Windows1252Encoding.cpp 5.7 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153
  1. //
  2. // Windows1252Encoding.cpp
  3. //
  4. // $Id: //poco/1.4/Foundation/src/Windows1252Encoding.cpp#1 $
  5. //
  6. // Library: Foundation
  7. // Package: Text
  8. // Module: Windows1252Encoding
  9. //
  10. // Copyright (c) 2005-2007, Applied Informatics Software Engineering GmbH.
  11. // and Contributors.
  12. //
  13. // SPDX-License-Identifier: BSL-1.0
  14. //
  15. #include "Poco/Windows1252Encoding.h"
  16. #include "Poco/String.h"
  17. #include <map>
  18. namespace Poco {
  19. const char* Windows1252Encoding::_names[] =
  20. {
  21. "windows-1252",
  22. "Windows-1252",
  23. "cp1252",
  24. "CP1252",
  25. NULL
  26. };
  27. const TextEncoding::CharacterMap Windows1252Encoding::_charMap =
  28. {
  29. /* 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f */
  30. /* 00 */ 0x0000, 0x0001, 0x0002, 0x0003, 0x0004, 0x0005, 0x0006, 0x0007, 0x0008, 0x0009, 0x000a, 0x000b, 0x000c, 0x000d, 0x000e, 0x000f,
  31. /* 10 */ 0x0010, 0x0011, 0x0012, 0x0013, 0x0014, 0x0015, 0x0016, 0x0017, 0x0018, 0x0019, 0x001a, 0x001b, 0x001c, 0x001d, 0x001e, 0x001f,
  32. /* 20 */ 0x0020, 0x0021, 0x0022, 0x0023, 0x0024, 0x0025, 0x0026, 0x0027, 0x0028, 0x0029, 0x002a, 0x002b, 0x002c, 0x002d, 0x002e, 0x002f,
  33. /* 30 */ 0x0030, 0x0031, 0x0032, 0x0033, 0x0034, 0x0035, 0x0036, 0x0037, 0x0038, 0x0039, 0x003a, 0x003b, 0x003c, 0x003d, 0x003e, 0x003f,
  34. /* 40 */ 0x0040, 0x0041, 0x0042, 0x0043, 0x0044, 0x0045, 0x0046, 0x0047, 0x0048, 0x0049, 0x004a, 0x004b, 0x004c, 0x004d, 0x004e, 0x004f,
  35. /* 50 */ 0x0050, 0x0051, 0x0052, 0x0053, 0x0054, 0x0055, 0x0056, 0x0057, 0x0058, 0x0059, 0x005a, 0x005b, 0x005c, 0x005d, 0x005e, 0x005f,
  36. /* 60 */ 0x0060, 0x0061, 0x0062, 0x0063, 0x0064, 0x0065, 0x0066, 0x0067, 0x0068, 0x0069, 0x006a, 0x006b, 0x006c, 0x006d, 0x006e, 0x006f,
  37. /* 70 */ 0x0070, 0x0071, 0x0072, 0x0073, 0x0074, 0x0075, 0x0076, 0x0077, 0x0078, 0x0079, 0x007a, 0x007b, 0x007c, 0x007d, 0x007e, 0x007f,
  38. /* 80 */ 0x20ac, 0x0081, 0x201a, 0x0192, 0x201e, 0x2026, 0x2020, 0x2021, 0x02c6, 0x2030, 0x0160, 0x2039, 0x0152, 0x008d, 0x017d, 0x008f,
  39. /* 90 */ 0x0090, 0x2018, 0x2019, 0x201c, 0x201d, 0x2022, 0x2013, 0x2014, 0x02dc, 0x2122, 0x0161, 0x203a, 0x0153, 0x009d, 0x017e, 0x0178,
  40. /* a0 */ 0x00a0, 0x00a1, 0x00a2, 0x00a3, 0x00a4, 0x00a5, 0x00a6, 0x00a7, 0x00a8, 0x00a9, 0x00aa, 0x00ab, 0x00ac, 0x00ad, 0x00ae, 0x00af,
  41. /* b0 */ 0x00b0, 0x00b1, 0x00b2, 0x00b3, 0x00b4, 0x00b5, 0x00b6, 0x00b7, 0x00b8, 0x00b9, 0x00ba, 0x00bb, 0x00bc, 0x00bd, 0x00be, 0x00bf,
  42. /* c0 */ 0x00c0, 0x00c1, 0x00c2, 0x00c3, 0x00c4, 0x00c5, 0x00c6, 0x00c7, 0x00c8, 0x00c9, 0x00ca, 0x00cb, 0x00cc, 0x00cd, 0x00ce, 0x00cf,
  43. /* d0 */ 0x00d0, 0x00d1, 0x00d2, 0x00d3, 0x00d4, 0x00d5, 0x00d6, 0x00d7, 0x00d8, 0x00d9, 0x00da, 0x00db, 0x00dc, 0x00dd, 0x00de, 0x00df,
  44. /* e0 */ 0x00e0, 0x00e1, 0x00e2, 0x00e3, 0x00e4, 0x00e5, 0x00e6, 0x00e7, 0x00e8, 0x00e9, 0x00ea, 0x00eb, 0x00ec, 0x00ed, 0x00ee, 0x00ef,
  45. /* f0 */ 0x00f0, 0x00f1, 0x00f2, 0x00f3, 0x00f4, 0x00f5, 0x00f6, 0x00f7, 0x00f8, 0x00f9, 0x00fa, 0x00fb, 0x00fc, 0x00fd, 0x00fe, 0x00ff,
  46. };
  47. Windows1252Encoding::Windows1252Encoding()
