From 29b56ef4e4113bcd091b19d6926f18814162ca53 Mon Sep 17 00:00:00 2001 From: Nicholas Noll Date: Tue, 26 Oct 2021 21:01:41 -0700 Subject: Feat(libunicode): Added an explicit unicode library Refactored code to pull out utf8 functions from base into a standalone library. Also left the required function inside arg.c so that code that calls ARG_BEGIN doesn't have to link to libunicode. --- sys/base/.generated/utf8.c | 391 --------------------------------------------- sys/base/arg.c | 70 ++++++++ sys/base/bufio.c | 2 + sys/base/genutf8.py | 140 ---------------- sys/base/os.c | 2 +- sys/base/string.c | 174 -------------------- 6 files changed, 73 insertions(+), 706 deletions(-) delete mode 100644 sys/base/.generated/utf8.c delete mode 100755 sys/base/genutf8.py (limited to 'sys/base') diff --git a/sys/base/.generated/utf8.c b/sys/base/.generated/utf8.c deleted file mode 100644 index e101e1a..0000000 --- a/sys/base/.generated/utf8.c +++ /dev/null @@ -1,391 +0,0 @@ -rune* -rbsearch(rune c, rune* t, int n, int nelem) -{ - rune* p; - int m; - - while (n > 1) { - m = n >> 1; - p = t + m * nelem; - if (c >= p[0]) { - t = p; - n -= m; - } else - n = m; - } - - if (n && c >= t[0]) - return t; - else - return 0; -} -static rune isspace_rtab[] = { - 0x0009, 0x000c, 0x2000, 0x200a, 0x2028, 0x2029, -}; - -static rune isspace_stab[] = { - 0x0020, 0x0085, 0x00a0, 0x1680, 0x202f, 0x205f, 0x3000, -}; - -int -utf8·isspace(rune c) -{ - rune* p; - - p = rbsearch(c, isspace_rtab, arrlen(isspace_rtab) / 2, 2); - if (p && c >= p[0] && c <= p[1]) return 1; - p = rbsearch(c, isspace_stab, arrlen(isspace_stab), 1); - if (p && c == p[0]) return 1; - - return 0; -} - -static rune islower_rtab[] = { - 0x0061, 0x007a, 0x00df, 0x00f6, 0x00f8, 0x00ff, 0x0137, 0x0138, 0x0148, 0x0149, - 0x017e, 0x0180, 0x018c, 0x018d, 0x0199, 0x019b, 0x01aa, 0x01ab, 0x01b9, 0x01ba, - 0x01bd, 0x01bf, 0x01dc, 0x01dd, 0x01ef, 0x01f0, 0x0233, 0x0239, 0x023f, 0x0240, - 0x024f, 0x0293, 0x0295, 0x02af, 0x037b, 0x037d, 0x03ac, 0x03ce, 0x03d0, 0x03d1, - 0x03d5, 0x03d7, 0x03ef, 0x03f3, 0x03fb, 0x03fc, 0x0430, 0x045f, 0x04ce, 0x04cf, - 0x0560, 0x0588, 0x10d0, 0x10fa, 0x10fd, 0x10ff, 0x13f8, 0x13fd, 0x1c80, 0x1c88, - 0x1d00, 0x1d2b, 0x1d6b, 0x1d77, 0x1d79, 0x1d9a, 0x1e95, 0x1e9d, 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, 0x1fc2, 0x1fc4, 0x1fc6, 0x1fc7, 0x1fd0, 0x1fd3, - 0x1fd6, 0x1fd7, 0x1fe0, 0x1fe7, 0x1ff2, 0x1ff4, 0x1ff6, 0x1ff7, 0x210e, 0x210f, - 0x213c, 0x213d, 0x2146, 0x2149, 0x2c30, 0x2c5e, 0x2c65, 0x2c66, 0x2c73, 0x2c74, - 0x2c76, 0x2c7b, 0x2ce3, 0x2ce4, 0x2d00, 0x2d25, 0xa72f, 0xa731, 0xa771, 0xa778, - 0xa793, 0xa795, 0xab30, 0xab5a, 0xab60, 0xab65, 0xab70, 0xabbf, 0xfb00, 0xfb06, - 0xfb13, 0xfb17, 0xff41, 0xff5a, 0x10428, 0x1044f, 0x104d8, 0x104fb, 0x10cc0, 0x10cf2, - 0x118c0, 0x118df, 0x16e60, 0x16e7f, 0x1d41a, 0x1d433, 0x1d44e, 0x1d454, 0x1d456, 0x1d467, - 0x1d482, 0x1d49b, 0x1d4b6, 0x1d4b9, 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, -}; - -static rune islower_stab[] = { - 0x00b5, 0x0101, 0x0103, 0x0105, 0x0107, 0x0109, 0x010b, 0x010d, 0x010f, 0x0111, 0x0113, - 0x0115, 0x0117, 0x0119, 0x011b, 0x011d, 0x011f, 0x0121, 0x0123, 0x0125, 0x0127, 0x0129, - 0x012b, 0x012d, 0x012f, 0x0131, 0x0133, 0x0135, 0x013a, 0x013c, 0x013e, 0x0140, 0x0142, - 0x0144, 0x0146, 0x014b, 0x014d, 0x014f, 0x0151, 0x0153, 0x0155, 0x0157, 0x0159, 0x015b, - 0x015d, 0x015f, 0x0161, 0x0163, 0x0165, 0x0167, 0x0169, 0x016b, 0x016d, 0x016f, 0x0171, - 0x0173, 0x0175, 0x0177, 0x017a, 0x017c, 0x0183, 0x0185, 0x0188, 0x0192, 0x0195, 0x019e, - 0x01a1, 0x01a3, 0x01a5, 0x01a8, 0x01ad, 0x01b0, 0x01b4, 0x01b6, 0x01c6, 0x01c9, 0x01cc, - 0x01ce, 0x01d0, 0x01d2, 0x01d4, 0x01d6, 0x01d8, 0x01da, 0x01df, 0x01e1, 0x01e3, 0x01e5, - 0x01e7, 0x01e9, 0x01eb, 0x01ed, 0x01f3, 0x01f5, 0x01f9, 0x01fb, 0x01fd, 0x01ff, 0x0201, - 0x0203, 0x0205, 0x0207, 0x0209, 0x020b, 0x020d, 0x020f, 0x0211, 0x0213, 0x0215, 0x0217, - 0x0219, 0x021b, 0x021d, 0x021f, 0x0221, 0x0223, 0x0225, 0x0227, 0x0229, 0x022b, 0x022d, - 0x022f, 0x0231, 0x023c, 0x0242, 0x0247, 0x0249, 0x024b, 0x024d, 0x0371, 0x0373, 0x0377, - 0x0390, 0x03d9, 0x03db, 0x03dd, 0x03df, 0x03e1, 0x03e3, 0x03e5, 0x03e7, 0x03e9, 