 380} 381 382template <class = void> // TODO PRINT template or availability markup fires too eagerly (http://llvm.org/PR61563). 383_LIBCPP_HIDE_FROM_ABI inline void vprint_nonunicode(string_view __fmt, format_args __args) { 384 std::vprint_nonunicode(stdout, __fmt, __args); 385} 386 387#endif // _LIBCPP_STD_VER >= 23 388 389_LIBCPP_END_NAMESPACE_STD 390 391#endif // _LIBCPP_PRINT 392