Index: src/encodings/mac_roman.m ================================================================== --- src/encodings/mac_roman.m +++ src/encodings/mac_roman.m @@ -15,20 +15,22 @@ */ #include "config.h" #import "OFString.h" + +#import "common.h" const of_char16_t of_mac_roman_table[] = { 0x00C4, 0x00C5, 0x00C7, 0x00C9, 0x00D1, 0x00D6, 0x00DC, 0x00E1, 0x00E0, 0x00E2, 0x00E4, 0x00E3, 0x00E5, 0x00E7, 0x00E9, 0x00E8, 0x00EA, 0x00EB, 0x00ED, 0x00EC, 0x00EE, 0x00EF, 0x00F1, 0x00F3, 0x00F2, 0x00F4, 0x00F6, 0x00F5, 0x00FA, 0x00F9, 0x00FB, 0x00FC, 0x2020, 0x00B0, 0x00A2, 0x00A3, 0x00A7, 0x2022, 0x00B6, 0x00DF, 0x00AE, 0x00A9, 0x2122, 0x00B4, 0x00A8, 0x2260, 0x00C6, 0x00D8, 0x221E, 0x00B1, 0x2264, 0x2265, 0x00A5, 0x00B5, 0x2202, 0x2211, - 0x220F, 0x03c0, 0x222B, 0x00AA, 0x00BA, 0x03A9, 0x00E6, 0x00F8, + 0x220F, 0x03C0, 0x222B, 0x00AA, 0x00BA, 0x03A9, 0x00E6, 0x00F8, 0x00BF, 0x00A1, 0x00AC, 0x221A, 0x0192, 0x2248, 0x2206, 0x00AB, 0x00BB, 0x2026, 0x00A0, 0x00C0, 0x00C3, 0x00D5, 0x0152, 0x0153, 0x2013, 0x2014, 0x201C, 0x201D, 0x2018, 0x2019, 0x00F7, 0x25CA, 0x00FF, 0x0178, 0x2044, 0x20AC, 0x2039, 0x203A, 0xFB01, 0xFB02, 0x2021, 0x00B7, 0x201A, 0x201E, 0x2030, 0x00C2, 0x00CA, 0x00C1, @@ -37,420 +39,155 @@ 0x00AF, 0x02D8, 0x02D9, 0x02DA, 0x00B8, 0x02DD, 0x02DB, 0x02C7 }; const size_t of_mac_roman_table_offset = 256 - (sizeof(of_mac_roman_table) / sizeof(*of_mac_roman_table)); +static const unsigned char page0[] = { + 0xCA, 0xC1, 0xA2, 0xA3, 0x00, 0xB4, 0x00, 0xA4, + 0xAC, 0xA9, 0xBB, 0xC7, 0xC2, 0x00, 0xA8, 0xF8, + 0xA1, 0xB1, 0x00, 0x00, 0xAB, 0xB5, 0xA6, 0xE1, + 0xFC, 0x00, 0xBC, 0xC8, 0x00, 0x00, 0x00, 0xC0, + 0xCB, 0xE7, 0xE5, 0xCC, 0x80, 0x81, 0xAE, 0x82, + 0xE9, 0x83, 0xE6, 0xE8, 0xED, 0xEA, 0xEB, 0xEC, + 0x00, 0x84, 0xF1, 0xEE, 0xEF, 0xCD, 0x85, 0x00, + 0xAF, 0xF4, 0xF2, 0xF3, 0x86, 0x00, 0x00, 0xA7, + 0x88, 0x87, 0x89, 0x8B, 0x8A, 0x8C, 0xBE, 0x8D, + 0x8F, 0x8E, 0x90, 0x91, 0x93, 0x92, 0x94, 0x95, + 0x00, 0x96, 0x98, 0x97, 0x99, 0x9B, 0x9A, 0xD6, + 0xBF, 0x9D, 0x9C, 0x9E, 0x9F, 0x00, 0x00, 0xD8 +}; +static const uint8_t page0Start = 0xA0; + +static const unsigned char page1[] = { + 0xF5, 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, 0xCE, 0xCF, 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, 0xD9, + 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, 0xC4 +}; +static const uint8_t page1Start = 0x31; + +static const unsigned char page2[] = { + 0xF6, 0xFF, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0xF9, 0xFA, 0xFB, 0xFE, 0xF7, 0xFD +}; +static const uint8_t page2Start = 0xC6; + +static const unsigned char page3[] = { + 0xBD, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xB9 +}; +static const uint8_t page3Start = 0xA9; + +static const unsigned char page20[] = { + 0xD0, 0xD1, 0x00, 0x00, 0x00, 0xD4, 0xD5, 0xE2, + 0x00, 0xD2, 0xD3, 0xE3, 0x00, 0xA0, 0xE0, 0xA5, + 0x00, 0x00, 0x00, 0xC9, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0xE4, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xDC, 0xDD, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0xDA, 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, 0xDB +}; +static const uint8_t page20Start = 0x13; + +static const unsigned char page21[] = { + 0xAA +}; +static const uint8_t page21Start = 0x22; + +static const unsigned char page22[] = { + 0xB6, 0x00, 0x00, 0x00, 0xC6, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0xB8, 