From c0aa88ae4f1c2eadf369aad17dda34666cb912f8 Mon Sep 17 00:00:00 2001 From: Danil Alexeev Date: Thu, 22 Feb 2024 12:18:45 +0300 Subject: [PATCH] Add Unicode support to `String.to_*_case()` methods --- core/string/char_range.inc | 1314 ++++++++++++++++++++++++++++++- core/string/char_utils.h | 43 +- core/string/ustring.cpp | 10 +- tests/core/string/test_string.h | 65 +- 4 files changed, 1387 insertions(+), 45 deletions(-) diff --git a/core/string/char_range.inc b/core/string/char_range.inc index 5dffe4f20d2..b7d6bbdb611 100644 --- a/core/string/char_range.inc +++ b/core/string/char_range.inc @@ -1,5 +1,5 @@ /**************************************************************************/ -/* color_names.inc */ +/* char_range.inc */ /**************************************************************************/ /* This file is part of: */ /* GODOT ENGINE */ @@ -689,7 +689,6 @@ inline constexpr CharRange xid_start[] = { { 0x2ceb0, 0x2ebe0 }, { 0x2f800, 0x2fa1d }, { 0x30000, 0x3134a }, - { 0x0, 0x0 }, }; inline constexpr CharRange xid_continue[] = { @@ -1450,7 +1449,1316 @@ inline constexpr CharRange xid_continue[] = { { 0x2f800, 0x2fa1d }, { 0x30000, 0x3134a }, { 0xe0100, 0xe01ef }, - { 0x0, 0x0 }, +}; + +inline constexpr CharRange uppercase_letter[] = { + { 0x41, 0x5a }, + { 0xc0, 0xd6 }, + { 0xd8, 0xde }, + { 0x100, 0x100 }, + { 0x102, 0x102 }, + { 0x104, 0x104 }, + { 0x106, 0x106 }, + { 0x108, 0x108 }, + { 0x10a, 0x10a }, + { 0x10c, 0x10c }, + { 0x10e, 0x10e }, + { 0x110, 0x110 }, + { 0x112, 0x112 }, + { 0x114, 0x114 }, + { 0x116, 0x116 }, + { 0x118, 0x118 }, + { 0x11a, 0x11a }, + { 0x11c, 0x11c }, + { 0x11e, 0x11e }, + { 0x120, 0x120 }, + { 0x122, 0x122 }, + { 0x124, 0x124 }, + { 0x126, 0x126 }, + { 0x128, 0x128 }, + { 0x12a, 0x12a }, + { 0x12c, 0x12c }, + { 0x12e, 0x12e }, + { 0x130, 0x130 }, + { 0x132, 0x132 }, + { 0x134, 0x134 }, + { 0x136, 0x136 }, + { 0x139, 0x139 }, + { 0x13b, 0x13b }, + { 0x13d, 0x13d }, + { 0x13f, 0x13f }, + { 0x141, 0x141 }, + { 0x143, 0x143 }, + { 0x145, 0x145 }, + { 0x147, 0x147 }, + { 0x14a, 0x14a }, + { 0x14c, 0x14c }, + { 0x14e, 0x14e }, + { 0x150, 0x150 }, + { 0x152, 0x152 }, + { 0x154, 0x154 }, + { 0x156, 0x156 }, + { 0x158, 0x158 }, + { 0x15a, 0x15a }, + { 0x15c, 0x15c }, + { 0x15e, 0x15e }, + { 0x160, 0x160 }, + { 0x162, 0x162 }, + { 0x164, 0x164 }, + { 0x166, 0x166 }, + { 0x168, 0x168 }, + { 0x16a, 0x16a }, + { 0x16c, 0x16c }, + { 0x16e, 0x16e }, + { 0x170, 0x170 }, + { 0x172, 0x172 }, + { 0x174, 0x174 }, + { 0x176, 0x176 }, + { 0x178, 0x179 }, + { 0x17b, 0x17b }, + { 0x17d, 0x17d }, + { 0x181, 0x182 }, + { 0x184, 0x184 }, + { 0x186, 0x187 }, + { 0x189, 0x18b }, + { 0x18e, 0x191 }, + { 0x193, 0x194 }, + { 0x196, 0x198 }, + { 0x19c, 0x19d }, + { 0x19f, 0x1a0 }, + { 0x1a2, 0x1a2 }, + { 0x1a4, 0x1a4 }, + { 0x1a6, 0x1a7 }, + { 0x1a9, 0x1a9 }, + { 0x1ac, 0x1ac }, + { 0x1ae, 0x1af }, + { 0x1b1, 0x1b3 }, + { 0x1b5, 0x1b5 }, + { 0x1b7, 0x1b8 }, + { 0x1bc, 0x1bc }, + { 0x1c4, 0x1c4 }, + { 0x1c7, 0x1c7 }, + { 0x1ca, 0x1ca }, + { 0x1cd, 0x1cd }, + { 0x1cf, 0x1cf }, + { 0x1d1, 0x1d1 }, + { 0x1d3, 0x1d3 }, + { 0x1d5, 0x1d5 }, + { 0x1d7, 0x1d7 }, + { 0x1d9, 0x1d9 }, + { 0x1db, 0x1db }, + { 0x1de, 0x1de }, + { 0x1e0, 0x1e0 }, + { 0x1e2, 0x1e2 }, + { 0x1e4, 0x1e4 }, + { 0x1e6, 0x1e6 }, + { 0x1e8, 0x1e8 }, + { 0x1ea, 0x1ea }, + { 0x1ec, 0x1ec }, + { 0x1ee, 0x1ee }, + { 0x1f1, 0x1f1 }, + { 0x1f4, 0x1f4 }, + { 0x1f6, 0x1f8 }, + { 0x1fa, 0x1fa }, + { 0x1fc, 0x1fc }, + { 0x1fe, 0x1fe }, + { 0x200, 0x200 }, + { 0x202, 0x202 }, + { 0x204, 0x204 }, + { 0x206, 0x206 }, + { 0x208, 0x208 }, + { 0x20a, 0x20a }, + { 0x20c, 0x20c }, + { 0x20e, 0x20e }, + { 0x210, 0x210 }, + { 0x212, 0x212 }, + { 0x214, 0x214 }, + { 0x216, 0x216 }, + { 0x218, 0x218 }, + { 0x21a, 0x21a }, + { 0x21c, 0x21c }, + { 0x21e, 0x21e }, + { 0x220, 0x220 }, + { 0x222, 0x222 }, + { 0x224, 0x224 }, + { 0x226, 0x226 }, + { 0x228, 0x228 }, + { 0x22a, 0x22a }, + { 0x22c, 0x22c }, + { 0x22e, 0x22e }, + { 0x230, 0x230 }, + { 0x232, 0x232 }, + { 0x23a, 0x23b }, + { 0x23d, 0x23e }, + { 0x241, 0x241 }, + { 0x243, 0x246 }, + { 0x248, 0x248 }, + { 0x24a, 0x24a }, + { 0x24c, 0x24c }, + { 0x24e, 0x24e }, + { 0x370, 0x370 }, + { 0x372, 0x372 }, + { 0x376, 0x376 }, + { 0x37f, 0x37f }, + { 0x386, 0x386 }, + { 0x388, 0x38a }, + { 0x38c, 0x38c }, + { 0x38e, 0x38f }, + { 0x391, 0x3a1 }, + { 0x3a3, 0x3ab }, + { 0x3cf, 0x3cf }, + { 0x3d2, 0x3d4 }, + { 0x3d8, 0x3d8 }, + { 0x3da, 0x3da }, + { 0x3dc, 0x3dc }, + { 