EIC code displayed by LXR |
|
|
|
|||
File indexing completed on 2026-05-03 08:13:51
0001 // -*- C++ -*- 0002 //===----------------------------------------------------------------------===// 0003 // 0004 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 0005 // See https://llvm.org/LICENSE.txt for license information. 0006 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 0007 // 0008 //===----------------------------------------------------------------------===// 0009 0010 // WARNING, this entire header is generated by 0011 // utils/generate_width_estimation_table.py 0012 // DO NOT MODIFY! 0013 0014 // UNICODE, INC. LICENSE AGREEMENT - DATA FILES AND SOFTWARE 0015 // 0016 // See Terms of Use <https://www.unicode.org/copyright.html> 0017 // for definitions of Unicode Inc.'s Data Files and Software. 0018 // 0019 // NOTICE TO USER: Carefully read the following legal agreement. 0020 // BY DOWNLOADING, INSTALLING, COPYING OR OTHERWISE USING UNICODE INC.'S 0021 // DATA FILES ("DATA FILES"), AND/OR SOFTWARE ("SOFTWARE"), 0022 // YOU UNEQUIVOCALLY ACCEPT, AND AGREE TO BE BOUND BY, ALL OF THE 0023 // TERMS AND CONDITIONS OF THIS AGREEMENT. 0024 // IF YOU DO NOT AGREE, DO NOT DOWNLOAD, INSTALL, COPY, DISTRIBUTE OR USE 0025 // THE DATA FILES OR SOFTWARE. 0026 // 0027 // COPYRIGHT AND PERMISSION NOTICE 0028 // 0029 // Copyright (c) 1991-2022 Unicode, Inc. All rights reserved. 0030 // Distributed under the Terms of Use in https://www.unicode.org/copyright.html. 0031 // 0032 // Permission is hereby granted, free of charge, to any person obtaining 0033 // a copy of the Unicode data files and any associated documentation 0034 // (the "Data Files") or Unicode software and any associated documentation 0035 // (the "Software") to deal in the Data Files or Software 0036 // without restriction, including without limitation the rights to use, 0037 // copy, modify, merge, publish, distribute, and/or sell copies of 0038 // the Data Files or Software, and to permit persons to whom the Data Files 0039 // or Software are furnished to do so, provided that either 0040 // (a) this copyright and permission notice appear with all copies 0041 // of the Data Files or Software, or 0042 // (b) this copyright and permission notice appear in associated 0043 // Documentation. 0044 // 0045 // THE DATA FILES AND SOFTWARE ARE PROVIDED "AS IS", WITHOUT WARRANTY OF 0046 // ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE 0047 // WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND 0048 // NONINFRINGEMENT OF THIRD PARTY RIGHTS. 0049 // IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS 0050 // NOTICE BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL 0051 // DAMAGES, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, 0052 // DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER 0053 // TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR 0054 // PERFORMANCE OF THE DATA FILES OR SOFTWARE. 0055 // 0056 // Except as contained in this notice, the name of a copyright holder 0057 // shall not be used in advertising or otherwise to promote the sale, 0058 // use or other dealings in these Data Files or Software without prior 0059 // written authorization of the copyright holder. 