master
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// WARNING, this entire header is generated by
11// utils/generate_width_estimation_table.py
12// DO NOT MODIFY!
13
14// UNICODE, INC. LICENSE AGREEMENT - DATA FILES AND SOFTWARE
15//
16// See Terms of Use <https://www.unicode.org/copyright.html>
17// for definitions of Unicode Inc.'s Data Files and Software.
18//
19// NOTICE TO USER: Carefully read the following legal agreement.
20// BY DOWNLOADING, INSTALLING, COPYING OR OTHERWISE USING UNICODE INC.'S
21// DATA FILES ("DATA FILES"), AND/OR SOFTWARE ("SOFTWARE"),
22// YOU UNEQUIVOCALLY ACCEPT, AND AGREE TO BE BOUND BY, ALL OF THE
23// TERMS AND CONDITIONS OF THIS AGREEMENT.
24// IF YOU DO NOT AGREE, DO NOT DOWNLOAD, INSTALL, COPY, DISTRIBUTE OR USE
25// THE DATA FILES OR SOFTWARE.
26//
27// COPYRIGHT AND PERMISSION NOTICE
28//
29// Copyright (c) 1991-2022 Unicode, Inc. All rights reserved.
30// Distributed under the Terms of Use in https://www.unicode.org/copyright.html.
31//
32// Permission is hereby granted, free of charge, to any person obtaining
33// a copy of the Unicode data files and any associated documentation
34// (the "Data Files") or Unicode software and any associated documentation
35// (the "Software") to deal in the Data Files or Software
36// without restriction, including without limitation the rights to use,
37// copy, modify, merge, publish, distribute, and/or sell copies of
38// the Data Files or Software, and to permit persons to whom the Data Files
39// or Software are furnished to do so, provided that either
40// (a) this copyright and permission notice appear with all copies
41// of the Data Files or Software, or
42// (b) this copyright and permission notice appear in associated
43// Documentation.
44//
45// THE DATA FILES AND SOFTWARE ARE PROVIDED "AS IS", WITHOUT WARRANTY OF
46// ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
47// WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
48// NONINFRINGEMENT OF THIRD PARTY RIGHTS.
49// IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS
50// NOTICE BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL
51// DAMAGES, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
52// DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
53// TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
54// PERFORMANCE OF THE DATA FILES OR SOFTWARE.
55//
56// Except as contained in this notice, the name of a copyright holder
57// shall not be used in advertising or otherwise to promote the sale,
58// use or other dealings in these Data Files or Software without prior
59// written authorization of the copyright holder.
60
61#ifndef _LIBCPP___FORMAT_WIDTH_ESTIMATION_TABLE_H
62#define _LIBCPP___FORMAT_WIDTH_ESTIMATION_TABLE_H
63
64#include <__algorithm/ranges_upper_bound.h>
65#include <__config>
66#include <__cstddef/ptrdiff_t.h>
67#include <cstdint>
68
69#if !defined(_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER)
70# pragma GCC system_header
71#endif
72
73_LIBCPP_BEGIN_NAMESPACE_STD
74
75#if _LIBCPP_STD_VER >= 20
76
77namespace __width_estimation_table {
78
79/// The entries of the characters with an estimated width of 2.
80///
81/// Contains the entries for [format.string.std]/12
82/// - Any code point with the East_Asian_Width="W" or East_Asian_Width="F"
83/// Derived Extracted Property as described by UAX #44
84/// - U+4DC0 - U+4DFF (Yijing Hexagram Symbols)
85/// - U+1F300 - U+1F5FF (Miscellaneous Symbols and Pictographs)
86/// - U+1F900 - U+1F9FF (Supplemental Symbols and Pictographs)
87///
88/// The data is generated from
89/// - https://www.unicode.org/Public/UCD/latest/ucd/EastAsianWidth.txt
90/// - The "overrides" in [format.string.std]/12
91///
92/// The format of EastAsianWidth.txt is two fields separated by a semicolon.
93/// Field 0: Unicode code point value or range of code point values
94/// Field 1: East_Asian_Width property, consisting of one of the following values:
95/// "A", "F", "H", "N", "Na", "W"
96/// - All code points, assigned or unassigned, that are not listed
97/// explicitly are given the value "N".