0x3de, 0x3de }, + { 0x3e0, 0x3e0 }, + { 0x3e2, 0x3e2 }, + { 0x3e4, 0x3e4 }, + { 0x3e6, 0x3e6 }, + { 0x3e8, 0x3e8 }, + { 0x3ea, 0x3ea }, + { 0x3ec, 0x3ec }, + { 0x3ee, 0x3ee }, + { 0x3f4, 0x3f4 }, + { 0x3f7, 0x3f7 }, + { 0x3f9, 0x3fa }, + { 0x3fd, 0x42f }, + { 0x460, 0x460 }, + { 0x462, 0x462 }, + { 0x464, 0x464 }, + { 0x466, 0x466 }, + { 0x468, 0x468 }, + { 0x46a, 0x46a }, + { 0x46c, 0x46c }, + { 0x46e, 0x46e }, + { 0x470, 0x470 }, + { 0x472, 0x472 }, + { 0x474, 0x474 }, + { 0x476, 0x476 }, + { 0x478, 0x478 }, + { 0x47a, 0x47a }, + { 0x47c, 0x47c }, + { 0x47e, 0x47e }, + { 0x480, 0x480 }, + { 0x48a, 0x48a }, + { 0x48c, 0x48c }, + { 0x48e, 0x48e }, + { 0x490, 0x490 }, + { 0x492, 0x492 }, + { 0x494, 0x494 }, + { 0x496, 0x496 }, + { 0x498, 0x498 }, + { 0x49a, 0x49a }, + { 0x49c, 0x49c }, + { 0x49e, 0x49e }, + { 0x4a0, 0x4a0 }, + { 0x4a2, 0x4a2 }, + { 0x4a4, 0x4a4 }, + { 0x4a6, 0x4a6 }, + { 0x4a8, 0x4a8 }, + { 0x4aa, 0x4aa }, + { 0x4ac, 0x4ac }, + { 0x4ae, 0x4ae }, + { 0x4b0, 0x4b0 }, + { 0x4b2, 0x4b2 }, + { 0x4b4, 0x4b4 }, + { 0x4b6, 0x4b6 }, + { 0x4b8, 0x4b8 }, + { 0x4ba, 0x4ba }, + { 0x4bc, 0x4bc }, + { 0x4be, 0x4be }, + { 0x4c0, 0x4c1 }, + { 0x4c3, 0x4c3 }, + { 0x4c5, 0x4c5 }, + { 0x4c7, 0x4c7 }, + { 0x4c9, 0x4c9 }, + { 0x4cb, 0x4cb }, + { 0x4cd, 0x4cd }, + { 0x4d0, 0x4d0 }, + { 0x4d2, 0x4d2 }, + { 0x4d4, 0x4d4 }, + { 0x4d6, 0x4d6 }, + { 0x4d8, 0x4d8 }, + { 0x4da, 0x4da }, + { 0x4dc, 0x4dc }, + { 0x4de, 0x4de }, + { 0x4e0, 0x4e0 }, + { 0x4e2, 0x4e2 }, + { 0x4e4, 0x4e4 }, + { 0x4e6, 0x4e6 }, + { 0x4e8, 0x4e8 }, + { 0x4ea, 0x4ea }, + { 0x4ec, 0x4ec }, + { 0x4ee, 0x4ee }, + { 0x4f0, 0x4f0 }, + { 0x4f2, 0x4f2 }, + { 0x4f4, 0x4f4 }, + { 0x4f6, 0x4f6 }, + { 0x4f8, 0x4f8 }, + { 0x4fa, 0x4fa }, + { 0x4fc, 0x4fc }, + { 0x4fe, 0x4fe }, + { 0x500, 0x500 }, + { 0x502, 0x502 }, + { 0x504, 0x504 }, + { 0x506, 0x506 }, + { 0x508, 0x508 }, + { 0x50a, 0x50a }, + { 0x50c, 0x50c }, + { 0x50e, 0x50e }, + { 0x510, 0x510 }, + { 0x512, 0x512 }, + { 0x514, 0x514 }, + { 0x516, 0x516 }, + { 0x518, 0x518 }, + { 0x51a, 0x51a }, + { 0x51c, 0x51c }, + { 0x51e, 0x51e }, + { 0x520, 0x520 }, + { 0x522, 0x522 }, + { 0x524, 0x524 }, + { 0x526, 0x526 }, + { 0x528, 0x528 }, + { 0x52a, 0x52a }, + { 0x52c, 0x52c }, + { 0x52e, 0x52e }, + { 0x531, 0x556 }, + { 0x10a0, 0x10c5 }, + { 0x10c7, 0x10c7 }, + { 0x10cd, 0x10cd }, + { 0x13a0, 0x13f5 }, + { 0x1c90, 0x1cba }, + { 0x1cbd, 0x1cbf }, + { 0x1e00, 0x1e00 }, + { 0x1e02, 0x1e02 }, + { 0x1e04, 0x1e04 }, + { 0x1e06, 0x1e06 }, + { 0x1e08, 0x1e08 }, + { 0x1e0a, 0x1e0a }, + { 0x1e0c, 0x1e0c }, + { 0x1e0e, 0x1e0e }, + { 0x1e10, 0x1e10 }, + { 0x1e12, 0x1e12 }, + { 0x1e14, 0x1e14 }, + { 0x1e16, 0x1e16 }, + { 0x1e18, 0x1e18 }, + { 0x1e1a, 0x1e1a }, + { 0x1e1c, 0x1e1c }, + { 0x1e1e, 0x1e1e }, + { 0x1e20, 0x1e20 }, + { 0x1e22, 0x1e22 }, + { 0x1e24, 0x1e24 }, + { 0x1e26, 0x1e26 }, + { 0x1e28, 0x1e28 }, + { 0x1e2a, 0x1e2a }, + { 0x1e2c, 0x1e2c }, + { 0x1e2e, 0x1e2e }, + { 0x1e30, 0x1e30 }, + { 0x1e32, 0x1e32 }, + { 0x1e34, 0x1e34 }, + { 0x1e36, 0x1e36 }, + { 0x1e38, 0x1e38 }, + { 0x1e3a, 0x1e3a }, + { 0x1e3c, 0x1e3c }, + { 0x1e3e, 0x1e3e }, + { 0x1e40, 0x1e40 }, + { 0x1e42, 0x1e42 }, + { 0x1e44, 0x1e44 }, + { 0x1e46, 0x1e46 }, + { 0x1e48, 0x1e48 }, + { 0x1e4a, 0x1e4a }, + { 0x1e4c, 0x1e4c }, + { 0x1e4e, 0x1e4e }, + { 0x1e50, 0x1e50 }, + { 0x1e52, 0x1e52 }, + { 0x1e54, 0x1e54 }, + { 0x1e56, 0x1e56 }, + { 0x1e58, 0x1e58 }, + { 0x1e5a, 0x1e5a }, + { 0x1e5c, 0x1e5c }, + { 0x1e5e, 0x1e5e }, + { 0x1e60, 0x1e60 }, + { 0x1e62, 0x1e62 }, + { 0x1e64, 0x1e64 }, + { 0x1e66, 0x1e66 }, + { 0x1e68, 0x1e68 }, + { 0x1e6a, 0x1e6a }, + { 0x1e6c, 0x1e6c }, + { 0x1e6e, 0x1e6e }, + { 0x1e70, 0x1e70 }, + { 0x1e72, 0x1e72 }, + { 0x1e74, 0x1e74 }, + { 0x1e76, 0x1e76 }, + { 0x1e78, 0x1e78 }, + { 0x1e7a, 0x1e7a }, + { 0x1e7c, 0x1e7c }, + { 0x1e7e, 0x1e7e }, + { 0x1e80, 0x1e80 }, + { 0x1e82, 0x1e82 }, + { 0x1e84, 0x1e84 }, + { 0x1e86, 0x1e86 }, + { 0x1e88, 0x1e88 }, + { 0x1e8a, 0x1e8a }, + { 0x1e8c, 0x1e8c }, + { 0x1e8e, 0x1e8e }, + { 0x1e90, 0x1e90 }, + { 0x1e92, 0x1e92 }, + { 0x1e94, 0x1e94 }, + { 0x1e9e, 0x1e9e }, + { 0x1ea0, 0x1ea0 }, + { 0x1ea2, 0x1ea2 }, + { 0x1ea4, 0x1ea4 }, + { 0x1ea6, 0x1ea6 }, + { 0x1ea8, 0x1ea8 }, + { 0x1eaa, 0x1eaa }, + { 0x1eac, 0x1eac }, + { 0x1eae, 0x1eae }, + { 0x1eb0, 0x1eb0 }, + { 0x1eb2, 0x1eb2 }, + { 0x1eb4, 0x1eb4 }, + { 0x1eb6, 0x1eb6 }, + { 0x1eb8, 0x1eb8 }, + { 0x1eba, 0x1eba }, + { 0x1ebc, 0x1ebc }, + { 0x1ebe, 0x1ebe }, + { 0x1ec0, 0x1ec0 }, + { 0x1ec2, 