0x00, 0xB7, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0xC3, 0x00, 0x00, 0x00, 0xB0, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0xBA, 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, 0xC5, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xAD, 0x00, + 0x00, 0x00, 0xB2, 0xB3 +}; +static const uint8_t page22Start = 0x02; + +static const unsigned char page25[] = { + 0xD7 +}; +static const uint8_t page25Start = 0xCA; + +static const unsigned char pageF8[] = { + 0xF0 +}; +static const uint8_t pageF8Start = 0xFF; + +static const unsigned char pageFB[] = { + 0xDE, 0xDF +}; +static const uint8_t pageFBStart = 0x01; + bool of_unicode_to_mac_roman(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; } - switch ((of_char16_t)c) { - case 0xC4: - output[i] = 0x80; - break; - case 0xC5: - output[i] = 0x81; - break; - case 0xC7: - output[i] = 0x82; - break; - case 0xC9: - output[i] = 0x83; - break; - case 0xD1: - output[i] = 0x84; - break; - case 0xD6: - output[i] = 0x85; - break; - case 0xDC: - output[i] = 0x86; - break; - case 0xE1: - output[i] = 0x87; - break; - case 0xE0: - output[i] = 0x88; - break; - case 0xE2: - output[i] = 0x89; - break; - case 0xE4: - output[i] = 0x8A; - break; - case 0xE3: - output[i] = 0x8B; - break; - case 0xE5: - output[i] = 0x8C; - break; - case 0xE7: - output[i] = 0x8D; - break; - case 0xE9: - output[i] = 0x8E; - break; - case 0xE8: - output[i] = 0x8F; - break; - case 0xEA: - output[i] = 0x90; - break; - case 0xEB: - output[i] = 0x91; - break; - case 0xED: - output[i] = 0x92; - break; - case 0xEC: - output[i] = 0x93; - break; - case 0xEE: - output[i] = 0x94; - break; - case 0xEF: - output[i] = 0x95; - break; - case 0xF1: - output[i] = 0x96; - break; - case 0xF3: - output[i] = 0x97; - break; - case 0xF2: - output[i] = 0x98; - break; - case 0xF4: - output[i] = 0x99; - break; - case 0xF6: - output[i] = 0x9A; - break; - case 0xF5: - output[i] = 0x9B; - break; - case 0xFA: - output[i] = 0x9C; - break; - case 0xF9: - output[i] = 0x9D; - break; - case 0xFB: - output[i] = 0x9E; - break; - case 0xFC: - output[i] = 0x9F; - break; - case 0x2020: - output[i] = 0xA0; - break; - case 0xB0: - output[i] = 0xA1; - break; - case 0xA2: - output[i] = 0xA2; - break; - case 0xA3: - output[i] = 0xA3; - break; - case 0xA7: - output[i] = 0xA4; - break; - case 0x2022: - output[i] = 0xA5; - break; - case 0xB6: - output[i] = 0xA6; - break; - case 0xDF: - output[i] = 0xA7; - break; - case 0xAE: - output[i] = 0xA8; - break; - case 0xA9: - output[i] = 0xA9; - break; - case 0x2122: - output[i] = 0xAA; - break; - case 0xB4: - output[i] = 0xAB; - break; - case 0xA8: - output[i] = 0xAC; - break; - case 0x2260: - output[i] = 0xAD; - break; - case 0xC6: - output[i] = 0xAE; - break; - case 0xD8: - output[i] = 0xAF; - break; - case 0x221E: - output[i] = 0xB0; - break; - case 0xB1: - output[i] = 0xB1; - break; - case 0x2264: - output[i] = 0xB2; - break; - case 0x2265: - output[i] = 0xB3; - break; - case 0xA5: - output[i] = 0xB4; - break; - case 0xB5: - output[i] = 0xB5; - break; - case 0x2202: - output[i] = 0xB6; - break; - case 0x2211: - output[i] = 0xB7; - break; - case 0x220F: - output[i] = 0xB8; - break; - case 0x3C0: - output[i] = 0xB9; - break; - case 0x222B: - output[i] = 0xBA; - break; - case 0xAA: - output[i] = 0xBB; - break; - case 0xBA: - output[i] = 