0x03eb, - 0x03ed, 0x03f5, 0x03f8, 0x0461, 0x0463, 0x0465, 0x0467, 0x0469, 0x046b, 0x046d, 0x046f, - 0x0471, 0x0473, 0x0475, 0x0477, 0x0479, 0x047b, 0x047d, 0x047f, 0x0481, 0x048b, 0x048d, - 0x048f, 0x0491, 0x0493, 0x0495, 0x0497, 0x0499, 0x049b, 0x049d, 0x049f, 0x04a1, 0x04a3, - 0x04a5, 0x04a7, 0x04a9, 0x04ab, 0x04ad, 0x04af, 0x04b1, 0x04b3, 0x04b5, 0x04b7, 0x04b9, - 0x04bb, 0x04bd, 0x04bf, 0x04c2, 0x04c4, 0x04c6, 0x04c8, 0x04ca, 0x04cc, 0x04d1, 0x04d3, - 0x04d5, 0x04d7, 0x04d9, 0x04db, 0x04dd, 0x04df, 0x04e1, 0x04e3, 0x04e5, 0x04e7, 0x04e9, - 0x04eb, 0x04ed, 0x04ef, 0x04f1, 0x04f3, 0x04f5, 0x04f7, 0x04f9, 0x04fb, 0x04fd, 0x04ff, - 0x0501, 0x0503, 0x0505, 0x0507, 0x0509, 0x050b, 0x050d, 0x050f, 0x0511, 0x0513, 0x0515, - 0x0517, 0x0519, 0x051b, 0x051d, 0x051f, 0x0521, 0x0523, 0x0525, 0x0527, 0x0529, 0x052b, - 0x052d, 0x052f, 0x1e01, 0x1e03, 0x1e05, 0x1e07, 0x1e09, 0x1e0b, 0x1e0d, 0x1e0f, 0x1e11, - 0x1e13, 0x1e15, 0x1e17, 0x1e19, 0x1e1b, 0x1e1d, 0x1e1f, 0x1e21, 0x1e23, 0x1e25, 0x1e27, - 0x1e29, 0x1e2b, 0x1e2d, 0x1e2f, 0x1e31, 0x1e33, 0x1e35, 0x1e37, 0x1e39, 0x1e3b, 0x1e3d, - 0x1e3f, 0x1e41, 0x1e43, 0x1e45, 0x1e47, 0x1e49, 0x1e4b, 0x1e4d, 0x1e4f, 0x1e51, 0x1e53, - 0x1e55, 0x1e57, 0x1e59, 0x1e5b, 0x1e5d, 0x1e5f, 0x1e61, 0x1e63, 0x1e65, 0x1e67, 0x1e69, - 0x1e6b, 0x1e6d, 0x1e6f, 0x1e71, 0x1e73, 0x1e75, 0x1e77, 0x1e79, 0x1e7b, 0x1e7d, 0x1e7f, - 0x1e81, 0x1e83, 0x1e85, 0x1e87, 0x1e89, 0x1e8b, 0x1e8d, 0x1e8f, 0x1e91, 0x1e93, 0x1e9f, - 0x1ea1, 0x1ea3, 0x1ea5, 0x1ea7, 0x1ea9, 0x1eab, 0x1ead, 0x1eaf, 0x1eb1, 0x1eb3, 0x1eb5, - 0x1eb7, 0x1eb9, 0x1ebb, 0x1ebd, 0x1ebf, 0x1ec1, 0x1ec3, 0x1ec5, 0x1ec7, 0x1ec9, 0x1ecb, - 0x1ecd, 0x1ecf, 0x1ed1, 0x1ed3, 0x1ed5, 0x1ed7, 0x1ed9, 0x1edb, 0x1edd, 0x1edf, 0x1ee1, - 0x1ee3, 0x1ee5, 0x1ee7, 0x1ee9, 0x1eeb, 0x1eed, 0x1eef, 0x1ef1, 0x1ef3, 0x1ef5, 0x1ef7, - 0x1ef9, 0x1efb, 0x1efd, 0x1fbe, 0x210a, 0x2113, 0x212f, 0x2134, 0x2139, 0x214e, 0x2184, - 0x2c61, 0x2c68, 0x2c6a, 0x2c6c, 0x2c71, 0x2c81, 0x2c83, 0x2c85, 0x2c87, 0x2c89, 0x2c8b, - 0x2c8d, 0x2c8f, 0x2c91, 0x2c93, 0x2c95, 0x2c97, 0x2c99, 0x2c9b, 0x2c9d, 0x2c9f, 0x2ca1, - 0x2ca3, 0x2ca5, 0x2ca7, 0x2ca9, 0x2cab, 0x2cad, 0x2caf, 0x2cb1, 0x2cb3, 0x2cb5, 0x2cb7, - 0x2cb9, 0x2cbb, 0x2cbd, 0x2cbf, 0x2cc1, 0x2cc3, 0x2cc5, 0x2cc7, 0x2cc9, 0x2ccb, 0x2ccd, - 0x2ccf, 0x2cd1, 0x2cd3, 0x2cd5, 0x2cd7, 0x2cd9, 0x2cdb, 0x2cdd, 0x2cdf, 0x2ce1, 0x2cec, - 0x2cee, 0x2cf3, 0x2d27, 0x2d2d, 0xa641, 0xa643, 0xa645, 0xa647, 0xa649, 0xa64b, 0xa64d, - 0xa64f, 0xa651, 0xa653, 0xa655, 0xa657, 0xa659, 0xa65b, 0xa65d, 0xa65f, 0xa661, 0xa663, - 0xa665, 0xa667, 0xa669, 0xa66b, 0xa66d, 0xa681, 0xa683, 0xa685, 0xa687, 0xa689, 0xa68b, - 0xa68d, 0xa68f, 0xa691, 0xa693, 0xa695, 0xa697, 0xa699, 0xa69b, 0xa723, 0xa725, 0xa727, - 0xa729, 0xa72b, 0xa72d, 0xa733, 0xa735, 0xa737, 0xa739, 0xa73b, 0xa73d, 0xa73f, 0xa741, - 0xa743, 0xa745, 0xa747, 0xa749, 0xa74b, 0xa74d, 0xa74f, 0xa751, 0xa753, 0xa755, 0xa757, - 0xa759, 0xa75b, 0xa75d, 0xa75f, 0xa761, 0xa763, 0xa765, 0xa767, 0xa769, 0xa76b, 0xa76d, - 0xa76f, 0xa77a, 0xa77c, 0xa77f, 0xa781, 0xa783, 0xa785, 0xa787, 0xa78c, 0xa78e, 0xa791, - 0xa797, 0xa799, 0xa79b, 0xa79d, 0xa79f, 0xa7a1, 0xa7a3, 0xa7a5, 0xa7a7, 0xa7a9, 0xa7af, - 0xa7b5, 0xa7b7, 0xa7b9, 0xa7fa, 0x1d4bb, 0x1d7cb, -}; - -int -utf8·islower(rune c) -{ - rune* p; - - p = rbsearch(c, islower_rtab, arrlen(islower_rtab) / 2, 2); - if (p && c >= p[0] && c <= p[1]) return 1; - p = rbsearch(c, islower_stab, arrlen(islower_stab), 1); - if (p && c == p[0]) return 1; - - return 0; -} - -static rune isupper_rtab[] = { - 0x0041, 0x005a, 0x00c0, 0x00d6, 0x00d8, 0x00de, 0x0178, 0x0179, 0x0181, 0x0182, - 0x0186, 0x0187, 0x0189, 0x018b, 0x018e, 0x0191, 0x0193, 0x0194, 0x0196, 0x0198, - 0x019c, 0x019d, 0x019f, 0x01a0, 0x01a6, 0x01a7, 0x01ae, 0x01af, 0x01b1, 0x01b3, - 0x01b7, 0x01b8, 0x01f6, 0x01f8, 0x023a, 0x023b, 0x023d, 0x023e, 0x0243, 0x0246, - 0x0388, 0x038a, 0x038e, 0x038f, 0x0391, 0x03a1, 0x03a3, 0x03ab, 0x03d2, 0x03d4, - 0x03f9, 0x03fa, 0x03fd, 0x042f, 0x04c0, 0x04c1, 0x0531, 0x0556, 0x10a0, 0x10c5, - 0x13a0, 0x13f5, 0x1c90, 0x1cba, 0x1cbd, 0x1cbf, 0x1f08, 0x1f0f, 0x1f18, 0x1f1d, - 0x1f28, 0x1f2f, 0x1f38, 0x1f3f, 0x1f48, 0x1f4d, 0x1f68, 0x1f6f, 0x1fb8, 