0x1ec2 }, + { 0x1ec4, 0x1ec4 }, + { 0x1ec6, 0x1ec6 }, + { 0x1ec8, 0x1ec8 }, + { 0x1eca, 0x1eca }, + { 0x1ecc, 0x1ecc }, + { 0x1ece, 0x1ece }, + { 0x1ed0, 0x1ed0 }, + { 0x1ed2, 0x1ed2 }, + { 0x1ed4, 0x1ed4 }, + { 0x1ed6, 0x1ed6 }, + { 0x1ed8, 0x1ed8 }, + { 0x1eda, 0x1eda }, + { 0x1edc, 0x1edc }, + { 0x1ede, 0x1ede }, + { 0x1ee0, 0x1ee0 }, + { 0x1ee2, 0x1ee2 }, + { 0x1ee4, 0x1ee4 }, + { 0x1ee6, 0x1ee6 }, + { 0x1ee8, 0x1ee8 }, + { 0x1eea, 0x1eea }, + { 0x1eec, 0x1eec }, + { 0x1eee, 0x1eee }, + { 0x1ef0, 0x1ef0 }, + { 0x1ef2, 0x1ef2 }, + { 0x1ef4, 0x1ef4 }, + { 0x1ef6, 0x1ef6 }, + { 0x1ef8, 0x1ef8 }, + { 0x1efa, 0x1efa }, + { 0x1efc, 0x1efc }, + { 0x1efe, 0x1efe }, + { 0x1f08, 0x1f0f }, + { 0x1f18, 0x1f1d }, + { 0x1f28, 0x1f2f }, + { 0x1f38, 0x1f3f }, + { 0x1f48, 0x1f4d }, + { 0x1f59, 0x1f59 }, + { 0x1f5b, 0x1f5b }, + { 0x1f5d, 0x1f5d }, + { 0x1f5f, 0x1f5f }, + { 0x1f68, 0x1f6f }, + { 0x1fb8, 0x1fbb }, + { 0x1fc8, 0x1fcb }, + { 0x1fd8, 0x1fdb }, + { 0x1fe8, 0x1fec }, + { 0x1ff8, 0x1ffb }, + { 0x2102, 0x2102 }, + { 0x2107, 0x2107 }, + { 0x210b, 0x210d }, + { 0x2110, 0x2112 }, + { 0x2115, 0x2115 }, + { 0x2119, 0x211d }, + { 0x2124, 0x2124 }, + { 0x2126, 0x2126 }, + { 0x2128, 0x2128 }, + { 0x212a, 0x212d }, + { 0x2130, 0x2133 }, + { 0x213e, 0x213f }, + { 0x2145, 0x2145 }, + { 0x2183, 0x2183 }, + { 0x2c00, 0x2c2f }, + { 0x2c60, 0x2c60 }, + { 0x2c62, 0x2c64 }, + { 0x2c67, 0x2c67 }, + { 0x2c69, 0x2c69 }, + { 0x2c6b, 0x2c6b }, + { 0x2c6d, 0x2c70 }, + { 0x2c72, 0x2c72 }, + { 0x2c75, 0x2c75 }, + { 0x2c7e, 0x2c80 }, + { 0x2c82, 0x2c82 }, + { 0x2c84, 0x2c84 }, + { 0x2c86, 0x2c86 }, + { 0x2c88, 0x2c88 }, + { 0x2c8a, 0x2c8a }, + { 0x2c8c, 0x2c8c }, + { 0x2c8e, 0x2c8e }, + { 0x2c90, 0x2c90 }, + { 0x2c92, 0x2c92 }, + { 0x2c94, 0x2c94 }, + { 0x2c96, 0x2c96 }, + { 0x2c98, 0x2c98 }, + { 0x2c9a, 0x2c9a }, + { 0x2c9c, 0x2c9c }, + { 0x2c9e, 0x2c9e }, + { 0x2ca0, 0x2ca0 }, + { 0x2ca2, 0x2ca2 }, + { 0x2ca4, 0x2ca4 }, + { 0x2ca6, 0x2ca6 }, + { 0x2ca8, 0x2ca8 }, + { 0x2caa, 0x2caa }, + { 0x2cac, 0x2cac }, + { 0x2cae, 0x2cae }, + { 0x2cb0, 0x2cb0 }, + { 0x2cb2, 0x2cb2 }, + { 0x2cb4, 0x2cb4 }, + { 0x2cb6, 0x2cb6 }, + { 0x2cb8, 0x2cb8 }, + { 0x2cba, 0x2cba }, + { 0x2cbc, 0x2cbc }, + { 0x2cbe, 0x2cbe }, + { 0x2cc0, 0x2cc0 }, + { 0x2cc2, 0x2cc2 }, + { 0x2cc4, 0x2cc4 }, + { 0x2cc6, 0x2cc6 }, + { 0x2cc8, 0x2cc8 }, + { 0x2cca, 0x2cca }, + { 0x2ccc, 0x2ccc }, + { 0x2cce, 0x2cce }, + { 0x2cd0, 0x2cd0 }, + { 0x2cd2, 0x2cd2 }, + { 0x2cd4, 0x2cd4 }, + { 0x2cd6, 0x2cd6 }, + { 0x2cd8, 0x2cd8 }, + { 0x2cda, 0x2cda }, + { 0x2cdc, 0x2cdc }, + { 0x2cde, 0x2cde }, + { 0x2ce0, 0x2ce0 }, + { 0x2ce2, 0x2ce2 }, + { 0x2ceb, 0x2ceb }, + { 0x2ced, 0x2ced }, + { 0x2cf2, 0x2cf2 }, + { 0xa640, 0xa640 }, + { 0xa642, 0xa642 }, + { 0xa644, 0xa644 }, + { 0xa646, 0xa646 }, + { 0xa648, 0xa648 }, + { 0xa64a, 0xa64a }, + { 0xa64c, 0xa64c }, + { 0xa64e, 0xa64e }, + { 0xa650, 0xa650 }, + { 0xa652, 0xa652 }, + { 0xa654, 0xa654 }, + { 0xa656, 0xa656 }, + { 0xa658, 0xa658 }, + { 0xa65a, 0xa65a }, + { 0xa65c, 0xa65c }, + { 0xa65e, 0xa65e }, + { 0xa660, 0xa660 }, + { 0xa662, 0xa662 }, + { 0xa664, 0xa664 }, + { 0xa666, 0xa666 }, + { 0xa668, 0xa668 }, + { 0xa66a, 0xa66a }, + { 0xa66c, 0xa66c }, + { 0xa680, 0xa680 }, + { 0xa682, 0xa682 }, + { 0xa684, 0xa684 }, + { 0xa686, 0xa686 }, + { 0xa688, 0xa688 }, + { 0xa68a, 0xa68a }, + { 0xa68c, 0xa68c }, + { 0xa68e, 0xa68e }, + { 0xa690, 0xa690 }, + { 0xa692, 0xa692 }, + { 0xa694, 0xa694 }, + { 0xa696, 0xa696 }, + { 0xa698, 0xa698 }, + { 0xa69a, 0xa69a }, + { 0xa722, 0xa722 }, + { 0xa724, 0xa724 }, + { 0xa726, 0xa726 }, + { 0xa728, 0xa728 }, + { 0xa72a, 0xa72a }, + { 0xa72c, 0xa72c }, + { 0xa72e, 0xa72e }, + { 0xa732, 0xa732 }, + { 0xa734, 0xa734 }, + { 0xa736, 0xa736 }, + { 0xa738, 0xa738 }, + { 0xa73a, 0xa73a }, + { 0xa73c, 0xa73c }, + { 0xa73e, 0xa73e }, + { 0xa740, 0xa740 }, + { 0xa742, 0xa742 }, + { 0xa744, 0xa744 }, + { 0xa746, 0xa746 }, + { 0xa748, 0xa748 }, + { 0xa74a, 0xa74a }, + { 0xa74c, 0xa74c }, + { 0xa74e, 0xa74e }, + { 0xa750, 0xa750 }, + { 0xa752, 0xa752 }, + { 0xa754, 0xa754 }, + { 0xa756, 0xa756 }, + { 0xa758, 0xa758 }, + { 0xa75a, 0xa75a }, + { 0xa75c, 0xa75c }, + { 0xa75e, 0xa75e }, + { 0xa760, 0xa760 }, + { 0xa762, 0xa762 }, + { 0xa764, 0xa764 }, + { 0xa766, 0xa766 }, + { 0xa768, 0xa768 }, + { 0xa76a, 0xa76a }, + { 0xa76c, 0xa76c }, + { 0xa76e, 0xa76e }, + { 0xa779, 0xa779 }, + { 0xa77b, 0xa77b }, + { 0xa77d, 0xa77e }, + { 0xa780, 0xa780 }, + { 0xa782, 0xa782 }, + { 0xa784, 