cp950ext.h 6.2 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160
  1. /*
  2. * Copyright (C) 1999-2001, 2012, 2016 Free Software Foundation, Inc.
  3. * This file is part of the GNU LIBICONV Library.
  4. *
  5. * The GNU LIBICONV Library is free software; you can redistribute it
  6. * and/or modify it under the terms of the GNU Lesser General Public
  7. * License as published by the Free Software Foundation; either version 2.1
  8. * of the License, or (at your option) any later version.
  9. *
  10. * The GNU LIBICONV Library is distributed in the hope that it will be
  11. * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
  12. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  13. * Lesser General Public License for more details.
  14. *
  15. * You should have received a copy of the GNU Lesser General Public
  16. * License along with the GNU LIBICONV Library; see the file COPYING.LIB.
  17. * If not, see <https://www.gnu.org/licenses/>.
  18. */
  19. /*
  20. * CP950 extensions
  21. */
  22. static const unsigned short cp950ext_2uni_pagef9[157-116] = {
  23. /* 0xf9 */
  24. 0x7881, 0x92b9, 0x88cf, 0x58bb,
  25. 0x6052, 0x7ca7, 0x5afa, 0x2554, 0x2566, 0x2557, 0x2560, 0x256c,
  26. 0x2563, 0x255a, 0x2569, 0x255d, 0x2552, 0x2564, 0x2555, 0x255e,
  27. 0x256a, 0x2561, 0x2558, 0x2567, 0x255b, 0x2553, 0x2565, 0x2556,
  28. 0x255f, 0x256b, 0x2562, 0x2559, 0x2568, 0x255c, 0x2551, 0x2550,
  29. 0x256d, 0x256e, 0x2570, 0x256f, 0x2593,
  30. };
  31. static int
  32. cp950ext_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
  33. {
  34. unsigned char c1 = s[0];
  35. if ((c1 == 0xf9)) {
  36. if (n >= 2) {
  37. unsigned char c2 = s[1];
  38. if ((c2 >= 0x40 && c2 < 0x7f) || (c2 >= 0xa1 && c2 < 0xff)) {
  39. unsigned int i = 157 * (c1 - 0xa1) + (c2 - (c2 >= 0xa1 ? 0x62 : 0x40));
  40. unsigned short wc = 0xfffd;
  41. {
  42. if (i >= 13932 && i < 13973)
  43. wc = cp950ext_2uni_pagef9[i-13932];
  44. }
  45. if (wc != 0xfffd) {
  46. *pwc = (ucs4_t) wc;
  47. return 2;
  48. }
  49. }
  50. return RET_ILSEQ;
  51. }
  52. return RET_TOOFEW(0);
  53. }
  54. return RET_ILSEQ;
  55. }
  56. static const unsigned short cp950ext_2charset[41] = {
  57. 0xf9f9, 0xf9f8, 0xf9e6, 0xf9ef, 0xf9dd, 0xf9e8, 0xf9f1, 0xf9df,
  58. 0xf9ec, 0xf9f5, 0xf9e3, 0xf9ee, 0xf9f7, 0xf9e5, 0xf9e9, 0xf9f2,
  59. 0xf9e0, 0xf9eb, 0xf9f4, 0xf9e2, 0xf9e7, 0xf9f0, 0xf9de, 0xf9ed,
  60. 0xf9f6, 0xf9e4, 0xf9ea, 0xf9f3, 0xf9e1, 0xf9fa, 0xf9fb, 0xf9fd,
  61. 0xf9fc, 0xf9fe, 0xf9d9, 0xf9dc, 0xf9da, 0xf9d6, 0xf9db, 0xf9d8,
  62. 0xf9d7,
  63. };
  64. static const Summary16 cp950ext_uni2indx_page25[10] = {
  65. /* 0x2500 */
  66. { 0, 0x0000 }, { 0, 0x0000 }, { 0, 0x0000 }, { 0, 0x0000 },
  67. { 0, 0x0000 }, { 0, 0xffff }, { 16, 0xffff }, { 32, 0x0001 },
  68. { 33, 0x0000 }, { 33, 0x0008 },
  69. };
  70. static const Summary16 cp950ext_uni2indx_page58[12] = {
  71. /* 0x5800 */
  72. { 34, 0x0000 }, { 34, 0x0000 }, { 34, 0x0000 }, { 34, 0x0000 },
  73. { 34, 0x0000 }, { 34, 0x0000 }, { 34, 0x0000 }, { 34, 0x0000 },
  74. { 34, 0x0000 }, { 34, 0x0000 }, { 34, 0x0000 }, { 34, 0x0800 },
  75. };
  76. static const Summary16 cp950ext_uni2indx_page5a[16] = {
  77. /* 0x5a00 */
  78. { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0000 },
