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___FORMAT_WRITE_ESCAPED_H
11#define _LIBCPP___FORMAT_WRITE_ESCAPED_H
12
13#include <__algorithm/ranges_copy.h>
14#include <__algorithm/ranges_for_each.h>
15#include <__charconv/to_chars_integral.h>
16#include <__charconv/to_chars_result.h>
17#include <__chrono/statically_widen.h>
18#include <__format/escaped_output_table.h>
19#include <__format/formatter_output.h>
20#include <__format/parser_std_format_spec.h>
21#include <__format/unicode.h>
22#include <__iterator/back_insert_iterator.h>
23#include <__memory/addressof.h>
24#include <__system_error/errc.h>
25#include <__type_traits/make_unsigned.h>
26#include <__utility/move.h>
27#include <string_view>
28
29#if !defined(_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER)
30# pragma GCC system_header
31#endif
32
33_LIBCPP_PUSH_MACROS
34#include <__undef_macros>
35
36_LIBCPP_BEGIN_NAMESPACE_STD
37
38namespace __formatter {
39
40#if _LIBCPP_STD_VER >= 20
41
42/// Writes a string using format's width estimation algorithm.
43///
44/// \note When \c _LIBCPP_HAS_NO_UNICODE is defined the function assumes the
45/// input is ASCII.
46template <class _CharT>
47_LIBCPP_HIDE_FROM_ABI auto
48__write_string(basic_string_view<_CharT> __str,
49 output_iterator<const _CharT&> auto __out_it,
50 __format_spec::__parsed_specifications<_CharT> __specs) -> decltype(__out_it) {
51 if (!__specs.__has_precision())
52 return __formatter::__write_string_no_precision(__str, std::move(__out_it), __specs);
53
54 int __size = __formatter::__truncate(__str, __specs.__precision_);
55
56 return __formatter::__write(__str.begin(), __str.end(), std::move(__out_it), __specs, __size);
57}
58
59#endif // _LIBCPP_STD_VER >= 20
60#if _LIBCPP_STD_VER >= 23
61
62struct __nul_terminator {};
63
64template <class _CharT>
65_LIBCPP_HIDE_FROM_ABI bool operator==(const _CharT* __cstr, __nul_terminator) {
66 return *__cstr == _CharT('\0');
67}
68
69template <class _CharT>
70_LIBCPP_HIDE_FROM_ABI void
71__write_escaped_code_unit(basic_string<_CharT>& __str, char32_t __value, const _CharT* __prefix) {
72 back_insert_iterator __out_it{__str};
73 std::ranges::copy(__prefix, __nul_terminator{}, __out_it);
74
75 char __buffer[8];
76 to_chars_result __r = std::to_chars(first: std::begin(array&: __buffer), last: std::end(array&: __buffer), __value, base: 16);
77 _LIBCPP_ASSERT_INTERNAL(__r.ec == errc(0), "Internal buffer too small");
78 std::ranges::copy(std::begin(array&: __buffer), __r.ptr, __out_it);
79
80 __str += _CharT('}');
81}
82
83// [format.string.escaped]/2.2.1.2
84// ...
85// then the sequence \u{hex-digit-sequence} is appended to E, where
86// hex-digit-sequence is the shortest hexadecimal representation of C using
87// lower-case hexadecimal digits.
88template <class _CharT>
89_LIBCPP_HIDE_FROM_ABI void __write_well_formed_escaped_code_unit(basic_string<_CharT>& __str, char32_t __value) {
90 __formatter::__write_escaped_code_unit(__str, __value, _LIBCPP_STATICALLY_WIDEN(_CharT, "\\u{"));
91}
92
93// [format.string.escaped]/2.2.3
94// Otherwise (X is a sequence of ill-formed code units), each code unit U is
95// appended to E in order as the sequence \x{hex-digit-sequence}, where
96// hex-digit-sequence is the shortest hexadecimal representation of U using
97// lower-case hexadecimal digits.
98template <class _CharT>
99_LIBCPP_HIDE_FROM_ABI void __write_escape_ill_formed_code_unit(basic_string<_CharT>& __str, char32_t __value) {
100 __formatter::__write_escaped_code_unit(__str, __value, _LIBCPP_STATICALLY_WIDEN(_CharT, "\\x{"));
101}
102
103template <class _CharT>
104[[nodiscard]] _LIBCPP_HIDE_FROM_ABI bool
105__is_escaped_sequence_written(basic_string<_CharT>& __str, bool __last_escaped, char32_t __value) {
106# ifdef _LIBCPP_HAS_NO_UNICODE
107 // For ASCII assume everything above 127 is printable.