98/// - The unassigned code points in the following blocks default to "W":
99/// CJK Unified Ideographs Extension A: U+3400..U+4DBF
100/// CJK Unified Ideographs: U+4E00..U+9FFF
101/// CJK Compatibility Ideographs: U+F900..U+FAFF
102/// - All undesignated code points in Planes 2 and 3, whether inside or
103/// outside of allocated blocks, default to "W":
104/// Plane 2: U+20000..U+2FFFD
105/// Plane 3: U+30000..U+3FFFD
106///
107/// The table is similar to the table
108/// __extended_grapheme_custer_property_boundary::__entries
109/// which explains the details of these classes. The only difference is this
110/// table lacks a property, thus having more bits available for the size.
111///
112/// The maximum code point that has an estimated width of 2 is U+3FFFD. This
113/// value can be encoded in 18 bits. Thus the upper 3 bits of the code point
114/// are always 0. These 3 bits are used to enlarge the offset range. This
115/// optimization reduces the table in Unicode 15 from 184 to 104 entries,
116/// saving 320 bytes.
117///
118/// The data has 2 values:
119/// - bits [0, 13] The size of the range, allowing 16384 elements.
120/// - bits [14, 31] The lower bound code point of the range. The upper bound of
121/// the range is lower bound + size.
122_LIBCPP_HIDE_FROM_ABI inline constexpr uint32_t __entries[110] = {
123 0x0440005f /* 00001100 - 0000115f [ 96] */, //
124 0x08c68001 /* 0000231a - 0000231b [ 2] */, //
125 0x08ca4001 /* 00002329 - 0000232a [ 2] */, //
126 0x08fa4003 /* 000023e9 - 000023ec [ 4] */, //
127 0x08fc0000 /* 000023f0 - 000023f0 [ 1] */, //
128 0x08fcc000 /* 000023f3 - 000023f3 [ 1] */, //
129 0x097f4001 /* 000025fd - 000025fe [ 2] */, //
130 0x09850001 /* 00002614 - 00002615 [ 2] */, //
131 0x098c0007 /* 00002630 - 00002637 [ 8] */, //
132 0x0992000b /* 00002648 - 00002653 [ 12] */, //
133 0x099fc000 /* 0000267f - 0000267f [ 1] */, //
134 0x09a28005 /* 0000268a - 0000268f [ 6] */, //
135 0x09a4c000 /* 00002693 - 00002693 [ 1] */, //
136 0x09a84000 /* 000026a1 - 000026a1 [ 1] */, //
137 0x09aa8001 /* 000026aa - 000026ab [ 2] */, //
138 0x09af4001 /* 000026bd - 000026be [ 2] */, //
139 0x09b10001 /* 000026c4 - 000026c5 [ 2] */, //
140 0x09b38000 /* 000026ce - 000026ce [ 1] */, //
141 0x09b50000 /* 000026d4 - 000026d4 [ 1] */, //
142 0x09ba8000 /* 000026ea - 000026ea [ 1] */, //
143 0x09bc8001 /* 000026f2 - 000026f3 [ 2] */, //
144 0x09bd4000 /* 000026f5 - 000026f5 [ 1] */, //
145 0x09be8000 /* 000026fa - 000026fa [ 1] */, //
146 0x09bf4000 /* 000026fd - 000026fd [ 1] */, //
147 0x09c14000 /* 00002705 - 00002705 [ 1] */, //
148 0x09c28001 /* 0000270a - 0000270b [ 2] */, //
149 0x09ca0000 /* 00002728 - 00002728 [ 1] */, //
150 0x09d30000 /* 0000274c - 0000274c [ 1] */, //
151 0x09d38000 /* 0000274e - 0000274e [ 1] */, //
152 0x09d4c002 /* 00002753 - 00002755 [ 3] */, //