0x1fbb, - 0x1fc8, 0x1fcb, 0x1fd8, 0x1fdb, 0x1fe8, 0x1fec, 0x1ff8, 0x1ffb, 0x210b, 0x210d, - 0x2110, 0x2112, 0x2119, 0x211d, 0x212a, 0x212d, 0x2130, 0x2133, 0x213e, 0x213f, - 0x2c00, 0x2c2e, 0x2c62, 0x2c64, 0x2c6d, 0x2c70, 0x2c7e, 0x2c80, 0xa77d, 0xa77e, - 0xa7aa, 0xa7ae, 0xa7b0, 0xa7b4, 0xff21, 0xff3a, 0x10400, 0x10427, 0x104b0, 0x104d3, - 0x10c80, 0x10cb2, 0x118a0, 0x118bf, 0x16e40, 0x16e5f, 0x1d400, 0x1d419, 0x1d434, 0x1d44d, - 0x1d468, 0x1d481, 0x1d49e, 0x1d49f, 0x1d4a5, 0x1d4a6, 0x1d4a9, 0x1d4ac, 0x1d4ae, 0x1d4b5, - 0x1d4d0, 0x1d4e9, 0x1d504, 0x1d505, 0x1d507, 0x1d50a, 0x1d50d, 0x1d514, 0x1d516, 0x1d51c, - 0x1d538, 0x1d539, 0x1d53b, 0x1d53e, 0x1d540, 0x1d544, 0x1d54a, 0x1d550, 0x1d56c, 0x1d585, - 0x1d5a0, 0x1d5b9, 0x1d5d4, 0x1d5ed, 0x1d608, 0x1d621, 0x1d63c, 0x1d655, 0x1d670, 0x1d689, - 0x1d6a8, 0x1d6c0, 0x1d6e2, 0x1d6fa, 0x1d71c, 0x1d734, 0x1d756, 0x1d76e, 0x1d790, 0x1d7a8, -}; - -static rune isupper_stab[] = { - 0x0100, 0x0102, 0x0104, 0x0106, 0x0108, 0x010a, 0x010c, 0x010e, 0x0110, 0x0112, 0x0114, - 0x0116, 0x0118, 0x011a, 0x011c, 0x011e, 0x0120, 0x0122, 0x0124, 0x0126, 0x0128, 0x012a, - 0x012c, 0x012e, 0x0130, 0x0132, 0x0134, 0x0136, 0x0139, 0x013b, 0x013d, 0x013f, 0x0141, - 0x0143, 0x0145, 0x0147, 0x014a, 0x014c, 0x014e, 0x0150, 0x0152, 0x0154, 0x0156, 0x0158, - 0x015a, 0x015c, 0x015e, 0x0160, 0x0162, 0x0164, 0x0166, 0x0168, 0x016a, 0x016c, 0x016e, - 0x0170, 0x0172, 0x0174, 0x0176, 0x017b, 0x017d, 0x0184, 0x01a2, 0x01a4, 0x01a9, 0x01ac, - 0x01b5, 0x01bc, 0x01c4, 0x01c7, 0x01ca, 0x01cd, 0x01cf, 0x01d1, 0x01d3, 0x01d5, 0x01d7, - 0x01d9, 0x01db, 0x01de, 0x01e0, 0x01e2, 0x01e4, 0x01e6, 0x01e8, 0x01ea, 0x01ec, 0x01ee, - 0x01f1, 0x01f4, 0x01fa, 0x01fc, 0x01fe, 0x0200, 0x0202, 0x0204, 0x0206, 0x0208, 0x020a, - 0x020c, 0x020e, 0x0210, 0x0212, 0x0214, 0x0216, 0x0218, 0x021a, 0x021c, 0x021e, 0x0220, - 0x0222, 0x0224, 0x0226, 0x0228, 0x022a, 0x022c, 0x022e, 0x0230, 0x0232, 0x0241, 0x0248, - 0x024a, 0x024c, 0x024e, 0x0370, 0x0372, 0x0376, 0x037f, 0x0386, 0x038c, 0x03cf, 0x03d8, - 0x03da, 0x03dc, 0x03de, 0x03e0, 0x03e2, 0x03e4, 0x03e6, 0x03e8, 0x03ea, 0x03ec, 0x03ee, - 0x03f4, 0x03f7, 0x0460, 0x0462, 0x0464, 0x0466, 0x0468, 0x046a, 0x046c, 0x046e, 0x0470, - 0x0472, 0x0474, 0x0476, 0x0478, 0x047a, 0x047c, 0x047e, 0x0480, 0x048a, 0x048c, 0x048e, - 0x0490, 0x0492, 0x0494, 0x0496, 0x0498, 0x049a, 0x049c, 0x049e, 0x04a0, 0x04a2, 0x04a4, - 0x04a6, 0x04a8, 0x04aa, 0x04ac, 0x04ae, 0x04b0, 0x04b2, 0x04b4, 0x04b6, 0x04b8, 0x04ba, - 0x04bc, 0x04be, 0x04c3, 0x04c5, 0x04c7, 0x04c9, 0x04cb, 0x04cd, 0x04d0, 0x04d2, 0x04d4, - 0x04d6, 0x04d8, 0x04da, 0x04dc, 0x04de, 0x04e0, 0x04e2, 0x04e4, 0x04e6, 0x04e8, 0x04ea, - 0x04ec, 0x04ee, 0x04f0, 0x04f2, 0x04f4, 0x04f6, 0x04f8, 0x04fa, 0x04fc, 0x04fe, 0x0500, - 0x0502, 0x0504, 0x0506, 0x0508, 0x050a, 0x050c, 0x050e, 0x0510, 0x0512, 0x0514, 0x0516, - 0x0518, 0x051a, 0x051c, 0x051e, 0x0520, 0x0522, 0x0524, 0x0526, 0x0528, 0x052a, 0x052c, - 0x052e, 0x10c7, 0x10cd, 0x1e00, 0x1e02, 0x1e04, 0x1e06, 0x1e08, 0x1e0a, 0x1e0c, 0x1e0e, - 0x1e10, 0x1e12, 0x1e14, 0x1e16, 0x1e18, 0x1e1a, 0x1e1c, 0x1e1e, 0x1e20, 0x1e22, 0x1e24, - 0x1e26, 0x1e28, 0x1e2a, 0x1e2c, 0x1e2e, 0x1e30, 0x1e32, 0x1e34, 0x1e36, 0x1e38, 0x1e3a, - 0x1e3c, 0x1e3e, 0x1e40, 0x1e42, 0x1e44, 0x1e46, 0x1e48, 0x1e4a, 0x1e4c, 0x1e4e, 0x1e50, - 0x1e52, 0x1e54, 0x1e56, 0x1e58, 0x1e5a, 0x1e5c, 0x1e5e, 0x1e60, 0x1e62, 0x1e64, 0x1e66, - 0x1e68, 0x1e6a, 0x1e6c, 0x1e6e, 0x1e70, 0x1e72, 0x1e74, 0x1e76, 0x1e78, 0x1e7a, 0x1e7c, - 0x1e7e, 0x1e80, 0x1e82, 0x1e84, 0x1e86, 0x1e88, 0x1e8a, 0x1e8c, 0x1e8e, 0x1e90, 0x1e92, - 0x1e94, 0x1e9e, 0x1ea0, 0x1ea2, 0x1ea4, 0x1ea6, 0x1ea8, 0x1eaa, 0x1eac, 0x1eae, 0x1eb0, - 0x1eb2, 0x1eb4, 0x1eb6, 0x1eb8, 0x1eba, 0x1ebc, 0x1ebe, 0x1ec0, 0x1ec2, 0x1ec4, 0x1ec6, - 0x1ec8, 0x1eca, 0x1ecc, 0x1ece, 0x1ed0, 0x1ed2, 0x1ed4, 0x1ed6, 0x1ed8, 0x1eda, 0x1edc, - 0x1ede, 0x1ee0, 0x1ee2, 0x1ee4, 0x1ee6, 0x1ee8, 0x1eea, 0x1eec, 0x1eee, 0x1ef0, 0x1ef2, - 0x1ef4, 0x1ef6, 0x1ef8, 0x1efa, 0x1efc, 0x1efe, 0x1f59, 0x1f5b, 0x1f5d, 0x1f5f, 0x2102, - 0x2107, 0x2115, 0x2124, 0x2126, 0x2128, 0x2145, 0x2183, 0x2c60, 0x2c67, 0x2c69, 0x2c6b, - 0x2c72, 0x2c75, 0x2c82, 0x2c84, 0x2c86, 0x2c88, 0x2c8a, 0x2c8c, 0x2c8e, 