Index: src/encodings/windows_1251.m ================================================================== --- src/encodings/windows_1251.m +++ src/encodings/windows_1251.m @@ -15,10 +15,12 @@ */ #include "config.h" #import "OFString.h" + +#import "common.h" const of_char16_t of_windows_1251_table[] = { 0x0402, 0x0403, 0x201A, 0x0453, 0x201E, 0x2026, 0x2020, 0x2021, 0x20AC, 0x2030, 0x0409, 0x2039, 0x040A, 0x040C, 0x040B, 0x040F, 0x0452, 0x2018, 0x2019, 0x201C, 0x201D, 0x2022, 0x2013, 0x2014, @@ -37,229 +39,102 @@ 0x0448, 0x0449, 0x044A, 0x044B, 0x044C, 0x044D, 0x044E, 0x044F }; const size_t of_windows_1251_table_offset = 256 - (sizeof(of_windows_1251_table) / sizeof(*of_windows_1251_table)); +static const unsigned char page0[] = { + 0xA0, 0x00, 0x00, 0x00, 0xA4, 0x00, 0xA6, 0xA7, + 0x00, 0xA9, 0x00, 0xAB, 0xAC, 0xAD, 0xAE, 0x00, + 0xB0, 0xB1, 0x00, 0x00, 0x00, 0xB5, 0xB6, 0xB7, + 0x00, 0x00, 0x00, 0xBB +}; +static const uint8_t page0Start = 0xA0; + +static const unsigned char page4[] = { + 0xA8, 0x80, 0x81, 0xAA, 0xBD, 0xB2, 0xAF, 0xA3, + 0x8A, 0x8C, 0x8E, 0x8D, 0x00, 0xA1, 0x8F, 0xC0, + 0xC1, 0xC2, 0xC3, 0xC4, 0xC5, 0xC6, 0xC7, 0xC8, + 0xC9, 0xCA, 0xCB, 0xCC, 0xCD, 0xCE, 0xCF, 0xD0, + 0xD1, 0xD2, 0xD3, 0xD4, 0xD5, 0xD6, 0xD7, 0xD8, + 0xD9, 0xDA, 0xDB, 0xDC, 0xDD, 0xDE, 0xDF, 0xE0, + 0xE1, 0xE2, 0xE3, 0xE4, 0xE5, 0xE6, 0xE7, 0xE8, + 0xE9, 0xEA, 0xEB, 0xEC, 0xED, 0xEE, 0xEF, 0xF0, + 0xF1, 0xF2, 0xF3, 0xF4, 0xF5, 0xF6, 0xF7, 0xF8, + 0xF9, 0xFA, 0xFB, 0xFC, 0xFD, 0xFE, 0xFF, 0x00, + 0xB8, 0x90, 0x83, 0xBA, 0xBE, 0xB3, 0xBF, 0xBC, + 0x9A, 0x9C, 0x9E, 0x9D, 0x00, 0xA2, 0x9F, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xA5, + 0xB4 +}; +static const uint8_t page4Start = 0x01; + +static const unsigned char page20[] = { + 0x96, 0x97, 0x00, 0x00, 0x00, 0x91, 0x92, 0x82, + 0x00, 0x93, 0x94, 0x84, 0x00, 0x86, 0x87, 0x95, + 0x00, 0x00, 0x00, 0x85, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x89, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x8B, 0x9B, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x88 +}; +static const uint8_t page20Start = 0x13; + +static const unsigned char page21[] = { + 0xB9, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x99 +}; +static const uint8_t page21Start = 0x16; + bool of_unicode_to_windows_1251(const of_unichar_t *input, unsigned char *output, size_t length, bool lossy) { for (size_t i = 0; i < length; i++) { of_unichar_t c = input[i]; if OF_UNLIKELY (c > 0x7F) { + uint8_t index; + if OF_UNLIKELY (c > 0xFFFF) { if (lossy) { output[i] = '?'; continue; } else return false; } - if OF_LIKELY (c >= 0x410 && c <= 0x44F) - output[i] = 0xC0 + (c - 0x410); - else { - switch ((of_char16_t)c) { - case 0x402: - output[i] = 0x80; - break; - case 0x403: - output[i] = 0x81; - break; - case 0x201A: - output[i] = 0x82; - break; - case 0x453: - output[i] = 0x83; - break; - case 0x201E: - output[i] = 0x84; - break; - case 0x2026: - output[i] = 0x85; - break; - case 0x2020: - output[i] = 0x86; - break; - case 0x2021: - output[i] = 0x87; - break; - case 0x20AC: - output[i] = 0x88; - break; - case 0x2030: - output[i] = 0x89; - break; - case 0x409: - output[i] = 0x8A; - break; - case 0x2039: - output[i] = 0x8B; - break; - case 0x40A: - output[i] = 0x8C; - break; - case 0x40C: - output[i] = 0x8D; - break; - case 0x40B: - output[i] = 0x8E; - break; - case 0x40F: - output[i] = 0x8F; - break; - case 0x452: - output[i] = 0x90; - break; - case 0x2018: - output[i] = 0x91; - break; - case 0x2019: - output[i] = 0x92; - break; - case 0x201C: - output[i] = 0x93; - break; - case 0x201D: - output[i] = 0x94; - break; - case 0x2022: - output[i] = 0x95; - break; - case 0x2013: - output[i] = 0x96; - break; - case 0x2014: - output[i] = 0x97; - break; - case 0x2122: - output[i] = 0x99; - break; - case 0x459: - output[i] = 0x9A; - break; - case 0x203A: - output[i] = 0x9B; - break; - case 0x45A: - output[i] = 0x9C; - break; - case 0x45C: - output[i] = 0x9D; - break; - case 0x45B: - output[i] = 0x9E; - break; - case 0x45F: - output[i] = 0x9F; - break; - case 0xA0: - output[i] = 0xA0; - break; - case 0x40E: - output[i] = 0xA1; - break; - case 0x45E: - output[i] = 0xA2; - break; - case 0x408: - output[i] = 0xA3; - break; - case 0xA4: - output[i] = 0xA4; - break; - case 0x490: - output[i] = 0xA5; - break; - case 0xA6: - output[i] = 0xA6; - break; - case 0xA7: - output[i] = 0xA7; - break; - case 0x401: - output[i] = 0xA8; - break; - case 0xA9: - output[i] = 0xA9; - break; - case 0x404: - output[i] = 0xAA; - break; - case 0xAB: - output[i] = 0xAB; - break; - case 0xAC: - output[i] = 0xAC; - break; - case 0xAD: - output[i] = 0xAD; - break; - case 0xAE: - output[i] = 0xAE; - break; - case 0x407: - output[i] = 0xAF; - break; - case 0xB0: - output[i] = 0xB0; - break; - case 0xB1: - output[i] = 0xB1; - break; - case 0x406: - output[i] = 0xB2; - break; - case 0x456: - output[i] = 0xB3; - break; - case 0x491: - output[i] = 0xB4; - break; - case 0xB5: - output[i] = 0xB5; - break; - case 0xB6: - output[i] = 0xB6; - break; - case 0xB7: - output[i] = 0xB7; - break; - case 0x451: - output[i] = 0xB8; - break; - case 0x2116: - output[i] = 0xB9; - break; - case 0x454: - output[i] = 0xBA; - break; - case 0xBB: - output[i] = 0xBB; - break; - case 0x458: - output[i] = 0xBC; - break; - case 0x405: - output[i] = 0xBD; - break; - case 0x455: - output[i] = 0xBE; - break; - case 0x457: - output[i] = 0xBF; - break; - default: - if (lossy) - output[i] = '?'; - else - return false; - - break; - } + switch (c >> 8) { + CASE_MISSING_IS_ERROR(0) + CASE_MISSING_IS_ERROR(4) + CASE_MISSING_IS_ERROR(20) + CASE_MISSING_IS_ERROR(21) + default: + if (lossy) { + output[i] = '?'; + continue; + } else + return false; } } else output[i] = (unsigned char)c; } return true; }