153 0x09d5c000 /* 00002757 - 00002757 [ 1] */, //
154 0x09e54002 /* 00002795 - 00002797 [ 3] */, //
155 0x09ec0000 /* 000027b0 - 000027b0 [ 1] */, //
156 0x09efc000 /* 000027bf - 000027bf [ 1] */, //
157 0x0ac6c001 /* 00002b1b - 00002b1c [ 2] */, //
158 0x0ad40000 /* 00002b50 - 00002b50 [ 1] */, //
159 0x0ad54000 /* 00002b55 - 00002b55 [ 1] */, //
160 0x0ba00019 /* 00002e80 - 00002e99 [ 26] */, //
161 0x0ba6c058 /* 00002e9b - 00002ef3 [ 89] */, //
162 0x0bc000d5 /* 00002f00 - 00002fd5 [ 214] */, //
163 0x0bfc004e /* 00002ff0 - 0000303e [ 79] */, //
164 0x0c104055 /* 00003041 - 00003096 [ 86] */, //
165 0x0c264066 /* 00003099 - 000030ff [ 103] */, //
166 0x0c41402a /* 00003105 - 0000312f [ 43] */, //
167 0x0c4c405d /* 00003131 - 0000318e [ 94] */, //
168 0x0c640055 /* 00003190 - 000031e5 [ 86] */, //
169 0x0c7bc02f /* 000031ef - 0000321e [ 48] */, //
170 0x0c880027 /* 00003220 - 00003247 [ 40] */, //
171 0x0c943fff /* 00003250 - 0000724f [16384] */, //
172 0x1c94323c /* 00007250 - 0000a48c [12861] */, //
173 0x29240036 /* 0000a490 - 0000a4c6 [ 55] */, //
174 0x2a58001c /* 0000a960 - 0000a97c [ 29] */, //
175 0x2b002ba3 /* 0000ac00 - 0000d7a3 [11172] */, //
176 0x3e4001ff /* 0000f900 - 0000faff [ 512] */, //
177 0x3f840009 /* 0000fe10 - 0000fe19 [ 10] */, //
178 0x3f8c0022 /* 0000fe30 - 0000fe52 [ 35] */, //
179 0x3f950012 /* 0000fe54 - 0000fe66 [ 19] */, //
180 0x3f9a0003 /* 0000fe68 - 0000fe6b [ 4] */, //
181 0x3fc0405f /* 0000ff01 - 0000ff60 [ 96] */, //
182 0x3ff80006 /* 0000ffe0 - 0000ffe6 [ 7] */, //
183 0x5bf80004 /* 00016fe0 - 00016fe4 [ 5] */, //
184 0x5bfc0001 /* 00016ff0 - 00016ff1 [ 2] */, //
185 0x5c0017f7 /* 00017000 - 000187f7 [ 6136] */, //
186 0x620004d5 /* 00018800 - 00018cd5 [ 1238] */, //
187 0x633fc009 /* 00018cff - 00018d08 [ 10] */, //
188 0x6bfc0003 /* 0001aff0 - 0001aff3 [ 4] */, //
189 0x6bfd4006 /* 0001aff5 - 0001affb [ 7] */, //
190 0x6bff4001 /* 0001affd - 0001affe [ 2] */, //
191 0x6c000122 /* 0001b000 - 0001b122 [ 291] */, //
192 0x6c4c8000 /* 0001b132 - 0001b132 [ 1] */, //
193 0x6c540002 /* 0001b150 - 0001b152 [ 3] */, //
194 0x6c554000 /* 0001b155 - 0001b155 [ 1] */, //
195 0x6c590003 /* 0001b164 - 0001b167 [ 4] */, //
196 0x6c5c018b /* 0001b170 - 0001b2fb [ 396] */, //
197 0x74c00056 /* 0001d300 - 0001d356 [ 87] */, //
198 0x74d80016 /* 0001d360 - 0001d376 [ 23] */, //
199 0x7c010000 /* 0001f004 - 0001f004 [ 1] */, //
200 0x7c33c000 /* 0001f0cf - 0001f0cf [ 1] */, //
201 0x7c638000 /* 0001f18e - 0001f18e [ 1] */, //
202 0x7c644009 /* 0001f191 - 0001f19a [ 10] */, //
203 0x7c800002 /* 0001f200 - 0001f202 [ 3] */, //
204 0x7c84002b /* 0001f210 - 0001f23b [ 44] */, //
205 0x7c900008 /* 0001f240 - 0001f248 [ 9] */, //
206 0x7c940001 /* 0001f250 - 0001f251 [ 2] */, //
207 0x7c980005 /* 0001f260 - 0001f265 [ 6] */, //