0x2c90, 0x2c92, - 0x2c94, 0x2c96, 0x2c98, 0x2c9a, 0x2c9c, 0x2c9e, 0x2ca0, 0x2ca2, 0x2ca4, 0x2ca6, 0x2ca8, - 0x2caa, 0x2cac, 0x2cae, 0x2cb0, 0x2cb2, 0x2cb4, 0x2cb6, 0x2cb8, 0x2cba, 0x2cbc, 0x2cbe, - 0x2cc0, 0x2cc2, 0x2cc4, 0x2cc6, 0x2cc8, 0x2cca, 0x2ccc, 0x2cce, 0x2cd0, 0x2cd2, 0x2cd4, - 0x2cd6, 0x2cd8, 0x2cda, 0x2cdc, 0x2cde, 0x2ce0, 0x2ce2, 0x2ceb, 0x2ced, 0x2cf2, 0xa640, - 0xa642, 0xa644, 0xa646, 0xa648, 0xa64a, 0xa64c, 0xa64e, 0xa650, 0xa652, 0xa654, 0xa656, - 0xa658, 0xa65a, 0xa65c, 0xa65e, 0xa660, 0xa662, 0xa664, 0xa666, 0xa668, 0xa66a, 0xa66c, - 0xa680, 0xa682, 0xa684, 0xa686, 0xa688, 0xa68a, 0xa68c, 0xa68e, 0xa690, 0xa692, 0xa694, - 0xa696, 0xa698, 0xa69a, 0xa722, 0xa724, 0xa726, 0xa728, 0xa72a, 0xa72c, 0xa72e, 0xa732, - 0xa734, 0xa736, 0xa738, 0xa73a, 0xa73c, 0xa73e, 0xa740, 0xa742, 0xa744, 0xa746, 0xa748, - 0xa74a, 0xa74c, 0xa74e, 0xa750, 0xa752, 0xa754, 0xa756, 0xa758, 0xa75a, 0xa75c, 0xa75e, - 0xa760, 0xa762, 0xa764, 0xa766, 0xa768, 0xa76a, 0xa76c, 0xa76e, 0xa779, 0xa77b, 0xa780, - 0xa782, 0xa784, 0xa786, 0xa78b, 0xa78d, 0xa790, 0xa792, 0xa796, 0xa798, 0xa79a, 0xa79c, - 0xa79e, 0xa7a0, 0xa7a2, 0xa7a4, 0xa7a6, 0xa7a8, 0xa7b6, 0xa7b8, 0x1d49c, 0x1d4a2, 0x1d546, - 0x1d7ca, -}; - -int -utf8·isupper(rune c) -{ - rune* p; - - p = rbsearch(c, isupper_rtab, arrlen(isupper_rtab) / 2, 2); - if (p && c >= p[0] && c <= p[1]) return 1; - p = rbsearch(c, isupper_stab, arrlen(isupper_stab), 1); - if (p && c == p[0]) return 1; - - return 0; -} - -static rune istitle_rtab[] = { - 0x1f88, 0x1f8f, 0x1f98, 0x1f9f, 0x1fa8, 0x1faf, -}; - -static rune istitle_stab[] = { - 0x01c5, 0x01c8, 0x01cb, 0x01f2, 0x1fbc, 0x1fcc, -}; - -int -utf8·istitle(rune c) -{ - rune* p; - - p = rbsearch(c, istitle_rtab, arrlen(istitle_rtab) / 2, 2); - if (p && c >= p[0] && c <= p[1]) return 1; - p = rbsearch(c, istitle_stab, arrlen(istitle_stab), 1); - if (p && c == p[0]) return 1; - - return 0; -} - -static rune isletter_rtab[] = { - 0x0041, 0x005a, 0x0061, 0x007a, 0x00c0, 0x00d6, 0x00d8, 0x00f6, 0x00f8, 0x02c1, - 0x02c6, 0x02d1, 0x02e0, 0x02e4, 0x0370, 0x0374, 0x0376, 0x0377, 0x037a, 0x037d, - 0x0388, 0x038a, 0x038e, 0x03a1, 0x03a3, 0x03f5, 0x03f7, 0x0481, 0x048a, 0x052f, - 0x0531, 0x0556, 0x0560, 0x0588, 0x05d0, 0x05ea, 0x05ef, 0x05f2, 0x0620, 0x064a, - 0x066e, 0x066f, 0x0671, 0x06d3, 0x06e5, 0x06e6, 0x06ee, 0x06ef, 0x06fa, 0x06fc, - 0x0712, 0x072f, 0x074d, 0x07a5, 0x07ca, 0x07ea, 0x07f4, 0x07f5, 0x0800, 0x0815, - 0x0840, 0x0858, 0x0860, 0x086a, 0x08a0, 0x08b4, 0x08b6, 0x08bd, 0x0904, 0x0939, - 0x0958, 0x0961, 0x0971, 0x0980, 0x0985, 0x098c, 0x098f, 0x0990, 0x0993, 0x09a8, - 0x09aa, 0x09b0, 0x09b6, 0x09b9, 0x09dc, 0x09dd, 0x09df, 0x09e1, 0x09f0, 0x09f1, - 0x0a05, 0x0a0a, 0x0a0f, 0x0a10, 0x0a13, 0x0a28, 0x0a2a, 0x0a30, 0x0a32, 0x0a33, - 0x0a35, 0x0a36, 0x0a38, 0x0a39, 0x0a59, 0x0a5c, 0x0a72, 0x0a74, 0x0a85, 0x0a8d, - 0x0a8f, 0x0a91, 0x0a93, 0x0aa8, 0x0aaa, 0x0ab0, 0x0ab2, 0x0ab3, 0x0ab5, 0x0ab9, - 0x0ae0, 0x0ae1, 0x0b05, 0x0b0c, 0x0b0f, 0x0b10, 0x0b13, 0x0b28, 0x0b2a, 0x0b30, - 0x0b32, 0x0b33, 0x0b35, 0x0b39, 0x0b5c, 0x0b5d, 0x0b5f, 0x0b61, 0x0b85, 0x0b8a, - 0x0b8e, 0x0b90, 0x0b92, 0x0b95, 0x0b99, 0x0b9a, 0x0b9e, 0x0b9f, 0x0ba3, 0x0ba4, - 0x0ba8, 0x0baa, 0x0bae, 0x0bb9, 0x0c05, 0x0c0c, 0x0c0e, 0x0c10, 0x0c12, 0x0c28, - 0x0c2a, 0x0c39, 0x0c58, 0x0c5a, 0x0c60, 0x0c61, 0x0c85, 0x0c8c, 0x0c8e, 0x0c90, - 0x0c92, 0x0ca8, 0x0caa, 0x0cb3, 0x0cb5, 0x0cb9, 0x0ce0, 0x0ce1, 0x0cf1, 0x0cf2, - 0x0d05, 0x0d0c, 0x0d0e, 0x0d10, 0x0d12, 0x0d3a, 0x0d54, 0x0d56, 0x0d5f, 0x0d61, - 0x0d7a, 0x0d7f, 0x0d85, 0x0d96, 0x0d9a, 0x0db1, 0x0db3, 0x0dbb, 0x0dc0, 0x0dc6, - 0x0e01, 0x0e30, 0x0e32, 0x0e33, 0x0e40, 0x0e46, 0x0e81, 0x0e82, 0x0e87, 0x0e88, - 0x0e94, 0x0e97, 0x0e99, 0x0e9f, 0x0ea1, 0x0ea3, 0x0eaa, 0x0eab, 0x0ead, 0x0eb0, - 0x0eb2, 0x0eb3, 0x0ec0, 0x0ec4, 0x0edc, 0x0edf, 0x0f40, 0x0f47, 0x0f49, 0x0f6c, - 0x0f88, 0x0f8c, 0x1000, 0x102a, 0x1050, 0x1055, 0x105a, 0x105d, 0x1065, 0x1066, - 0x106e, 0x1070, 0x1075, 0x1081, 0x10a0, 0x10c5, 0x10d0, 0x10fa, 0x10fc, 0x1248, - 0x124a, 0x124d, 0x1250, 0x1256, 0x125a, 0x125d, 0x1260, 0x1288, 0x128a, 0x128d, - 0x1290, 0x12b0, 0x12b2, 0x12b5, 0x12b8, 0x12be, 0x12c2, 0x12c5, 0x12c8, 0x12d6, - 