  48. {
  49. }
  50. Windows1252Encoding::~Windows1252Encoding()
  51. {
  52. }
  53. const char* Windows1252Encoding::canonicalName() const
  54. {
  55. return _names[0];
  56. }
  57. bool Windows1252Encoding::isA(const std::string& encodingName) const
  58. {
  59. for (const char** name = _names; *name; ++name)
  60. {
  61. if (Poco::icompare(encodingName, *name) == 0)
  62. return true;
  63. }
  64. return false;
  65. }
  66. const TextEncoding::CharacterMap& Windows1252Encoding::characterMap() const
  67. {
  68. return _charMap;
  69. }
  70. int Windows1252Encoding::convert(const unsigned char* bytes) const
  71. {
  72. return _charMap[*bytes];
  73. }
  74. int Windows1252Encoding::convert(int ch, unsigned char* bytes, int length) const
  75. {
  76. if (ch >= 0 && ch <= 255 && _charMap[ch] == ch)
  77. {
  78. if (bytes && length >= 1)
  79. *bytes = ch;
  80. return 1;
  81. }
  82. else switch (ch)
  83. {
  84. case 0x20ac: if (bytes && length >= 1) *bytes = 0x80; return 1;
  85. case 0x201a: if (bytes && length >= 1) *bytes = 0x82; return 1;
  86. case 0x0192: if (bytes && length >= 1) *bytes = 0x83; return 1;
  87. case 0x201e: if (bytes && length >= 1) *bytes = 0x84; return 1;
  88. case 0x2026: if (bytes && length >= 1) *bytes = 0x85; return 1;
  89. case 0x2020: if (bytes && length >= 1) *bytes = 0x86; return 1;
  90. case 0x2021: if (bytes && length >= 1) *bytes = 0x87; return 1;
  91. case 0x02c6: if (bytes && length >= 1) *bytes = 0x88; return 1;
  92. case 0x2030: if (bytes && length >= 1) *bytes = 0x89; return 1;
  93. case 0x0160: if (bytes && length >= 1) *bytes = 0x8a; return 1;
  94. case 0x2039: if (bytes && length >= 1) *bytes = 0x8b; return 1;
  95. case 0x0152: if (bytes && length >= 1) *bytes = 0x8c; return 1;
  96. case 0x017d: if (bytes && length >= 1) *bytes = 0x8e; return 1;
  97. case 0x2018: if (bytes && length >= 1) *bytes = 0x91; return 1;
  98. case 0x2019: if (bytes && length >= 1) *bytes = 0x92; return 1;
  99. case 0x201c: if (bytes && length >= 1) *bytes = 0x93; return 1;
  100. case 0x201d: if (bytes && length >= 1) *bytes = 0x94; return 1;
  101. case 0x2022: if (bytes && length >= 1) *bytes = 0x95; return 1;
  102. case 0x2013: if (bytes && length >= 1) *bytes = 0x96; return 1;
  103. case 0x2014: if (bytes && length >= 1) *bytes = 0x97; return 1;
  104. case 0x02dc: if (bytes && length >= 1) *bytes = 0x98; return 1;
  105. case 0x2122: if (bytes && length >= 1) *bytes = 0x99; return 1;
  106. case 0x0161: if (bytes && length >= 1) *bytes = 0x9a; return 1;
  107. case 0x203a: if (bytes && length >= 1) *bytes = 0x9b; return 1;
  108. case 0x0153: if (bytes && length >= 1) *bytes = 0x9c; return 1;
  109. case 0x017e: if (bytes && length >= 1) *bytes = 0x9e; return 1;
  110. case 0x0178: if (bytes && length >= 1) *bytes = 0x9f; return 1;
  111. default: return 0;
  112. }
  113. }
  114. int Windows1252Encoding::queryConvert(const unsigned char* bytes, int length) const
  115. {
  116. if (1 <= length)
  117. return _charMap[*bytes];
  118. else
  119. return -1;
  120. }
  121. int Windows1252Encoding::sequenceLength(const unsigned char* bytes, int length) const
  122. {
  123. return 1;
  124. }
  125. } // namespace Poco