208 0x7cc0034f /* 0001f300 - 0001f64f [ 848] */, //
209 0x7da00045 /* 0001f680 - 0001f6c5 [ 70] */, //
210 0x7db30000 /* 0001f6cc - 0001f6cc [ 1] */, //
211 0x7db40002 /* 0001f6d0 - 0001f6d2 [ 3] */, //
212 0x7db54002 /* 0001f6d5 - 0001f6d7 [ 3] */, //
213 0x7db70003 /* 0001f6dc - 0001f6df [ 4] */, //
214 0x7dbac001 /* 0001f6eb - 0001f6ec [ 2] */, //
215 0x7dbd0008 /* 0001f6f4 - 0001f6fc [ 9] */, //
216 0x7df8000b /* 0001f7e0 - 0001f7eb [ 12] */, //
217 0x7dfc0000 /* 0001f7f0 - 0001f7f0 [ 1] */, //
218 0x7e4000ff /* 0001f900 - 0001f9ff [ 256] */, //
219 0x7e9c000c /* 0001fa70 - 0001fa7c [ 13] */, //
220 0x7ea00009 /* 0001fa80 - 0001fa89 [ 10] */, //
221 0x7ea3c037 /* 0001fa8f - 0001fac6 [ 56] */, //
222 0x7eb3800e /* 0001face - 0001fadc [ 15] */, //
223 0x7eb7c00a /* 0001fadf - 0001fae9 [ 11] */, //
224 0x7ebc0008 /* 0001faf0 - 0001faf8 [ 9] */, //
225 0x80003fff /* 00020000 - 00023fff [16384] */, //
226 0x90003fff /* 00024000 - 00027fff [16384] */, //
227 0xa0003fff /* 00028000 - 0002bfff [16384] */, //
228 0xb0003ffd /* 0002c000 - 0002fffd [16382] */, //
229 0xc0003fff /* 00030000 - 00033fff [16384] */, //
230 0xd0003fff /* 00034000 - 00037fff [16384] */, //
231 0xe0003fff /* 00038000 - 0003bfff [16384] */, //
232 0xf0003ffd /* 0003c000 - 0003fffd [16382] */};
233
234/// The upper bound entry of EastAsianWidth.txt.
235///
236/// Values greater than this value may have more than 18 significant bits.
237/// They always have a width of 1. This property makes it possible to store
238/// the table in its compact form.
239inline constexpr uint32_t __table_upper_bound = 0x0003fffd;
240
241/// Returns the estimated width of a Unicode code point.
242///
243/// \\pre The code point is a valid Unicode code point.
244[[nodiscard]] _LIBCPP_HIDE_FROM_ABI constexpr int __estimated_width(const char32_t __code_point) noexcept {
245 // Since __table_upper_bound contains the unshifted range do the
246 // comparison without shifting.
247 if (__code_point > __table_upper_bound) [[unlikely]]
248 return 1;
249
250 // When the code-point is less than the first element in the table
251 // the lookup is quite expensive. Since quite some scripts are in
252 // that range, it makes sense to validate that first.
253 // The std_format_spec_string_unicode benchmark gives a measurable
254 // improvement.
255 if (__code_point < (__entries[0] >> 14))
256 return 1;
257
258 ptrdiff_t __i = std::ranges::upper_bound(__entries, (__code_point << 14) | 0x3fffu) - __entries;
259 if (__i == 0)
260 return 1;
261
262 --__i;
263 uint32_t __upper_bound = (__entries[__i] >> 14) + (__entries[__i] & 0x3fffu);
264 return 1 + (__code_point <= __upper_bound);
265}
266
267} // namespace __width_estimation_table
268
269#endif // _LIBCPP_STD_VER >= 20
270
271_LIBCPP_END_NAMESPACE_STD
272
273#endif // _LIBCPP___FORMAT_WIDTH_ESTIMATION_TABLE_H