0x12d8, 0x1310, 0x1312, 0x1315, 0x1318, 0x135a, 0x1380, 0x138f, 0x13a0, 0x13f5, - 0x13f8, 0x13fd, 0x1401, 0x166c, 0x166f, 0x167f, 0x1681, 0x169a, 0x16a0, 0x16ea, - 0x16f1, 0x16f8, 0x1700, 0x170c, 0x170e, 0x1711, 0x1720, 0x1731, 0x1740, 0x1751, - 0x1760, 0x176c, 0x176e, 0x1770, 0x1780, 0x17b3, 0x1820, 0x1878, 0x1880, 0x1884, - 0x1887, 0x18a8, 0x18b0, 0x18f5, 0x1900, 0x191e, 0x1950, 0x196d, 0x1970, 0x1974, - 0x1980, 0x19ab, 0x19b0, 0x19c9, 0x1a00, 0x1a16, 0x1a20, 0x1a54, 0x1b05, 0x1b33, - 0x1b45, 0x1b4b, 0x1b83, 0x1ba0, 0x1bae, 0x1baf, 0x1bba, 0x1be5, 0x1c00, 0x1c23, - 0x1c4d, 0x1c4f, 0x1c5a, 0x1c7d, 0x1c80, 0x1c88, 0x1c90, 0x1cba, 0x1cbd, 0x1cbf, - 0x1ce9, 0x1cec, 0x1cee, 0x1cf1, 0x1cf5, 0x1cf6, 0x1d00, 0x1dbf, 0x1e00, 0x1f15, - 0x1f18, 0x1f1d, 0x1f20, 0x1f45, 0x1f48, 0x1f4d, 0x1f50, 0x1f57, 0x1f5f, 0x1f7d, - 0x1f80, 0x1fb4, 0x1fb6, 0x1fbc, 0x1fc2, 0x1fc4, 0x1fc6, 0x1fcc, 0x1fd0, 0x1fd3, - 0x1fd6, 0x1fdb, 0x1fe0, 0x1fec, 0x1ff2, 0x1ff4, 0x1ff6, 0x1ffc, 0x2090, 0x209c, - 0x210a, 0x2113, 0x2119, 0x211d, 0x212a, 0x212d, 0x212f, 0x2139, 0x213c, 0x213f, - 0x2145, 0x2149, 0x2183, 0x2184, 0x2c00, 0x2c2e, 0x2c30, 0x2c5e, 0x2c60, 0x2ce4, - 0x2ceb, 0x2cee, 0x2cf2, 0x2cf3, 0x2d00, 0x2d25, 0x2d30, 0x2d67, 0x2d80, 0x2d96, - 0x2da0, 0x2da6, 0x2da8, 0x2dae, 0x2db0, 0x2db6, 0x2db8, 0x2dbe, 0x2dc0, 0x2dc6, - 0x2dc8, 0x2dce, 0x2dd0, 0x2dd6, 0x2dd8, 0x2dde, 0x3005, 0x3006, 0x3031, 0x3035, - 0x303b, 0x303c, 0x3041, 0x3096, 0x309d, 0x309f, 0x30a1, 0x30fa, 0x30fc, 0x30ff, - 0x3105, 0x312f, 0x3131, 0x318e, 0x31a0, 0x31ba, 0x31f0, 0x31ff, 0xa000, 0xa48c, - 0xa4d0, 0xa4fd, 0xa500, 0xa60c, 0xa610, 0xa61f, 0xa62a, 0xa62b, 0xa640, 0xa66e, - 0xa67f, 0xa69d, 0xa6a0, 0xa6e5, 0xa717, 0xa71f, 0xa722, 0xa788, 0xa78b, 0xa7b9, - 0xa7f7, 0xa801, 0xa803, 0xa805, 0xa807, 0xa80a, 0xa80c, 0xa822, 0xa840, 0xa873, - 0xa882, 0xa8b3, 0xa8f2, 0xa8f7, 0xa8fd, 0xa8fe, 0xa90a, 0xa925, 0xa930, 0xa946, - 0xa960, 0xa97c, 0xa984, 0xa9b2, 0xa9e0, 0xa9e4, 0xa9e6, 0xa9ef, 0xa9fa, 0xa9fe, - 0xaa00, 0xaa28, 0xaa40, 0xaa42, 0xaa44, 0xaa4b, 0xaa60, 0xaa76, 0xaa7e, 0xaaaf, - 0xaab5, 0xaab6, 0xaab9, 0xaabd, 0xaadb, 0xaadd, 0xaae0, 0xaaea, 0xaaf2, 0xaaf4, - 0xab01, 0xab06, 0xab09, 0xab0e, 0xab11, 0xab16, 0xab20, 0xab26, 0xab28, 0xab2e, - 0xab30, 0xab5a, 0xab5c, 0xab65, 0xab70, 0xabe2, 0xd7b0, 0xd7c6, 0xd7cb, 0xd7fb, - 0xf900, 0xfa6d, 0xfa70, 0xfad9, 0xfb00, 0xfb06, 0xfb13, 0xfb17, 0xfb1f, 0xfb28, - 0xfb2a, 0xfb36, 0xfb38, 0xfb3c, 0xfb40, 0xfb41, 0xfb43, 0xfb44, 0xfb46, 0xfbb1, - 0xfbd3, 0xfd3d, 0xfd50, 0xfd8f, 0xfd92, 0xfdc7, 0xfdf0, 0xfdfb, 0xfe70, 0xfe74, - 0xfe76, 0xfefc, 0xff21, 0xff3a, 0xff41, 0xff5a, 0xff66, 0xffbe, 0xffc2, 0xffc7, - 0xffca, 0xffcf, 0xffd2, 0xffd7, 0xffda, 0xffdc, 0x10000, 0x1000b, 0x1000d, 0x10026, - 0x10028, 0x1003a, 0x1003c, 0x1003d, 0x1003f, 0x1004d, 0x10050, 0x1005d, 0x10080, 0x100fa, - 0x10280, 0x1029c, 0x102a0, 0x102d0, 0x10300, 0x1031f, 0x1032d, 0x10340, 0x10342, 0x10349, - 0x10350, 0x10375, 0x10380, 0x1039d, 0x103a0, 0x103c3, 0x103c8, 0x103cf, 0x10400, 0x1049d, - 0x104b0, 0x104d3, 0x104d8, 0x104fb, 0x10500, 0x10527, 0x10530, 0x10563, 0x10600, 0x10736, - 0x10740, 0x10755, 0x10760, 0x10767, 0x10800, 0x10805, 0x1080a, 0x10835, 0x10837, 0x10838, - 0x1083f, 0x10855, 0x10860, 0x10876, 0x10880, 0x1089e, 0x108e0, 0x108f2, 0x108f4, 0x108f5, - 0x10900, 0x10915, 0x10920, 0x10939, 0x10980, 0x109b7, 0x109be, 0x109bf, 0x10a10, 0x10a13, - 0x10a15, 0x10a17, 0x10a19, 0x10a35, 0x10a60, 0x10a7c, 0x10a80, 0x10a9c, 0x10ac0, 0x10ac7, - 0x10ac9, 0x10ae4, 0x10b00, 0x10b35, 0x10b40, 0x10b55, 0x10b60, 0x10b72, 0x10b80, 0x10b91, - 0x10c00, 0x10c48, 0x10c80, 0x10cb2, 0x10cc0, 0x10cf2, 0x10d00, 0x10d23, 0x10f00, 0x10f1c, - 0x10f30, 0x10f45, 0x11003, 0x11037, 0x11083, 0x110af, 0x110d0, 0x110e8, 0x11103, 0x11126, - 0x11150, 0x11172, 0x11183, 0x111b2, 0x111c1, 0x111c4, 0x11200, 0x11211, 0x11213, 0x1122b, - 0x11280, 0x11286, 0x1128a, 0x1128d, 0x1128f, 0x1129d, 0x1129f, 0x112a8, 0x112b0, 0x112de, - 0x11305, 0x1130c, 0x1130f, 0x11310, 0x11313, 0x11328, 0x1132a, 0x11330, 0x11332, 0x11333, - 0x11335, 0x11339, 0x1135d, 0x11361, 0x11400, 0x11434, 0x11447, 0x1144a, 0x11480, 0x114af, - 0x114c4, 0x114c5, 0x11580, 0x115ae, 0x115d8, 0x115db, 0x11600, 0x1162f, 