  79. { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0000 },
  80. { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0000 },
  81. { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0000 }, { 35, 0x0400 },
  82. };
  83. static const Summary16 cp950ext_uni2indx_page60[6] = {
  84. /* 0x6000 */
  85. { 36, 0x0000 }, { 36, 0x0000 }, { 36, 0x0000 }, { 36, 0x0000 },
  86. { 36, 0x0000 }, { 36, 0x0004 },
  87. };
  88. static const Summary16 cp950ext_uni2indx_page78[9] = {
  89. /* 0x7800 */
  90. { 37, 0x0000 }, { 37, 0x0000 }, { 37, 0x0000 }, { 37, 0x0000 },
  91. { 37, 0x0000 }, { 37, 0x0000 }, { 37, 0x0000 }, { 37, 0x0000 },
  92. { 37, 0x0002 },
  93. };
  94. static const Summary16 cp950ext_uni2indx_page7c[11] = {
  95. /* 0x7c00 */
  96. { 38, 0x0000 }, { 38, 0x0000 }, { 38, 0x0000 }, { 38, 0x0000 },
  97. { 38, 0x0000 }, { 38, 0x0000 }, { 38, 0x0000 }, { 38, 0x0000 },
  98. { 38, 0x0000 }, { 38, 0x0000 }, { 38, 0x0080 },
  99. };
  100. static const Summary16 cp950ext_uni2indx_page88[13] = {
  101. /* 0x8800 */
  102. { 39, 0x0000 }, { 39, 0x0000 }, { 39, 0x0000 }, { 39, 0x0000 },
  103. { 39, 0x0000 }, { 39, 0x0000 }, { 39, 0x0000 }, { 39, 0x0000 },
  104. { 39, 0x0000 }, { 39, 0x0000 }, { 39, 0x0000 }, { 39, 0x0000 },
  105. { 39, 0x8000 },
  106. };
  107. static const Summary16 cp950ext_uni2indx_page92[12] = {
  108. /* 0x9200 */
  109. { 40, 0x0000 }, { 40, 0x0000 }, { 40, 0x0000 }, { 40, 0x0000 },
  110. { 40, 0x0000 }, { 40, 0x0000 }, { 40, 0x0000 }, { 40, 0x0000 },
  111. { 40, 0x0000 }, { 40, 0x0000 }, { 40, 0x0000 }, { 40, 0x0200 },
  112. };
  113. static int
  114. cp950ext_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
  115. {
  116. if (n >= 2) {
  117. const Summary16 *summary = NULL;
  118. if (wc >= 0x2500 && wc < 0x25a0)
  119. summary = &cp950ext_uni2indx_page25[(wc>>4)-0x250];
  120. else if (wc >= 0x5800 && wc < 0x58c0)
  121. summary = &cp950ext_uni2indx_page58[(wc>>4)-0x580];
  122. else if (wc >= 0x5a00 && wc < 0x5b00)
  123. summary = &cp950ext_uni2indx_page5a[(wc>>4)-0x5a0];
  124. else if (wc >= 0x6000 && wc < 0x6060)
  125. summary = &cp950ext_uni2indx_page60[(wc>>4)-0x600];
  126. else if (wc >= 0x7800 && wc < 0x7890)
  127. summary = &cp950ext_uni2indx_page78[(wc>>4)-0x780];
  128. else if (wc >= 0x7c00 && wc < 0x7cb0)
  129. summary = &cp950ext_uni2indx_page7c[(wc>>4)-0x7c0];
  130. else if (wc >= 0x8800 && wc < 0x88d0)
  131. summary = &cp950ext_uni2indx_page88[(wc>>4)-0x880];
  132. else if (wc >= 0x9200 && wc < 0x92c0)
  133. summary = &cp950ext_uni2indx_page92[(wc>>4)-0x920];
  134. if (summary) {
  135. unsigned short used = summary->used;
  136. unsigned int i = wc & 0x0f;
  137. if (used & ((unsigned short) 1 << i)) {
  138. unsigned short c;
  139. /* Keep in 'used' only the bits 0..i-1. */
  140. used &= ((unsigned short) 1 << i) - 1;
  141. /* Add 'summary->indx' and the number of bits set in 'used'. */
  142. used = (used & 0x5555) + ((used & 0xaaaa) >> 1);
  143. used = (used & 0x3333) + ((used & 0xcccc) >> 2);
  144. used = (used & 0x0f0f) + ((used & 0xf0f0) >> 4);
  145. used = (used & 0x00ff) + (used >> 8);
  146. c = cp950ext_2charset[summary->indx + used];
  147. r[0] = (c >> 8); r[1] = (c & 0xff);
  148. return 2;
  149. }
  150. }
  151. return RET_ILUNI;
  152. }
  153. return RET_TOOSMALL;
  154. }