0xBC; - break; - case 0x3A9: - output[i] = 0xBD; - break; - case 0xE6: - output[i] = 0xBE; - break; - case 0xF8: - output[i] = 0xBF; - break; - case 0xBF: - output[i] = 0xC0; - break; - case 0xA1: - output[i] = 0xC1; - break; - case 0xAC: - output[i] = 0xC2; - break; - case 0x221A: - output[i] = 0xC3; - break; - case 0x192: - output[i] = 0xC4; - break; - case 0x2248: - output[i] = 0xC5; - break; - case 0x2206: - output[i] = 0xC6; - break; - case 0xAB: - output[i] = 0xC7; - break; - case 0xBB: - output[i] = 0xC8; - break; - case 0x2026: - output[i] = 0xC9; - break; - case 0xA0: - output[i] = 0xCA; - break; - case 0xC0: - output[i] = 0xCB; - break; - case 0xC3: - output[i] = 0xCC; - break; - case 0xD5: - output[i] = 0xCD; - break; - case 0x152: - output[i] = 0xCE; - break; - case 0x153: - output[i] = 0xCF; - break; - case 0x2013: - output[i] = 0xD0; - break; - case 0x2014: - output[i] = 0xD1; - break; - case 0x201C: - output[i] = 0xD2; - break; - case 0x201D: - output[i] = 0xD3; - break; - case 0x2018: - output[i] = 0xD4; - break; - case 0x2019: - output[i] = 0xD5; - break; - case 0xF7: - output[i] = 0xD6; - break; - case 0x25CA: - output[i] = 0xD7; - break; - case 0xFF: - output[i] = 0xD8; - break; - case 0x178: - output[i] = 0xD9; - break; - case 0x2044: - output[i] = 0xDA; - break; - case 0x20AC: - output[i] = 0xDB; - break; - case 0x2039: - output[i] = 0xDC; - break; - case 0x203A: - output[i] = 0xDD; - break; - case 0xFB01: - output[i] = 0xDE; - break; - case 0xFB02: - output[i] = 0xDF; - break; - case 0x2021: - output[i] = 0xE0; - break; - case 0xB7: - output[i] = 0xE1; - break; - case 0x201A: - output[i] = 0xE2; - break; - case 0x201E: - output[i] = 0xE3; - break; - case 0x2030: - output[i] = 0xE4; - break; - case 0xC2: - output[i] = 0xE5; - break; - case 0xCA: - output[i] = 0xE6; - break; - case 0xC1: - output[i] = 0xE7; - break; - case 0xCB: - output[i] = 0xE8; - break; - case 0xC8: - output[i] = 0xE9; - break; - case 0xCD: - output[i] = 0xEA; - break; - case 0xCE: - output[i] = 0xEB; - break; - case 0xCF: - output[i] = 0xEC; - break; - case 0xCC: - output[i] = 0xED; - break; - case 0xD3: - output[i] = 0xEE; - break; - case 0xD4: - output[i] = 0xEF; - break; - case 0xF8FF: - output[i] = 0xF0; - break; - case 0xD2: - output[i] = 0xF1; - break; - case 0xDA: - output[i] = 0xF2; - break; - case 0xDB: - output[i] = 0xF3; - break; - case 0xD9: - output[i] = 0xF4; - break; - case 0x131: - output[i] = 0xF5; - break; - case 0x2C6: - output[i] = 0xF6; - break; - case 0x2DC: - output[i] = 0xF7; - break; - case 0xAF: - output[i] = 0xF8; - break; - case 0x2D8: - output[i] = 0xF9; - break; - case 0x2D9: - output[i] = 0xFA; - break; - case 0x2DA: - output[i] = 0xFB; - break; - case 0xB8: - output[i] = 0xFC; - break; - case 0x2DD: - output[i] = 0xFD; - break; - case 0x2DB: - output[i] = 0xFE; - break; - case 0x2C7: - output[i] = 0xFF; - break; + switch (c >> 8) { + CASE_MISSING_IS_ERROR(0) + CASE_MISSING_IS_ERROR(1) + CASE_MISSING_IS_ERROR(2) + CASE_MISSING_IS_ERROR(3) + CASE_MISSING_IS_ERROR(20) + CASE_MISSING_IS_ERROR(21) + CASE_MISSING_IS_ERROR(22) + CASE_MISSING_IS_ERROR(25) + CASE_MISSING_IS_ERROR(F8) + CASE_MISSING_IS_ERROR(FB) default: - if (lossy) + if (lossy) { output[i] = '?'; - else + continue; + } else return false; - - break; } } else output[i] = (unsigned char)c; } return true; }