0x11680, 0x116aa, - 0x11700, 0x1171a, 0x11800, 0x1182b, 0x118a0, 0x118df, 0x11a0b, 0x11a32, 0x11a5c, 0x11a83, - 0x11a86, 0x11a89, 0x11ac0, 0x11af8, 0x11c00, 0x11c08, 0x11c0a, 0x11c2e, 0x11c72, 0x11c8f, - 0x11d00, 0x11d06, 0x11d08, 0x11d09, 0x11d0b, 0x11d30, 0x11d60, 0x11d65, 0x11d67, 0x11d68, - 0x11d6a, 0x11d89, 0x11ee0, 0x11ef2, 0x12000, 0x12399, 0x12480, 0x12543, 0x13000, 0x1342e, - 0x14400, 0x14646, 0x16800, 0x16a38, 0x16a40, 0x16a5e, 0x16ad0, 0x16aed, 0x16b00, 0x16b2f, - 0x16b40, 0x16b43, 0x16b63, 0x16b77, 0x16b7d, 0x16b8f, 0x16e40, 0x16e7f, 0x16f00, 0x16f44, - 0x16f93, 0x16f9f, 0x16fe0, 0x16fe1, 0x18800, 0x18af2, 0x1b000, 0x1b11e, 0x1b170, 0x1b2fb, - 0x1bc00, 0x1bc6a, 0x1bc70, 0x1bc7c, 0x1bc80, 0x1bc88, 0x1bc90, 0x1bc99, 0x1d400, 0x1d454, - 0x1d456, 0x1d49c, 0x1d49e, 0x1d49f, 0x1d4a5, 0x1d4a6, 0x1d4a9, 0x1d4ac, 0x1d4ae, 0x1d4b9, - 0x1d4bd, 0x1d4c3, 0x1d4c5, 0x1d505, 0x1d507, 0x1d50a, 0x1d50d, 0x1d514, 0x1d516, 0x1d51c, - 0x1d51e, 0x1d539, 0x1d53b, 0x1d53e, 0x1d540, 0x1d544, 0x1d54a, 0x1d550, 0x1d552, 0x1d6a5, - 0x1d6a8, 0x1d6c0, 0x1d6c2, 0x1d6da, 0x1d6dc, 0x1d6fa, 0x1d6fc, 0x1d714, 0x1d716, 0x1d734, - 0x1d736, 0x1d74e, 0x1d750, 0x1d76e, 0x1d770, 0x1d788, 0x1d78a, 0x1d7a8, 0x1d7aa, 0x1d7c2, - 0x1d7c4, 0x1d7cb, 0x1e800, 0x1e8c4, 0x1e900, 0x1e943, 0x1ee00, 0x1ee03, 0x1ee05, 0x1ee1f, - 0x1ee21, 0x1ee22, 0x1ee29, 0x1ee32, 0x1ee34, 0x1ee37, 0x1ee4d, 0x1ee4f, 0x1ee51, 0x1ee52, - 0x1ee61, 0x1ee62, 0x1ee67, 0x1ee6a, 0x1ee6c, 0x1ee72, 0x1ee74, 0x1ee77, 0x1ee79, 0x1ee7c, - 0x1ee80, 0x1ee89, 0x1ee8b, 0x1ee9b, 0x1eea1, 0x1eea3, 0x1eea5, 0x1eea9, 0x1eeab, 0x1eebb, -}; - -static rune isletter_stab[] = { - 0x00aa, 0x00b5, 0x00ba, 0x02ec, 0x02ee, 0x037f, 0x0386, 0x038c, 0x0559, 0x06d5, - 0x06ff, 0x0710, 0x07b1, 0x07fa, 0x081a, 0x0824, 0x0828, 0x093d, 0x0950, 0x09b2, - 0x09bd, 0x09ce, 0x09fc, 0x0a5e, 0x0abd, 0x0ad0, 0x0af9, 0x0b3d, 0x0b71, 0x0b83, - 0x0b9c, 0x0bd0, 0x0c3d, 0x0c80, 0x0cbd, 0x0cde, 0x0d3d, 0x0d4e, 0x0dbd, 0x0e84, - 0x0e8a, 0x0e8d, 0x0ea5, 0x0ea7, 0x0ebd, 0x0ec6, 0x0f00, 0x103f, 0x1061, 0x108e, - 0x10c7, 0x10cd, 0x1258, 0x12c0, 0x17d7, 0x17dc, 0x18aa, 0x1aa7, 0x1f59, 0x1f5b, - 0x1f5d, 0x1fbe, 0x2071, 0x207f, 0x2102, 0x2107, 0x2115, 0x2124, 0x2126, 0x2128, - 0x214e, 0x2d27, 0x2d2d, 0x2d6f, 0x2e2f, 0x3400, 0x4db5, 0x4e00, 0x9fef, 0xa8fb, - 0xa9cf, 0xaa7a, 0xaab1, 0xaac0, 0xaac2, 0xac00, 0xd7a3, 0xfb1d, 0xfb3e, 0x10808, - 0x1083c, 0x10a00, 0x10f27, 0x11144, 0x11176, 0x111da, 0x111dc, 0x11288, 0x1133d, 0x11350, - 0x114c7, 0x11644, 0x118ff, 0x11a00, 0x11a3a, 0x11a50, 0x11a9d, 0x11c40, 0x11d46, 0x11d98, - 0x16f50, 0x17000, 0x187f1, 0x1d4a2, 0x1d4bb, 0x1d546, 0x1ee24, 0x1ee27, 0x1ee39, 0x1ee3b, - 0x1ee42, 0x1ee47, 0x1ee49, 0x1ee4b, 0x1ee54, 0x1ee57, 0x1ee59, 0x1ee5b, 0x1ee5d, 0x1ee5f, - 0x1ee64, 0x1ee7e, 0x20000, 0x2a6d6, 0x2a700, 0x2b734, 0x2b740, 0x2b81d, 0x2b820, 0x2cea1, - 0x2ceb0, 0x2ebe0, -}; - -int -utf8·isletter(rune c) -{ - rune* p; - - p = rbsearch(c, isletter_rtab, arrlen(isletter_rtab) / 2, 2); - if (p && c >= p[0] && c <= p[1]) return 1; - p = rbsearch(c, isletter_stab, arrlen(isletter_stab), 1); - if (p && c == p[0]) return 1; - - return 0; -} - -static rune isdigit_rtab[] = { - 0x0030, 0x0039, 0x0660, 0x0669, 0x06f0, 0x06f9, 0x07c0, 0x07c9, 0x0966, 0x096f, - 0x09e6, 0x09ef, 0x0a66, 0x0a6f, 0x0ae6, 0x0aef, 0x0b66, 0x0b6f, 0x0be6, 0x0bef, - 0x0c66, 0x0c6f, 0x0ce6, 0x0cef, 0x0d66, 0x0d6f, 0x0de6, 0x0def, 0x0e50, 0x0e59, - 0x0ed0, 0x0ed9, 0x0f20, 0x0f29, 0x1040, 0x1049, 0x1090, 0x1099, 0x17e0, 0x17e9, - 0x1810, 0x1819, 0x1946, 0x194f, 0x19d0, 0x19d9, 0x1a80, 0x1a89, 0x1a90, 0x1a99, - 0x1b50, 0x1b59, 0x1bb0, 0x1bb9, 0x1c40, 0x1c49, 0x1c50, 0x1c59, 0xa620, 0xa629, - 0xa8d0, 0xa8d9, 0xa900, 0xa909, 0xa9d0, 0xa9d9, 0xa9f0, 0xa9f9, 0xaa50, 0xaa59, - 0xabf0, 0xabf9, 0xff10, 0xff19, 0x104a0, 0x104a9, 0x10d30, 0x10d39, 0x11066, 0x1106f, - 0x110f0, 0x110f9, 0x11136, 0x1113f, 0x111d0, 0x111d9, 0x112f0, 0x112f9, 0x11450, 0x11459, - 0x114d0, 0x114d9, 0x11650, 0x11659, 0x116c0, 0x116c9, 0x11730, 0x11739, 0x118e0, 0x118e9, - 0x11c50, 0x11c59, 0x11d50, 0x11d59, 0x11da0, 0x11da9, 0x16a60, 0x16a69, 0x16b50, 0x16b59, - 0x1d7ce, 0x1d7ff, -}; - -int -utf8·isdigit(rune c) -{ - rune* p; - - p = rbsearch(c, isdigit_rtab, arrlen(isdigit_rtab) / 2, 2); - if (p && c >= p[0] && c <= p[1]) return 1; - - return 0; -} diff --git a/sys/base/arg.c b/sys/base/arg.c index 64e4dd6..269043e 100644 --- a/sys/base/arg.c +++ b/sys/base/arg.c @@ -1 +1,71 @@ +#include +#include + +// NOTE: this utf8 bit is copied from libunicode to remove the hard dependency just for ARG_BEGIN. + +#define UTFmax 4 +#define RuneSync 0x80u +#define RuneSelf 0x80u +#define RuneErr 0xFFFDu +#define RuneMax 0x10FFFFu +#define RuneMask 0x1FFFFFu + +#define Bit(i) (7-(i)) +/* N 0's preceded by i 1's e.g. T(Bit(2)) is 1100 0000 */ +#define Tbyte(i) (((1 << (Bit(i)+1))-1) ^ 0xFF) +/* 0000 0000 0000 0111 1111 1111 */ +#define RuneX(i) ((1 << (Bit(i) + ((i)-1)*Bitx))-1) +enum +{ + Bitx = Bit(1), + Tx = Tbyte(1), + Rune1 = (1 << (Bit(0)+0*Bitx)) - 1, + + Maskx = (1 << Bitx) - 1, /* 0011 1111 */ + Testx = Maskx ^ 0xff, /* 1100 0000 */ + + SurrogateMin = 0xD800, + SurrogateMax = 0xDFFF, + Bad = RuneErr, +}; + + +int +arg·bytetorune(uint32* r, byte* s) +{ + int c[4], i; + uint32 l; + + c[0] = *(ubyte*)(s); + if(c[0] < Tx) { + *r = c[0]; + return 1; + } + + l = c[0]; + for(i = 1; i < UTFmax; i++) { + c[i] = *(ubyte*)(s+i); + c[i] ^= Tx; + if (c[i] & Testx) goto bad; + + l = (l << Bitx) | c[i]; + if(c[0] < Tbyte(i + 2)) { + l &= RuneX(i + 1); + if (i == 1) { + if (c[0] < Tbyte(2) || l <= Rune1) + goto bad; + } else if (l <= RuneX(i) || l > RuneMax) + goto bad; + if (i == 2 && SurrogateMin <= l && l <= SurrogateMax) + goto bad; + + *r = l; + return i + 1; + } + } +bad: + *r = RuneErr; + return 1; +} + char *argv0; diff --git a/sys/base/bufio.c b/sys/base/bufio.c index 7ebee68..cc08f66 100644 --- a/sys/base/bufio.c +++ b/sys/base/bufio.c @@ -94,6 +94,7 @@ bufio·ungetbyte(io·Buffer *buf, byte c) return 0; } +#if 0 rune bufio·getrune(io·Buffer *buf) { @@ -152,6 +153,7 @@ bufio·ungetrune(io·Buffer *buf, rune r) buf->pos -= buf->runesize; return 0; } +#endif int bufio·read(io·Buffer *buf, int sz, int n, void *out) diff --git a/sys/base/genutf8.py b/sys/base/genutf8.py deleted file mode 100755 index 9ce9975..0000000 --- a/sys/base/genutf8.py +++ /dev/null @@ -1,140 +0,0 @@ -#!/bin/python - -preamble = """ -Rune* -rbsearch(Rune c, Rune *t, int n, int nelem) -{ - Rune *p; - int m; - - while (n > 1) { - m = n >> 1; - p = t + m*nelem; - if (c >= p[0]) { - t = p; - n -= m; - } else n = m; - } - - if (n && c >= t[0]) return t; - else return 0; -} -""" - -def findrange(s): - rs = [] - ss = [] - b, c = None, None - for elt in sorted(s): - if b is None: - b, c = elt, elt - continue - if elt == c + 1: - c = elt - continue - else: - if b == c: - ss.append(b) - else: - rs.append( (b, c) ) - b, c = elt, elt - return rs, ss - -def puttab(s, fd, name): - lname = name.lower() - rs, ss = findrange(s) - - rtabn = "%s_rtab"%lname - stabn = "%s_stab"%lname - - fd.write("static Rune %s[] = {" % rtabn) - for i, elts in enumerate(sorted(rs)): - fd.write("0x%04x, 0x%04x,\n" % (elts[0], elts[1])) - fd.write("};\n\n") - - if len(ss) > 0: - fd.write("static Rune %s[] = {" % stabn) - for i, elt in enumerate(sorted(ss)): - if i % 2 == 0: - fd.write("\n") - fd.write("0x%04x, "%(elt)) - fd.write("\n};\n\n") - - fd.write("int\n") - fd.write("utf8·%s(Rune c)\n"%name) - fd.write("{\n") - fd.write("\tRune *p;\n\n") - fd.write("\tp = rbsearch(c, %s, arrlen(%s)/2, 2);\n"%(rtabn,rtabn)) - fd.write("\tif (p && c >= p[0] && c <= p[1])\n") - fd.write("\t\t\treturn 1;\n") - if len(ss) > 0: - fd.write("\tp = rbsearch(c, %s, arrlen(%s), 1);\n"%(stabn,stabn)) - fd.write("\tif (p && c == p[0])\n") - fd.write("\t\treturn 1;\n") - fd.write("\n\treturn 0;\n") - fd.write("}\n\n") - -RUNEMAX = 0x10FFFF - -if __name__ == "__main__": - alphas = set() - lowers = set() - uppers = set() - titles = set() - digits = set() - spaces = set() - - tolower = {} - toupper = {} - totitle = {} - - with open("data/UnicodeData.txt") as utf8: - for runedata in utf8: - fields = runedata[:-1].split(";") - rune = int(fields[0], 16) - kind = fields[2] - - assert rune < RUNEMAX - - if kind[0] == 'Z': - spaces.add(rune) - elif kind[0] == 'L': - alphas.add(rune) - - if kind[1] == "l": - lowers.add(rune) - elif