108 if (__value > 127)
109 return false;
110# endif
111
112 // [format.string.escaped]/2.2.1.2.1
113 // CE is UTF-8, UTF-16, or UTF-32 and C corresponds to a Unicode scalar
114 // value whose Unicode property General_Category has a value in the groups
115 // Separator (Z) or Other (C), as described by UAX #44 of the Unicode Standard,
116 if (!__escaped_output_table::__needs_escape(code_point: __value))
117 // [format.string.escaped]/2.2.1.2.2
118 // CE is UTF-8, UTF-16, or UTF-32 and C corresponds to a Unicode scalar
119 // value with the Unicode property Grapheme_Extend=Yes as described by UAX
120 // #44 of the Unicode Standard and C is not immediately preceded in S by a
121 // character P appended to E without translation to an escape sequence,
122 if (!__last_escaped || __extended_grapheme_custer_property_boundary::__get_property(code_point: __value) !=
123 __extended_grapheme_custer_property_boundary::__property::__Extend)
124 return false;
125
126 __formatter::__write_well_formed_escaped_code_unit(__str, __value);
127 return true;
128}
129
130template <class _CharT>
131[[nodiscard]] _LIBCPP_HIDE_FROM_ABI constexpr char32_t __to_char32(_CharT __value) {
132 return static_cast<make_unsigned_t<_CharT>>(__value);
133}
134
135enum class __escape_quotation_mark { __apostrophe, __double_quote };
136
137// [format.string.escaped]/2
138template <class _CharT>
139[[nodiscard]] _LIBCPP_HIDE_FROM_ABI bool __is_escaped_sequence_written(
140 basic_string<_CharT>& __str, char32_t __value, bool __last_escaped, __escape_quotation_mark __mark) {
141 // 2.2.1.1 - Mapped character in [tab:format.escape.sequences]
142 switch (__value) {
143 case _CharT('\t'):
144 __str += _LIBCPP_STATICALLY_WIDEN(_CharT, "\\t");
145 return true;
146 case _CharT('\n'):
147 __str += _LIBCPP_STATICALLY_WIDEN(_CharT, "\\n");
148 return true;
149 case _CharT('\r'):
150 __str += _LIBCPP_STATICALLY_WIDEN(_CharT, "\\r");
151 return true;
152 case _CharT('\''):
153 if (__mark == __escape_quotation_mark::__apostrophe)
154 __str += _LIBCPP_STATICALLY_WIDEN(_CharT, R"(\')");
155 else
156 __str += __value;
157 return true;
158 case _CharT('"'):
159 if (__mark == __escape_quotation_mark::__double_quote)
160 __str += _LIBCPP_STATICALLY_WIDEN(_CharT, R"(\")");
161 else
162 __str += __value;
163 return true;
164 case _CharT('\\'):
165 __str += _LIBCPP_STATICALLY_WIDEN(_CharT, R"(\\)");
166 return true;
167
168 // 2.2.1.2 - Space
169 case _CharT(' '):
170 __str += __value;
171 return true;
172 }
173
174 // 2.2.2
175 // Otherwise, if X is a shift sequence, the effect on E and further
176 // decoding of S is unspecified.
177 // For now shift sequences are ignored and treated as Unicode. Other parts
178 // of the format library do the same. It's unknown how ostream treats them.
179 // TODO FMT determine what to do with shift sequences.
180
181 // 2.2.1.2.1 and 2.2.1.2.2 - Escape
182 return __formatter::__is_escaped_sequence_written(__str, __last_escaped, __formatter::__to_char32(__value));
183}
184
185template <class _CharT>
186_LIBCPP_HIDE_FROM_ABI void
187__escape(basic_string<_CharT>& __str, basic_string_view<_CharT> __values, __escape_quotation_mark __mark) {
188 __unicode::__code_point_view<_CharT> __view{__values.begin(), __values.end()};
189
190 // When the first code unit has the property Grapheme_Extend=Yes it needs to
191 // be escaped. This happens when the previous code unit was also escaped.
192 bool __escape = true;
193 while (!__view.__at_end()) {
194 auto __first = __view.__position();
195 typename __unicode::__consume_result __result = __view.__consume();
196 if (__result.__status == __unicode::__consume_result::__ok) {
197 __escape = __formatter::__is_escaped_sequence_written(__str, __result.__code_point, __escape, __mark);
198 if (!__escape)
199 // 2.2.1.3 - Add the character
200 ranges::copy(__first, __view.__position(), std::back_insert_iterator(__str));
201 } else {
202 // 2.2.3 sequence of ill-formed code units
203 ranges::for_each(__first, __view.__position(), [&](_CharT __value) {
204 __formatter::__write_escape_ill_formed_code_unit(__str, __formatter::__to_char32(__value));
205 });
206 }
207 }
208}
209
210template <class _CharT>
211_LIBCPP_HIDE_FROM_ABI auto
212__format_escaped_char(_CharT __value,
213 output_iterator<const _CharT&> auto __out_it,
214 __format_spec::__parsed_specifications<_CharT> __specs) -> decltype(__out_it) {
215 basic_string<_CharT> __str;
216 __str += _CharT('\'');
217 __formatter::__escape(__str, basic_string_view{std::addressof(__value), 1}, __escape_quotation_mark::__apostrophe);
218 __str += _CharT('\'');
219 return __formatter::__write(__str.data(), __str.data() + __str.size(), std::move(__out_it), __specs, __str.size());
220}
221
222template <class _CharT>
223_LIBCPP_HIDE_FROM_ABI auto
224__format_escaped_string(basic_string_view<_CharT> __values,
225 output_iterator<const _CharT&> auto __out_it,
226 __format_spec::__parsed_specifications<_CharT> __specs) -> decltype(__out_it) {
227 basic_string<_CharT> __str;
228 __str += _CharT('"');
229 __formatter::__escape(__str, __values, __escape_quotation_mark::__double_quote);
230 __str += _CharT('"');
231 return __formatter::__write_string(basic_string_view{__str}, std::move(__out_it), __specs);
232}
233
234#endif // _LIBCPP_STD_VER >= 23
235
236} // namespace __formatter
237
238_LIBCPP_END_NAMESPACE_STD
239
240_LIBCPP_POP_MACROS
241
242#endif // _LIBCPP___FORMAT_WRITE_ESCAPED_H
243