0xa784 }, + { 0xa786, 0xa786 }, + { 0xa78b, 0xa78b }, + { 0xa78d, 0xa78d }, + { 0xa790, 0xa790 }, + { 0xa792, 0xa792 }, + { 0xa796, 0xa796 }, + { 0xa798, 0xa798 }, + { 0xa79a, 0xa79a }, + { 0xa79c, 0xa79c }, + { 0xa79e, 0xa79e }, + { 0xa7a0, 0xa7a0 }, + { 0xa7a2, 0xa7a2 }, + { 0xa7a4, 0xa7a4 }, + { 0xa7a6, 0xa7a6 }, + { 0xa7a8, 0xa7a8 }, + { 0xa7aa, 0xa7ae }, + { 0xa7b0, 0xa7b4 }, + { 0xa7b6, 0xa7b6 }, + { 0xa7b8, 0xa7b8 }, + { 0xa7ba, 0xa7ba }, + { 0xa7bc, 0xa7bc }, + { 0xa7be, 0xa7be }, + { 0xa7c0, 0xa7c0 }, + { 0xa7c2, 0xa7c2 }, + { 0xa7c4, 0xa7c7 }, + { 0xa7c9, 0xa7c9 }, + { 0xa7d0, 0xa7d0 }, + { 0xa7d6, 0xa7d6 }, + { 0xa7d8, 0xa7d8 }, + { 0xa7f5, 0xa7f5 }, + { 0xff21, 0xff3a }, + { 0x10400, 0x10427 }, + { 0x104b0, 0x104d3 }, + { 0x10570, 0x1057a }, + { 0x1057c, 0x1058a }, + { 0x1058c, 0x10592 }, + { 0x10594, 0x10595 }, + { 0x10c80, 0x10cb2 }, + { 0x118a0, 0x118bf }, + { 0x16e40, 0x16e5f }, + { 0x1d400, 0x1d419 }, + { 0x1d434, 0x1d44d }, + { 0x1d468, 0x1d481 }, + { 0x1d49c, 0x1d49c }, + { 0x1d49e, 0x1d49f }, + { 0x1d4a2, 0x1d4a2 }, + { 0x1d4a5, 0x1d4a6 }, + { 0x1d4a9, 0x1d4ac }, + { 0x1d4ae, 0x1d4b5 }, + { 0x1d4d0, 0x1d4e9 }, + { 0x1d504, 0x1d505 }, + { 0x1d507, 0x1d50a }, + { 0x1d50d, 0x1d514 }, + { 0x1d516, 0x1d51c }, + { 0x1d538, 0x1d539 }, + { 0x1d53b, 0x1d53e }, + { 0x1d540, 0x1d544 }, + { 0x1d546, 0x1d546 }, + { 0x1d54a, 0x1d550 }, + { 0x1d56c, 0x1d585 }, + { 0x1d5a0, 0x1d5b9 }, + { 0x1d5d4, 0x1d5ed }, + { 0x1d608, 0x1d621 }, + { 0x1d63c, 0x1d655 }, + { 0x1d670, 0x1d689 }, + { 0x1d6a8, 0x1d6c0 }, + { 0x1d6e2, 0x1d6fa }, + { 0x1d71c, 0x1d734 }, + { 0x1d756, 0x1d76e }, + { 0x1d790, 0x1d7a8 }, + { 0x1d7ca, 0x1d7ca }, + { 0x1e900, 0x1e921 }, +}; + +inline constexpr CharRange lowercase_letter[] = { + { 0x61, 0x7a }, + { 0xb5, 0xb5 }, + { 0xdf, 0xf6 }, + { 0xf8, 0xff }, + { 0x101, 0x101 }, + { 0x103, 0x103 }, + { 0x105, 0x105 }, + { 0x107, 0x107 }, + { 0x109, 0x109 }, + { 0x10b, 0x10b }, + { 0x10d, 0x10d }, + { 0x10f, 0x10f }, + { 0x111, 0x111 }, + { 0x113, 0x113 }, + { 0x115, 0x115 }, + { 0x117, 0x117 }, + { 0x119, 0x119 }, + { 0x11b, 0x11b }, + { 0x11d, 0x11d }, + { 0x11f, 0x11f }, + { 0x121, 0x121 }, + { 0x123, 0x123 }, + { 0x125, 0x125 }, + { 0x127, 0x127 }, + { 0x129, 0x129 }, + { 0x12b, 0x12b }, + { 0x12d, 0x12d }, + { 0x12f, 0x12f }, + { 0x131, 0x131 }, + { 0x133, 0x133 }, + { 0x135, 0x135 }, + { 0x137, 0x138 }, + { 0x13a, 0x13a }, + { 0x13c, 0x13c }, + { 0x13e, 0x13e }, + { 0x140, 0x140 }, + { 0x142, 0x142 }, + { 0x144, 0x144 }, + { 0x146, 0x146 }, + { 0x148, 0x149 }, + { 0x14b, 0x14b }, + { 0x14d, 0x14d }, + { 0x14f, 0x14f }, + { 0x151, 0x151 }, + { 0x153, 0x153 }, + { 0x155, 0x155 }, + { 0x157, 0x157 }, + { 0x159, 0x159 }, + { 0x15b, 0x15b }, + { 0x15d, 0x15d }, + { 0x15f, 0x15f }, + { 0x161, 0x161 }, + { 0x163, 0x163 }, + { 0x165, 0x165 }, + { 0x167, 0x167 }, + { 0x169, 0x169 }, + { 0x16b, 0x16b }, + { 0x16d, 0x16d }, + { 0x16f, 0x16f }, + { 0x171, 0x171 }, + { 0x173, 0x173 }, + { 0x175, 0x175 }, + { 0x177, 0x177 }, + { 0x17a, 0x17a }, + { 0x17c, 0x17c }, + { 0x17e, 0x180 }, + { 0x183, 0x183 }, + { 0x185, 0x185 }, + { 0x188, 0x188 }, + { 0x18c, 0x18d }, + { 0x192, 0x192 }, + { 0x195, 0x195 }, + { 0x199, 0x19b }, + { 0x19e, 0x19e }, + { 0x1a1, 0x1a1 }, + { 0x1a3, 0x1a3 }, + { 0x1a5, 0x1a5 }, + { 0x1a8, 0x1a8 }, + { 0x1aa, 0x1ab }, + { 0x1ad, 0x1ad }, + { 0x1b0, 0x1b0 }, + { 0x1b4, 0x1b4 }, + { 0x1b6, 0x1b6 }, + { 0x1b9, 0x1ba }, + { 0x1bd, 0x1bf }, + { 0x1c6, 0x1c6 }, + { 0x1c9, 0x1c9 }, + { 0x1cc, 0x1cc }, + { 0x1ce, 0x1ce }, + { 0x1d0, 0x1d0 }, + { 0x1d2, 0x1d2 }, + { 0x1d4, 0x1d4 }, + { 0x1d6, 0x1d6 }, + { 0x1d8, 0x1d8 }, + { 0x1da, 0x1da }, + { 0x1dc, 0x1dd }, + { 0x1df, 0x1df }, + { 0x1e1, 0x1e1 }, + { 0x1e3, 0x1e3 }, + { 0x1e5, 0x1e5 }, + { 0x1e7, 0x1e7 }, + { 0x1e9, 0x1e9 }, + { 0x1eb, 0x1eb }, + { 0x1ed, 0x1ed }, + { 0x1ef, 0x1f0 }, + { 0x1f3, 0x1f3 }, + { 0x1f5, 0x1f5 }, + { 0x1f9, 0x1f9 }, + { 0x1fb, 0x1fb }, + { 0x1fd, 0x1fd }, + { 0x1ff, 0x1ff }, + { 0x201, 0x201 }, + { 0x203, 0x203 }, + { 0x205, 0x205 }, + { 0x207, 0x207 }, + { 0x209, 0x209 }, + { 0x20b, 0x20b }, + { 0x20d, 0x20d }, + { 0x20f, 0x20f }, + { 0x211, 0x211 }, + { 0x213, 0x213 }, + { 0x215, 0x215 }, + { 0x217, 0x217 }, + { 0x219, 0x219 }, + { 0x21b, 0x21b }, + { 0x21d, 0x21d }, + { 0x21f, 0x21f }, + { 0x221, 0x221 }, + { 0x223, 0x223 }, + { 0x225, 0x225 }, + { 0x227, 0x227 }, + { 0x229, 0x229 }, + { 0x22b, 0x22b }, + { 0x22d, 0x22d }, + { 0x22f, 0x22f }, + { 0x231, 0x231 }, + { 0x233, 0x239 }, + { 0x23c, 0x23c }, + { 0x23f, 0x240 }, + { 0x242, 0x242 }, + { 0x247, 0x247 }, + { 0x249, 0x249 }, + { 0x24b, 0x24b }, + { 0x24d, 0x24d }, + { 0x24f, 0x293 }, + { 0x295, 0x2af }, + { 0x371, 0x371 }, + { 0x373, 0x373 }, + { 0x377, 0x377 }, + { 0x37b, 0x37d }, + { 0x390, 0x390 }, + { 0x3ac, 0x3ce }, + { 0x3d0, 0x3d1 }, + { 0x3d5, 0x3d7 }, + { 0x3d9, 0x3d9 }, + { 0x3db, 0x3db }, + { 0x3dd, 0x3dd }, + { 0x3df, 0x3df }, + { 0x3e1, 0x3e1 }, + { 0x3e3, 0x3e3 }, + { 0x3e5, 0x3e5 }, + { 0x3e7, 0x3e7 }, + { 0x3e9, 0x3e9 }, + { 0x3eb, 0x3eb }, + { 0x3ed, 0x3ed }, + { 0x3ef, 0x3f3 }, + { 0x3f5, 0x3f5 }, + { 0x3f8, 0x3f8 }, + { 0x3fb, 0x3fc }, + { 0x430, 0x45f }, + { 0x461, 0x461 }, + { 0x463, 0x463 }, + { 0x465, 0x465 }, + { 0x467, 0x467 }, + { 0x469, 0x469 }, + { 0x46b, 0x46b }, + { 0x46d, 0x46d }, + { 0x46f, 0x46f }, + { 0x471, 0x471 }, + { 0x473, 0x473 }, + { 0x475, 0x475 }, + { 0x477, 0x477 }, + { 0x479, 0x479 }, + { 0x47b, 0x47b }, + { 0x47d, 0x47d }, + { 0x47f, 0x47f }, + { 0x481, 0x481 }, + { 0x48b, 0x48b }, + { 0x48d, 0x48d }, + { 0x48f, 0x48f }, + { 0x491, 0x491 }, + { 0x493, 0x493 }, + { 0x495, 0x495 }, + { 0x497, 0x497 }, + { 0x499, 0x499 }, + { 0x49b, 0x49b }, + { 0x49d, 0x49d }, + { 0x49f, 0x49f }, + { 0x4a1, 0x4a1 }, + { 0x4a3, 0x4a3 }, + { 0x4a5, 0x4a5 }, + { 0x4a7, 0x4a7 }, + { 0x4a9, 0x4a9 }, + { 0x4ab, 0x4ab }, + { 0x4ad, 0x4ad }, + { 0x4af, 0x4af }, + { 0x4b1, 0x4b1 }, + { 0x4b3, 0x4b3 }, + { 0x4b5, 0x4b5 }, + { 0x4b7, 0x4b7 }, + { 0x4b9, 0x4b9 }, + { 0x4bb, 0x4bb }, + { 0x4bd, 0x4bd }, + { 0x4bf, 0x4bf }, + { 0x4c2, 0x4c2 }, + { 0x4c4, 0x4c4 }, + { 0x4c6, 0x4c6 }, + { 0x4c8, 0x4c8 }, + { 0x4ca, 0x4ca }, + { 0x4cc, 0x4cc }, + { 0x4ce, 0x4cf }, + { 0x4d1, 0x4d1 }, + { 0x4d3, 0x4d3 }, + { 0x4d5, 0x4d5 }, + { 0x4d7, 0x4d7 }, + { 0x4d9, 0x4d9 }, + { 0x4db, 0x4db }, + { 0x4dd, 0x4dd }, + { 0x4df, 0x4df }, + { 0x4e1, 0x4e1 }, + { 0x4e3, 0x4e3 }, + { 0x4e5, 0x4e5 }, + { 0x4e7, 0x4e7 }, + { 0x4e9, 0x4e9 }, + { 0x4eb, 0x4eb }, + { 0x4ed, 0x4ed }, + { 0x4ef, 0x4ef }, + { 0x4f1, 0x4f1 }, + { 0x4f3, 0x4f3 }, + { 0x4f5, 0x4f5 }, + { 0x4f7, 0x4f7 }, + { 0x4f9, 0x4f9 }, + { 0x4fb, 0x4fb }, + { 0x4fd, 0x4fd }, + { 0x4ff, 0x4ff }, + { 0x501, 0x501 }, + { 0x503, 0x503 }, + { 0x505, 0x505 }, + { 0x507, 0x507 }, + { 0x509, 0x509 }, + { 0x50b, 0x50b }, + { 0x50d, 0x50d }, + { 0x50f, 0x50f }, + { 0x511, 0x511 }, + { 0x513, 0x513 }, + { 0x515, 0x515 }, + { 0x517, 0x517 }, + { 0x519, 0x519 }, + { 0x51b, 0x51b }, + { 0x51d, 0x51d }, + { 0x51f, 0x51f }, + { 0x521, 0x521 }, + { 0x523, 0x523 }, + { 0x525, 0x525 }, + { 0x527, 0x527 }, + { 0x529, 0x529 }, + { 0x52b, 0x52b }, + { 0x52d, 0x52d }, + { 0x52f, 0x52f }, + { 0x560, 0x588 }, + { 0x10d0, 0x10fa }, + { 0x10fd, 0x10ff }, + { 0x13f8, 0x13fd }, + { 0x1c80, 0x1c88 }, + { 0x1d00, 0x1d2b }, + { 0x1d6b, 0x1d77 }, + { 0x1d79, 0x1d9a }, + { 0x1e01, 0x1e01 }, + { 0x1e03, 0x1e03 }, + { 0x1e05, 0x1e05 }, + { 0x1e07, 0x1e07 }, + { 0x1e09, 0x1e09 }, + { 0x1e0b, 0x1e0b }, + { 0x1e0d, 0x1e0d }, + { 0x1e0f, 0x1e0f }, + { 0x1e11, 0x1e11 }, + { 0x1e13, 0x1e13 }, + { 0x1e15, 0x1e15 }, + { 0x1e17, 0x1e17 }, + { 0x1e19, 0x1e19 }, + { 0x1e1b, 0x1e1b }, + { 0x1e1d, 0x1e1d }, + { 0x1e1f, 0x1e1f }, + { 0x1e21, 0x1e21 }, + { 0x1e23, 0x1e23 }, + { 0x1e25, 0x1e25 }, + { 0x1e27, 0x1e27 }, + { 0x1e29, 0x1e29 }, + { 0x1e2b, 0x1e2b }, + { 0x1e2d, 0x1e2d }, + { 0x1e2f, 0x1e2f }, + { 0x1e31, 0x1e31 }, + { 0x1e33, 0x1e33 }, + { 0x1e35, 0x1e35 }, + { 0x1e37, 0x1e37 }, + { 0x1e39, 0x1e39 }, + { 0x1e3b, 0x1e3b }, + { 0x1e3d, 0x1e3d }, + { 0x1e3f, 0x1e3f }, + { 0x1e41, 0x1e41 }, + { 0x1e43, 0x1e43 }, + { 0x1e45, 0x1e45 }, + { 0x1e47, 0x1e47 }, + { 0x1e49, 0x1e49 }, + { 0x1e4b, 0x1e4b }, + { 0x1e4d, 0x1e4d }, + { 0x1e4f, 0x1e4f }, + { 0x1e51, 0x1e51 }, + { 0x1e53, 0x1e53 }, + { 0x1e55, 0x1e55 }, + { 0x1e57, 0x1e57 }, + { 0x1e59, 0x1e59 }, + { 0x1e5b, 0x1e5b }, + { 0x1e5d, 0x1e5d }, + { 0x1e5f, 0x1e5f }, + { 0x1e61, 0x1e61 }, + { 0x1e63, 0x1e63 }, + { 0x1e65, 0x1e65 }, + { 0x1e67, 0x1e67 }, + { 0x1e69, 0x1e69 }, + { 0x1e6b, 0x1e6b }, + { 0x1e6d, 0x1e6d }, + { 0x1e6f, 0x1e6f }, + { 0x1e71, 0x1e71 }, + { 0x1e73, 0x1e73 }, + { 0x1e75, 0x1e75 }, + { 0x1e77, 0x1e77 }, + { 0x1e79, 0x1e79 }, + { 0x1e7b, 0x1e7b }, + { 0x1e7d, 0x1e7d }, + { 0x1e7f, 0x1e7f }, + { 0x1e81, 0x1e81 }, + { 0x1e83, 0x1e83 }, + { 0x1e85, 0x1e85 }, + { 0x1e87, 0x1e87 }, + { 0x1e89, 0x1e89 }, + { 0x1e8b, 0x1e8b }, + { 0x1e8d, 0x1e8d }, + { 0x1e8f, 0x1e8f }, + { 0x1e91, 0x1e91 }, + { 0x1e93, 0x1e93 }, + { 0x1e95, 0x1e9d }, + { 0x1e9f, 0x1e9f }, + { 0x1ea1, 0x1ea1 }, + { 0x1ea3, 0x1ea3 }, + { 0x1ea5, 0x1ea5 }, + { 0x1ea7, 0x1ea7 }, + { 0x1ea9, 0x1ea9 }, + { 0x1eab, 0x1eab }, + { 0x1ead, 0x1ead }, + { 0x1eaf, 0x1eaf }, + { 0x1eb1, 0x1eb1 }, + { 0x1eb3, 0x1eb3 }, + { 0x1eb5, 0x1eb5 }, + { 0x1eb7, 0x1eb7 }, + { 0x1eb9, 0x1eb9 }, + { 0x1ebb, 0x1ebb }, + { 0x1ebd, 0x1ebd }, + { 0x1ebf, 0x1ebf }, + { 0x1ec1, 0x1ec1 }, + { 0x1ec3, 0x1ec3 }, + { 0x1ec5, 0x1ec5 }, + { 0x1ec7, 0x1ec7 }, + { 0x1ec9, 0x1ec9 }, + { 0x1ecb, 0x1ecb }, + { 0x1ecd, 0x1ecd }, + { 0x1ecf, 0x1ecf }, + { 0x1ed1, 0x1ed1 }, + { 0x1ed3, 0x1ed3 }, + { 0x1ed5, 0x1ed5 }, + { 0x1ed7, 0x1ed7 }, + { 0x1ed9, 0x1ed9 }, + { 0x1edb, 0x1edb }, + { 0x1edd, 0x1edd }, + { 0x1edf, 0x1edf }, + { 0x1ee1, 0x1ee1 }, + { 0x1ee3, 0x1ee3 }, + { 0x1ee5, 0x1ee5 }, + { 0x1ee7, 0x1ee7 }, + { 0x1ee9, 0x1ee9 }, + { 0x1eeb, 0x1eeb }, + { 0x1eed, 0x1eed }, + { 0x1eef, 0x1eef }, + { 0x1ef1, 0x1ef1 }, + { 0x1ef3, 0x1ef3 }, + { 0x1ef5, 0x1ef5 }, + { 0x1ef7, 0x1ef7 }, + { 0x1ef9, 0x1ef9 }, + { 0x1efb, 0x1efb }, + { 0x1efd, 0x1efd }, + { 0x1eff, 0x1f07 }, + { 0x1f10, 0x1f15 }, + { 0x1f20, 0x1f27 }, + { 0x1f30, 0x1f37 }, + { 0x1f40, 0x1f45 }, + { 0x1f50, 0x1f57 }, + { 0x1f60, 0x1f67 }, + { 0x1f70, 0x1f7d }, + { 0x1f80, 0x1f87 }, + { 0x1f90, 0x1f97 }, + { 0x1fa0, 0x1fa7 }, + { 0x1fb0, 0x1fb4 }, + { 0x1fb6, 0x1fb7 }, + { 0x1fbe, 0x1fbe }, + { 0x1fc2, 0x1fc4 }, + { 0x1fc6, 0x1fc7 }, + { 0x1fd0, 0x1fd3 }, + { 0x1fd6, 0x1fd7 }, + { 0x1fe0, 0x1fe7 }, + { 0x1ff2, 0x1ff4 }, + { 0x1ff6, 0x1ff7 }, + { 0x210a, 0x210a }, + { 0x210e, 0x210f }, + { 0x2113, 0x2113 }, + { 0x212f, 0x212f }, + { 0x2134, 0x2134 }, + { 0x2139, 0x2139 }, + { 0x213c, 0x213d }, + { 0x2146, 0x2149 }, + { 0x214e, 0x214e }, + { 0x2184, 0x2184 }, + { 0x2c30, 0x2c5f }, + { 0x2c61, 0x2c61 }, + { 0x2c65, 0x2c66 }, + { 0x2c68, 0x2c68 }, + { 0x2c6a, 0x2c6a }, + { 0x2c6c, 0x2c6c }, + { 0x2c71, 0x2c71 }, + { 0x2c73, 0x2c74 }, + { 0x2c76, 0x2c7b }, + { 0x2c81, 0x2c81 }, + { 0x2c83, 0x2c83 }, + { 0x2c85, 0x2c85 }, + { 0x2c87, 0x2c87 }, + { 0x2c89, 0x2c89 }, + { 0x2c8b, 0x2c8b }, + { 0x2c8d, 0x2c8d }, + { 0x2c8f, 0x2c8f }, + { 0x2c91, 0x2c91 }, + { 0x2c93, 0x2c93 }, + { 0x2c95, 0x2c95 }, + { 0x2c97, 0x2c97 }, + { 0x2c99, 0x2c99 }, + { 0x2c9b, 0x2c9b }, + { 0x2c9d, 0x2c9d }, + { 0x2c9f, 0x2c9f }, + { 0x2ca1, 0x2ca1 }, + { 0x2ca3, 0x2ca3 }, + { 0x2ca5, 0x2ca5 }, + { 0x2ca7, 0x2ca7 }, + { 0x2ca9, 0x2ca9 }, + { 0x2cab, 0x2cab }, + { 0x2cad, 0x2cad }, + { 0x2caf, 0x2caf }, + { 0x2cb1, 0x2cb1 }, + { 0x2cb3, 0x2cb3 }, + { 0x2cb5, 0x2cb5 }, + { 0x2cb7, 0x2cb7 }, + { 0x2cb9, 0x2cb9 }, + { 0x2cbb, 0x2cbb }, + { 0x2cbd, 0x2cbd }, + { 0x2cbf, 0x2cbf }, + { 0x2cc1, 0x2cc1 }, + { 0x2cc3, 0x2cc3 }, + { 0x2cc5, 0x2cc5 }, + { 0x2cc7, 0x2cc7 }, + { 0x2cc9, 0x2cc9 }, + { 0x2ccb, 0x2ccb }, + { 0x2ccd, 0x2ccd }, + { 0x2ccf, 0x2ccf }, + { 0x2cd1, 0x2cd1 }, + { 0x2cd3, 0x2cd3 }, + { 0x2cd5, 0x2cd5 }, + { 0x2cd7, 0x2cd7 }, + { 0x2cd9, 0x2cd9 }, + { 0x2cdb, 0x2cdb }, + { 0x2cdd, 0x2cdd }, + { 0x2cdf, 0x2cdf }, + { 0x2ce1, 0x2ce1 }, + { 0x2ce3, 0x2ce4 }, + { 0x2cec, 0x2cec }, + { 0x2cee, 0x2cee }, + { 0x2cf3, 0x2cf3 }, + { 0x2d00, 0x2d25 }, + { 0x2d27, 0x2d27 }, + { 0x2d2d, 0x2d2d }, + { 0xa641, 0xa641 }, + { 0xa643, 0xa643 }, + { 0xa645, 0xa645 }, + { 0xa647, 0xa647 }, + { 0xa649, 0xa649 }, + { 0xa64b, 0xa64b }, + { 0xa64d, 0xa64d }, + { 0xa64f, 0xa64f }, + { 0xa651, 0xa651 }, + { 0xa653, 0xa653 }, + { 0xa655, 0xa655 }, + { 0xa657, 0xa657 }, + { 0xa659, 0xa659 }, + { 0xa65b, 0xa65b }, + { 0xa65d, 0xa65d }, + { 0xa65f, 0xa65f }, + { 0xa661, 0xa661 }, + { 0xa663, 0xa663 }, + { 0xa665, 0xa665 }, + { 0xa667, 0xa667 }, + { 0xa669, 0xa669 }, + { 0xa66b, 0xa66b }, + { 0xa66d, 0xa66d }, + { 0xa681, 0xa681 }, + { 0xa683, 0xa683 }, + { 0xa685, 0xa685 }, + { 0xa687, 0xa687 }, + { 0xa689, 0xa689 }, + { 0xa68b, 0xa68b }, + { 0xa68d, 0xa68d }, + { 0xa68f, 0xa68f }, + { 0xa691, 0xa691 }, + { 0xa693, 0xa693 }, + { 0xa695, 0xa695 }, + { 0xa697, 0xa697 }, + { 0xa699, 0xa699 }, + { 0xa69b, 0xa69b }, + { 0xa723, 0xa723 }, + { 0xa725, 0xa725 }, + { 0xa727, 0xa727 }, + { 0xa729, 0xa729 }, + { 0xa72b, 0xa72b }, + { 0xa72d, 0xa72d }, + { 0xa72f, 0xa731 }, + { 0xa733, 0xa733 }, + { 0xa735, 0xa735 }, + { 0xa737, 0xa737 }, + { 0xa739, 0xa739 }, + { 0xa73b, 0xa73b }, + { 0xa73d, 0xa73d }, + { 0xa73f, 0xa73f }, + { 0xa741, 0xa741 }, + { 0xa743, 0xa743 }, + { 0xa745, 0xa745 }, + { 0xa747, 0xa747 }, + { 0xa749, 0xa749 }, + { 0xa74b, 0xa74b }, + { 0xa74d, 0xa74d }, + { 0xa74f, 0xa74f }, + { 0xa751, 0xa751 }, + { 0xa753, 0xa753 }, + { 0xa755, 0xa755 }, + { 0xa757, 0xa757 }, + { 0xa759, 0xa759 }, + { 0xa75b, 0xa75b }, + { 0xa75d, 0xa75d }, + { 0xa75f, 0xa75f }, + { 0xa761, 0xa761 }, + { 0xa763, 0xa763 }, + { 0xa765, 0xa765 }, + { 0xa767, 0xa767 }, + { 0xa769, 0xa769 }, + { 0xa76b, 0xa76b }, + { 0xa76d, 0xa76d }, + { 0xa76f, 0xa76f }, + { 0xa771, 0xa778 }, + { 0xa77a, 0xa77a }, + { 0xa77c, 0xa77c }, + { 0xa77f, 0xa77f }, + { 0xa781, 0xa781 }, + { 0xa783, 0xa783 }, + { 0xa785, 0xa785 }, + { 0xa787, 0xa787 }, + { 0xa78c, 0xa78c }, + { 0xa78e, 0xa78e }, + { 0xa791, 0xa791 }, + { 0xa793, 0xa795 }, + { 0xa797, 0xa797 }, + { 0xa799, 0xa799 }, + { 0xa79b, 0xa79b }, + { 0xa79d, 0xa79d }, + { 0xa79f, 0xa79f }, + { 0xa7a1, 0xa7a1 }, + { 0xa7a3, 0xa7a3 }, + { 0xa7a5, 0xa7a5 }, + { 0xa7a7, 0xa7a7 }, + { 0xa7a9, 0xa7a9 }, + { 0xa7af, 0xa7af }, + { 0xa7b5, 0xa7b5 }, + { 0xa7b7, 0xa7b7 }, + { 0xa7b9, 0xa7b9 }, + { 0xa7bb, 0xa7bb }, + { 0xa7bd, 0xa7bd }, + { 0xa7bf, 0xa7bf }, + { 0xa7c1, 0xa7c1 }, + { 0xa7c3, 0xa7c3 }, + { 0xa7c8, 0xa7c8 }, + { 0xa7ca, 0xa7ca }, + { 0xa7d1, 0xa7d1 }, + { 0xa7d3, 0xa7d3 }, + { 0xa7d5, 0xa7d5 }, + { 0xa7d7, 0xa7d7 }, + { 0xa7d9, 0xa7d9 }, + { 0xa7f6, 0xa7f6 }, + { 0xa7fa, 0xa7fa }, + { 0xab30, 0xab5a }, + { 0xab60, 0xab68 }, + { 0xab70, 0xabbf }, + { 0xfb00, 0xfb06 }, + { 0xfb13, 0xfb17 }, + { 0xff41, 0xff5a }, + { 0x10428, 0x1044f }, + { 0x104d8, 0x104fb }, + { 0x10597, 0x105a1 }, + { 0x105a3, 0x105b1 }, + { 0x105b3, 0x105b9 }, + { 0x105bb, 0x105bc }, + { 0x10cc0, 0x10cf2 }, + { 0x118c0, 0x118df }, + { 0x16e60, 0x16e7f }, + { 0x1d41a, 0x1d433 }, + { 0x1d44e, 0x1d454 }, + { 0x1d456, 0x1d467 }, + { 0x1d482, 0x1d49b }, + { 0x1d4b6, 0x1d4b9 }, + { 0x1d4bb, 0x1d4bb }, + { 0x1d4bd, 0x1d4c3 }, + { 0x1d4c5, 0x1d4cf }, + { 0x1d4ea, 0x1d503 }, + { 0x1d51e, 0x1d537 }, + { 0x1d552, 0x1d56b }, + { 0x1d586, 0x1d59f }, + { 0x1d5ba, 0x1d5d3 }, + { 0x1d5ee, 0x1d607 }, + { 0x1d622, 0x1d63b }, + { 0x1d656, 0x1d66f }, + { 0x1d68a, 0x1d6a5 }, + { 0x1d6c2, 0x1d6da }, + { 0x1d6dc, 0x1d6e1 }, + { 0x1d6fc, 0x1d714 }, + { 0x1d716, 0x1d71b }, + { 0x1d736, 0x1d74e }, + { 0x1d750, 0x1d755 }, + { 0x1d770, 0x1d788 }, + { 0x1d78a, 0x1d78f }, + { 0x1d7aa, 0x1d7c2 }, + { 0x1d7c4, 0x1d7c9 }, + { 0x1d7cb, 0x1d7cb }, + { 0x1df00, 0x1df09 }, + { 0x1df0b, 0x1df1e }, + { 0x1df25, 0x1df2a }, + { 0x1e922, 0x1e943 }, }; #endif // CHAR_RANGE_INC diff --git a/core/string/char_utils.h b/core/string/char_utils.h index 4ff8fb73202..aa9bc198ca7 100644 --- a/core/string/char_utils.h +++ b/core/string/char_utils.h @@ -35,24 +35,43 @@ #include "char_range.inc" +#define BSEARCH_CHAR_RANGE(m_array) \ + int low = 0; \ + int high = sizeof(m_array) / sizeof(m_array[0]) - 1; \ + int middle; \ + \ + while (low <= high) { \ + middle = (low + high) / 2; \ + \ + if (c < m_array[middle].start) { \ + high = middle - 1; \ + } else if (c > m_array[middle].end) { \ + low = middle + 1; \ + } else { \ + return true; \ + } \ + } \ + \ + return false + static _FORCE_INLINE_ bool is_unicode_identifier_start(char32_t c) { - for (int i = 0; xid_start[i].start != 0; i++) { - if (c >= xid_start[i].start && c <= xid_start[i].end) { - return true; - } - } - return false; + BSEARCH_CHAR_RANGE(xid_start); } static _FORCE_INLINE_ bool is_unicode_identifier_continue(char32_t c) { - for (int i = 0; xid_continue[i].start != 0; i++) { - if (c >= xid_continue[i].start && c <= xid_continue[i].end) { - return true; - } - } - return false; + BSEARCH_CHAR_RANGE(xid_continue); } +static _FORCE_INLINE_ bool is_unicode_upper_case(char32_t c) { + BSEARCH_CHAR_RANGE(uppercase_letter); +} + +static _FORCE_INLINE_ bool is_unicode_lower_case(char32_t c) { + BSEARCH_CHAR_RANGE(lowercase_letter); +} + +#undef BSEARCH_CHAR_RANGE + static _FORCE_INLINE_ bool is_ascii_upper_case(char32_t c) { return (c >= 'A' && c <= 'Z'); } diff --git a/core/string/ustring.cpp b/core/string/ustring.cpp index 67b0fdee207..540891b3b09 100644 --- a/core/string/ustring.cpp +++ b/core/string/ustring.cpp @@ -1044,17 +1044,17 @@ String String::_camelcase_to_underscore() const { int start_index = 0; for (int i = 1; i < size(); i++) { - bool is_prev_upper = is_ascii_upper_case(cstr[i - 1]); - bool is_prev_lower = is_ascii_lower_case(cstr[i - 1]); + bool is_prev_upper = is_unicode_upper_case(cstr[i - 1]); + bool is_prev_lower = is_unicode_lower_case(cstr[i - 1]); bool is_prev_digit = is_digit(cstr[i - 1]); - bool is_curr_upper = is_ascii_upper_case(cstr[i]); - bool is_curr_lower = is_ascii_lower_case(cstr[i]); + bool is_curr_upper = is_unicode_upper_case(cstr[i]); + bool is_curr_lower = is_unicode_lower_case(cstr[i]); bool is_curr_digit = is_digit(cstr[i]); bool is_next_lower = false; if (i + 1 < size()) { - is_next_lower = is_ascii_lower_case(cstr[i + 1]); + is_next_lower = is_unicode_lower_case(cstr[i + 1]); } const bool cond_a = is_prev_lower && is_curr_upper; // aA diff --git a/tests/core/string/test_string.h b/tests/core/string/test_string.h index 8a11491bb2f..8d2607b8743 100644 --- a/tests/core/string/test_string.h +++ b/tests/core/string/test_string.h @@ -1300,39 +1300,54 @@ TEST_CASE("[String] Capitalize against many strings") { input = "snake_case_function( snake_case_arg )"; output = "Snake Case Function( Snake Case Arg )"; CHECK(input.capitalize() == output); + + input = U"словоСлово_слово слово"; + output = U"Слово Слово Слово Слово"; + CHECK(input.capitalize() == output); + + input = U"λέξηΛέξη_λέξη λέξη"; + output = U"Λέξη Λέξη Λέξη Λέξη"; + CHECK(input.capitalize() == output); + + input = U"բառԲառ_բառ բառ"; + output = U"Բառ Բառ Բառ Բառ"; + CHECK(input.capitalize() == output); } struct StringCasesTestCase { - const char *input; - const char *camel_case; - const char *pascal_case; - const char *snake_case; + const char32_t *input; + const char32_t *camel_case; + const char32_t *pascal_case; + const char32_t *snake_case; }; TEST_CASE("[String] Checking case conversion methods") { StringCasesTestCase test_cases[] = { /* clang-format off */ - { "2D", "2d", "2d", "2d" }, - { "2d", "2d", "2d", "2d" }, - { "2db", "2Db", "2Db", "2_db" }, - { "Vector3", "vector3", "Vector3", "vector_3" }, - { "sha256", "sha256", "Sha256", "sha_256" }, - { "Node2D", "node2d", "Node2d", "node_2d" }, - { "RichTextLabel", "richTextLabel", "RichTextLabel", "rich_text_label" }, - { "HTML5", "html5", "Html5", "html_5" }, - { "Node2DPosition", "node2dPosition", "Node2dPosition", "node_2d_position" }, - { "Number2Digits", "number2Digits", "Number2Digits", "number_2_digits" }, - { "get_property_list", "getPropertyList", "GetPropertyList", "get_property_list" }, - { "get_camera_2d", "getCamera2d", "GetCamera2d", "get_camera_2d" }, - { "_physics_process", "physicsProcess", "PhysicsProcess", "_physics_process" }, - { "bytes2var", "bytes2Var", "Bytes2Var", "bytes_2_var" }, - { "linear2db", "linear2Db", "Linear2Db", "linear_2_db" }, - { "sha256sum", "sha256Sum", "Sha256Sum", "sha_256_sum" }, - { "camelCase", "camelCase", "CamelCase", "camel_case" }, - { "PascalCase", "pascalCase", "PascalCase", "pascal_case" }, - { "snake_case", "snakeCase", "SnakeCase", "snake_case" }, - { "Test TEST test", "testTestTest", "TestTestTest", "test_test_test" }, - { nullptr, nullptr, nullptr, nullptr }, + { U"2D", U"2d", U"2d", U"2d" }, + { U"2d", U"2d", U"2d", U"2d" }, + { U"2db", U"2Db", U"2Db", U"2_db" }, + { U"Vector3", U"vector3", U"Vector3", U"vector_3" }, + { U"sha256", U"sha256", U"Sha256", U"sha_256" }, + { U"Node2D", U"node2d", U"Node2d", U"node_2d" }, + { U"RichTextLabel", U"richTextLabel", U"RichTextLabel", U"rich_text_label" }, + { U"HTML5", U"html5", U"Html5", U"html_5" }, + { U"Node2DPosition", U"node2dPosition", U"Node2dPosition", U"node_2d_position" }, + { U"Number2Digits", U"number2Digits", U"Number2Digits", U"number_2_digits" }, + { U"get_property_list", U"getPropertyList", U"GetPropertyList", U"get_property_list" }, + { U"get_camera_2d", U"getCamera2d", U"GetCamera2d", U"get_camera_2d" }, + { U"_physics_process", U"physicsProcess", U"PhysicsProcess", U"_physics_process" }, + { U"bytes2var", U"bytes2Var", U"Bytes2Var", U"bytes_2_var" }, + { U"linear2db", U"linear2Db", U"Linear2Db", U"linear_2_db" }, + { U"sha256sum", U"sha256Sum", U"Sha256Sum", U"sha_256_sum" }, + { U"camelCase", U"camelCase", U"CamelCase", U"camel_case" }, + { U"PascalCase", U"pascalCase", U"PascalCase", U"pascal_case" }, + { U"snake_case", U"snakeCase", U"SnakeCase", U"snake_case" }, + { U"Test TEST test", U"testTestTest", U"TestTestTest", U"test_test_test" }, + { U"словоСлово_слово слово", U"словоСловоСловоСлово", U"СловоСловоСловоСлово", U"слово_слово_слово_слово" }, + { U"λέξηΛέξη_λέξη λέξη", U"λέξηΛέξηΛέξηΛέξη", U"ΛέξηΛέξηΛέξηΛέξη", U"λέξη_λέξη_λέξη_λέξη" }, + { U"բառԲառ_բառ բառ", U"բառԲառԲառԲառ", U"ԲառԲառԲառԲառ", U"բառ_բառ_բառ_բառ" }, + { nullptr, nullptr, nullptr, nullptr }, /* clang-format on */ };