0060 0061 #ifndef _LIBCPP___FORMAT_WIDTH_ESTIMATION_TABLE_H 0062 #define _LIBCPP___FORMAT_WIDTH_ESTIMATION_TABLE_H 0063 0064 #include <__algorithm/ranges_upper_bound.h> 0065 #include <__config> 0066 #include <__cstddef/ptrdiff_t.h> 0067 #include <cstdint> 0068 0069 #if !defined(_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER) 0070 # pragma GCC system_header 0071 #endif 0072 0073 _LIBCPP_BEGIN_NAMESPACE_STD 0074 0075 #if _LIBCPP_STD_VER >= 20 0076 0077 namespace __width_estimation_table { 0078 0079 /// The entries of the characters with an estimated width of 2. 0080 /// 0081 /// Contains the entries for [format.string.std]/12 0082 /// - Any code point with the East_Asian_Width="W" or East_Asian_Width="F" 0083 /// Derived Extracted Property as described by UAX #44 0084 /// - U+4DC0 - U+4DFF (Yijing Hexagram Symbols) 0085 /// - U+1F300 - U+1F5FF (Miscellaneous Symbols and Pictographs) 0086 /// - U+1F900 - U+1F9FF (Supplemental Symbols and Pictographs) 0087 /// 0088 /// The data is generated from 0089 /// - https://www.unicode.org/Public/UCD/latest/ucd/EastAsianWidth.txt 0090 /// - The "overrides" in [format.string.std]/12 0091 /// 0092 /// The format of EastAsianWidth.txt is two fields separated by a semicolon. 0093 /// Field 0: Unicode code point value or range of code point values 0094 /// Field 1: East_Asian_Width property, consisting of one of the following values: 0095 /// "A", "F", "H", "N", "Na", "W" 0096 /// - All code points, assigned or unassigned, that are not listed 0097 /// explicitly are given the value "N". 0098 /// - The unassigned code points in the following blocks default to "W": 0099 /// CJK Unified Ideographs Extension A: U+3400..U+4DBF 0100 /// CJK Unified Ideographs: U+4E00..U+9FFF 0101 /// CJK Compatibility Ideographs: U+F900..U+FAFF 0102 /// - All undesignated code points in Planes 2 and 3, whether inside or 0103 /// outside of allocated blocks, default to "W": 0104 /// Plane 2: U+20000..U+2FFFD 0105 /// Plane 3: U+30000..U+3FFFD 0106 /// 0107 /// The table is similar to the table 0108 /// __extended_grapheme_custer_property_boundary::__entries 0109 /// which explains the details of these classes. The only difference is this 0110 /// table lacks a property, thus having more bits available for the size. 0111 /// 0112 /// The maximum code point that has an estimated width of 2 is U+3FFFD. This 0113 /// value can be encoded in 18 bits. Thus the upper 3 bits of the code point 0114 /// are always 0. These 3 bits are used to enlarge the offset range. This 0115 /// optimization reduces the table in Unicode 15 from 184 to 104 entries, 0116 /// saving 320 bytes. 0117 /// 0118 /// The data has 2 values: 0119 /// - bits [0, 13] The size of the range, allowing 16384 elements. 0120 /// - bits [14, 31] The lower bound code point of the range. The upper bound of 0121 /// the range is lower bound + size. 0122 _LIBCPP_HIDE_FROM_ABI inline constexpr uint32_t __entries[107] = { 0123 0x0440005f /* 00001100 - 0000115f [ 96] */, // 0124 0x08c68001 /* 0000231a - 0000231b [ 2] */, // 0125 0x08ca4001 /* 00002329 - 0000232a [ 2] */, // 0126 0x08fa4003 /* 000023e9 - 000023ec [ 4] */, // 0127 0x08fc0000 /* 000023f0 - 000023f0 [ 1] */, // 0128 0x08fcc000 /* 000023f3 - 000023f3 [ 1] */, // 0129 0x097f4001 /* 000025fd - 000025fe [ 2] */, // 0130 0x09850001 /* 00002614 - 00002615 [ 2] */, // 0131 0x0992000b /* 00002648 - 00002653 [ 12] */, // 0132 0x099fc000 /* 0000267f - 0000267f [ 1] */, // 0133 0x09a4c000 /* 00002693 - 00002693 [ 1] */, // 0134 0x09a84000 /* 000026a1 - 000026a1 [ 1] */, // 0135 0x09aa8001 /* 000026aa - 000026ab [ 2] */, // 0136 0x09af4001 /* 000026bd - 000026be [ 2] */, // 0137 0x09b10001 /* 000026c4 - 000026c5 [ 2] */, // 0138 0x09b38000 /* 000026ce - 000026ce [ 1] */, // 0139 0x09b50000 /* 000026d4 - 000026d4 [ 1] */, // 0140 0x09ba8000 /* 000026ea - 000026ea [ 1] */, // 0141 0x09bc8001 /* 000026f2 - 000026f3 [ 2] */, // 0142 0x09bd4000 /* 000026f5 - 000026f5 [ 1] */, // 0143 0x09be8000 /* 000026fa - 000026fa [ 1] */, // 0144 0x09bf4000 /* 000026fd - 000026fd [ 1] */, // 0145 0x09c14000 /* 00002705 - 00002705 [ 1] */, // 0146 0x09c28001 /* 0000270a - 0000270b [ 2] */, // 0147 0x09ca0000 /* 00002728 - 00002728 [ 1] */, // 0148 0x09d30000 /* 0000274c - 0000274c [ 1] */, // 0149 0x09d38000 /* 0000274e - 0000274e [ 1] */, // 0150 0x09d4c002 /* 00002753 - 00002755 [ 3] */, // 0151 0x09d5c000 /* 00002757 - 00002757 [ 1] */, // 0152 0x09e54002 /* 00002795 - 00002797 [ 3] */, // 0153 0x09ec0000 /* 000027b0 - 000027b0 [ 1] */, // 0154 0x09efc000 /* 000027bf - 000027bf [ 1] */, // 0155 0x0ac6c001 /* 00002b1b - 00002b1c [ 2] */, // 0156 0x0ad40000 /* 00002b50 - 00002b50 [ 1] */, // 0157 0x0ad54000 /* 00002b55 - 00002b55 [ 1] */, // 0158 0x0ba00019 /* 00002e80 - 00002e99 [ 26] */, // 0159 0x0ba6c058 /* 00002e9b - 00002ef3 [ 89] */, // 0160 0x0bc000d5 /* 00002f00 - 00002fd5 [ 214] */, // 0161 0x0bfc004e /* 00002ff0 - 0000303e [ 79] */, // 0162 0x0c104055 /* 00003041 - 00003096 [ 86] */, // 0163 0x0c264066 /* 00003099 - 000030ff [ 103] */, // 0164 0x0c41402a /* 00003105 - 0000312f [ 43] */, // 0165 0x0c4c405d /* 00003131 - 0000318e [ 94] */, // 0166 0x0c640053 /* 00003190 - 000031e3 [ 84] */, // 0167 0x0c7bc02f /* 000031ef - 0000321e [ 48] */, // 0168 0x0c880027 /* 00003220 - 00003247 [ 40] */, // 0169 0x0c943fff /* 00003250 - 0000724f [16384] */, // 0170 0x1c94323c /* 00007250 - 0000a48c [12861] */, // 0171 0x29240036 /* 0000a490 - 0000a4c6 [ 55] */, // 0172 0x2a58001c /* 0000a960 - 0000a97c [ 29] */, // 0173 0x2b002ba3 /* 0000ac00 - 0000d7a3 [11172] */, // 0174 0x3e4001ff /* 0000f900 - 0000faff [ 512] */, // 0175 0x3f840009 /* 0000fe10 - 0000fe19 [ 10] */, // 0176 0x3f8c0022 /* 0000fe30 - 0000fe52 [ 35] */, // 0177 0x3f950012 /* 0000fe54 - 0000fe66 [ 19] */, // 0178 0x3f9a0003 /* 0000fe68 - 0000fe6b [ 4] */, // 0179 0x3fc0405f /* 0000ff01 - 0000ff60 [ 96] */, // 0180 0x3ff80006 /* 0000ffe0 - 0000ffe6 [ 7] */, // 0181 0x5bf80004 /* 00016fe0 - 00016fe4 [ 5] */, // 0182 0x5bfc0001 /* 00016ff0 - 00016ff1 [ 2] */, // 0183 0x5c0017f7 /* 00017000 - 000187f7 [ 6136] */, // 0184 0x620004d5 /* 00018800 - 00018cd5 [ 1238] */, // 0185 0x63400008 /* 00018d00 - 00018d08 [ 9] */, // 0186 0x6bfc0003 /* 0001aff0 - 0001aff3 [ 4] */, // 0187 0x6bfd4006 /* 0001aff5 - 0001affb [ 7] */, // 0188 0x6bff4001 /* 0001affd - 0001affe [ 2] */, // 0189 0x6c000122 /* 0001b000 - 0001b122 [ 291] */, // 0190 0x6c4c8000 /* 0001b132 - 0001b132 [ 1] */, // 0191 0x6c540002 /* 0001b150 - 0001b152 [ 3] */, // 0192 0x6c554000 /* 0001b155 - 0001b155 [ 1] */, // 0193 0x6c590003 /* 0001b164 - 0001b167 [ 4] */, // 0194 0x6c5c018b /* 0001b170 - 0001b2fb [ 396] */, // 0195 0x7c010000 /* 0001f004 - 0001f004 [ 1] */, // 0196 0x7c33c000 /* 0001f0cf - 0001f0cf [ 1] */, // 0197 0x7c638000 /* 0001f18e - 0001f18e [ 1] */, // 0198 0x7c644009 /* 0001f191 - 0001f19a [ 10] */, // 0199 0x7c800002 /* 0001f200 - 0001f202 [ 3] */, // 0200 0x7c84002b /* 0001f210 - 0001f23b [ 44] */, // 0201 0x7c900008 /* 0001f240 - 0001f248 [ 9] */, // 0202 0x7c940001 /* 0001f250 - 0001f251 [ 2] */, // 0203 0x7c980005 /* 0001f260 - 0001f265 [ 6] */, // 0204 0x7cc0034f /* 0001f300 - 0001f64f [ 848] */, // 0205 0x7da00045 /* 0001f680 - 0001f6c5 [ 70] */, // 0206 0x7db30000 /* 0001f6cc - 0001f6cc [ 1] */, // 0207 0x7db40002 /* 0001f6d0 - 0001f6d2 [ 3] */, // 0208 0x7db54002 /* 0001f6d5 - 0001f6d7 [ 3] */, // 0209 0x7db70003 /* 0001f6dc - 0001f6df [ 4] */, // 0210 0x7dbac001 /* 0001f6eb - 0001f6ec [ 2] */, // 0211 0x7dbd0008 /* 0001f6f4 - 0001f6fc [ 9] */, // 0212 0x7df8000b /* 0001f7e0 - 0001f7eb [ 12] */, // 0213 0x7dfc0000 /* 0001f7f0 - 0001f7f0 [ 1] */, // 0214 0x7e4000ff /* 0001f900 - 0001f9ff [ 256] */, // 0215 0x7e9c000c /* 0001fa70 - 0001fa7c [ 13] */, // 0216 0x7ea00008 /* 0001fa80 - 0001fa88 [ 9] */, // 0217 0x7ea4002d /* 0001fa90 - 0001fabd [ 46] */, // 0218 0x7eafc006 /* 0001fabf - 0001fac5 [ 7] */, // 0219 0x7eb3800d /* 0001face - 0001fadb [ 14] */, // 0220 0x7eb80008 /* 0001fae0 - 0001fae8 [ 9] */, // 0221 0x7ebc0008 /* 0001faf0 - 0001faf8 [ 9] */, // 0222 0x80003fff /* 00020000 - 00023fff [16384] */, // 0223 0x90003fff /* 00024000 - 00027fff [16384] */, // 0224 0xa0003fff /* 00028000 - 0002bfff [16384] */, // 0225 0xb0003ffd /* 0002c000 - 0002fffd [16382] */, // 0226 0xc0003fff /* 00030000 - 00033fff [16384] */, // 0227 0xd0003fff /* 00034000 - 00037fff [16384] */, // 0228 0xe0003fff /* 00038000 - 0003bfff [16384] */, // 0229 0xf0003ffd /* 0003c000 - 0003fffd [16382] */}; 0230 0231 /// The upper bound entry of EastAsianWidth.txt. 0232 /// 0233 /// Values greater than this value may have more than 18 significant bits. 0234 /// They always have a width of 1. This property makes it possible to store 0235 /// the table in its compact form. 0236 inline constexpr uint32_t __table_upper_bound = 0x0003fffd; 0237 0238 /// Returns the estimated width of a Unicode code point. 0239 /// 0240 /// \\pre The code point is a valid Unicode code point. 0241 [[nodiscard]] _LIBCPP_HIDE_FROM_ABI constexpr int __estimated_width(const char32_t __code_point) noexcept { 0242 // Since __table_upper_bound contains the unshifted range do the 0243 // comparison without shifting. 0244 if (__code_point > __table_upper_bound) [[unlikely]] 0245 return 1; 0246 0247 // When the code-point is less than the first element in the table 0248 // the lookup is quite expensive. Since quite some scripts are in 0249 // that range, it makes sense to validate that first. 0250 // The std_format_spec_string_unicode benchmark gives a measurable 0251 // improvement. 0252 if (__code_point < (__entries[0] >> 14)) 0253 return 1; 0254 0255 ptrdiff_t __i = std::ranges::upper_bound(__entries, (__code_point << 14) | 0x3fffu) - __entries; 0256 if (__i == 0) 0257 return 1; 0258 0259 --__i; 0260 uint32_t __upper_bound = (__entries[__i] >> 14) + (__entries[__i] & 0x3fffu); 0261 return 1 + (__code_point <= __upper_bound); 0262 } 0263 0264 } // namespace __width_estimation_table 0265 0266 #endif // _LIBCPP_STD_VER >= 20 0267 0268 _LIBCPP_END_NAMESPACE_STD 0269 0270 #endif // _LIBCPP___FORMAT_WIDTH_ESTIMATION_TABLE_H
| [ Source navigation ] | [ Diff markup ] | [ Identifier search ] | [ general search ] |
|
This page was automatically generated by the 2.3.7 LXR engine. The LXR team |
|