kind[1] == "u": - uppers.add(rune) - elif kind[1] == "t": - titles.add(rune) - - if (fields[12] != ""): - toupper[rune] = int(fields[12], 16) - - if (fields[13] != ""): - tolower[rune] = int(fields[13], 16) - - if (fields[14] != ""): - totitle[rune] = int(fields[14], 16) - - elif kind == 'Nd': - digits.add(rune) - - # Some annoying exceptions... - # These are counted as control characters - spaces.add(ord('\t')) - spaces.add(ord('\n')) - spaces.add(ord('\f')) - spaces.add(ord('\v')) - spaces.add(0x85) - spaces.add(0xfeff) - - # TODO: Add toLower/toUpper/toTitle... - with open("utf8_gen.c", "w") as utf8: - utf8.write(preamble) - puttab(spaces, utf8, "IsSpace") - puttab(lowers, utf8, "IsLower") - puttab(uppers, utf8, "IsUpper") - puttab(titles, utf8, "IsTitle") - puttab(alphas, utf8, "IsLetter") - puttab(digits, utf8, "IsDigit") diff --git a/sys/base/os.c b/sys/base/os.c index d5277b5..a76e183 100644 --- a/sys/base/os.c +++ b/sys/base/os.c @@ -25,6 +25,6 @@ os·basename(byte *path) { byte *sep; - sep = utf8·findrrune(path, os·sep()); + sep = strrchr(path, os·sep()); return (sep == nil) ? path : sep+1; } diff --git a/sys/base/string.c b/sys/base/string.c index 8973a4e..1ef9068 100644 --- a/sys/base/string.c +++ b/sys/base/string.c @@ -10,180 +10,6 @@ typedef struct Hdr byte buf[]; } Hdr; -// ------------------------------------------------------------------------- -// UTF-8 functions - -#define Bit(i) (7-(i)) -/* N 0's preceded by i 1's e.g. T(Bit(2)) is 1100 0000 */ -#define Tbyte(i) (((1 << (Bit(i)+1))-1) ^ 0xFF) -/* 0000 0000 0000 0111 1111 1111 */ -#define RuneX(i) ((1 << (Bit(i) + ((i)-1)*Bitx))-1) - -enum -{ - Bitx = Bit(1), - Tx = Tbyte(1), - Rune1 = (1 << (Bit(0)+0*Bitx)) - 1, - - Maskx = (1 << Bitx) - 1, /* 0011 1111 */ - Testx = Maskx ^ 0xff, /* 1100 0000 */ - - SurrogateMin = 0xD800, - SurrogateMax = 0xDFFF, - Bad = RuneErr, -}; - -int -utf8·bytetorune(rune* r, byte* s) -{ - int c[UTFmax], i; - rune l; - - c[0] = *(ubyte*)(s); - if(c[0] < Tx) { - *r = c[0]; - return 1; - } - - l = c[0]; - for(i = 1; i < UTFmax; i++) { - c[i] = *(ubyte*)(s+i); - c[i] ^= Tx; - if (c[i] & Testx) goto bad; - - l = (l << Bitx) | c[i]; - if(c[0] < Tbyte(i + 2)) { - l &= RuneX(i + 1); - if (i == 1) { - if (c[0] < Tbyte(2) || l <= Rune1) - goto bad; - } else if (l <= RuneX(i) || l > RuneMax) - goto bad; - if (i == 2 && SurrogateMin <= l && l <= SurrogateMax) - goto bad; - - *r = l; - return i + 1; - } - } -bad: - *r = RuneErr; - return 1; -} - -int -utf8·runetobyte(byte* s, rune* r) -{ - int i, j; - rune c; - - c = *r; - if (c <= Rune1) { - s[0] = c; - return 1; - } - - for (i = 2; i < UTFmax + 1; i++){ - if (i == 3){ - if (c > RuneMax) - c = RuneErr; - if (SurrogateMin <= c && c <= SurrogateMax) - c = RuneErr; - } - if (c <= RuneX(i) || i == UTFmax) { - s[0] = Tbyte(i) | (c >> (i - 1)*Bitx); - for(j = 1; j < i; j++) - s[j] = Tx | ((c >> (i - j - 1)*Bitx) & Maskx); - return i; - } - } - - return UTFmax; -} - -int -utf8·runelen(rune r) -{ - byte s[10]; - return utf8·runetobyte(s, &r); -} - -int -utf8·fullrune(byte* s, int n) -{ - int i; - rune c; - - if (n <= 0) return 0; - c = *(ubyte*) s; - if (c < Tx) return 1; - - for (i = 3; i < UTFmax + 1; i++) { - if (c < Tbyte(i)) return n >= i - 1; - } - - return n >= UTFmax; -} - -byte* -utf8·findrune(byte* s, long c) -{ - long c1; - rune r; - int n; - - if (c < RuneSync) return strchr(s, c); - - for (;;) { - c1 = *(ubyte*)s; - if (c1 < RuneSelf) { - if (c1 == 0) return nil; - if (c1 == c) return s; - s++; - continue; - } - n = utf8·bytetorune(&r, s); - if (r == c) return s; - s += n; - } - - return nil; -} - -byte* -utf8·findrrune(byte* s, long c) -{ - long c1; - rune r; - byte *l; - - if (c < RuneSync) - return strrchr(s, c); - - l = nil; - for (;;) { - c1 = *(ubyte*)s; - if (c1 < RuneSelf) { - if (c1 == 0) return l; - if (c1 == c) l = s; - s++; - continue; - } - c1 = utf8·bytetorune(&r, s); - if (r == c) - l = s; - s += c1; - } - - return nil; -} - -#undef Bit -#undef Tbyte -#undef RuneX - -#include ".generated/utf8.c" - // ------------------------------------------------------------------------- // Dynamic string functions -- cgit v1.2.1