lib/java.h

   1 /*
   2  * Copyright (C) 1999-2002, 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 /*
  21  * JAVA
  22  * This is ISO 8859-1 with \uXXXX escape sequences, denoting Unicode BMP
  23  * characters. Consecutive pairs of \uXXXX escape sequences in the surrogate
  24  * range, as in UTF-16, denote Unicode characters outside the BMP.
  25  */
  26
  27 static int
  28 java_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
  29 {
  30   unsigned char c;
  31   ucs4_t wc, wc2;
  32   int i;
  33
  34   c = s[0];
  35   if (c != '\\') {
  36     *pwc = c;
  37     return 1;
  38   }
  39   if (n < 2)
  40     return RET_TOOFEW(0);
  41   if (s[1] != 'u')
  42     goto simply_backslash;
  43   wc = 0;
  44   for (i = 2; i < 6; i++) {
  45     if (n <= i)
  46       return RET_TOOFEW(0);
  47     c = s[i];
  48     if (c >= '0' && c <= '9')
  49       c -= '0';
  50     else if (c >= 'A' && c <= 'Z')
  51       c -= 'A'-10;
  52     else if (c >= 'a' && c <= 'z')
  53       c -= 'a'-10;
  54     else
  55       goto simply_backslash;
  56     wc |= (ucs4_t) c << (4 * (5-i));
  57   }
  58   if (!(wc >= 0xd800 && wc < 0xe000)) {
  59     *pwc = wc;
  60     return 6;
  61   }
  62   if (wc >= 0xdc00)
  63     goto simply_backslash;
  64   if (n < 7)
  65     return RET_TOOFEW(0);
  66   if (s[6] != '\\')
  67     goto simply_backslash;
  68   if (n < 8)
  69     return RET_TOOFEW(0);
  70   if (s[7] != 'u')
  71     goto simply_backslash;
  72   wc2 = 0;
  73   for (i = 8; i < 12; i++) {
  74     if (n <= i)
  75       return RET_TOOFEW(0);
  76     c = s[i];
  77     if (c >= '0' && c <= '9')
  78       c -= '0';
  79     else if (c >= 'A' && c <= 'Z')
  80       c -= 'A'-10;
  81     else if (c >= 'a' && c <= 'z')
  82       c -= 'a'-10;
  83     else
  84       goto simply_backslash;
  85     wc2 |= (ucs4_t) c << (4 * (11-i));
  86   }
  87   if (!(wc2 >= 0xdc00 && wc2 < 0xe000))
  88     goto simply_backslash;
  89   *pwc = 0x10000 + ((wc - 0xd800) << 10) + (wc2 - 0xdc00);
  90   return 12;
  91 simply_backslash:
  92   *pwc = '\\';
  93   return 1;
  94 }
  95
  96 static int
  97 java_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
  98 {
  99   if (wc < 0x80) {
 100     *r = wc;
 101     return 1;
 102   } else if (wc < 0x10000) {
 103     if (n >= 6) {
 104       unsigned int i;
 105       r[0] = '\\';
 106       r[1] = 'u';
 107       i = (wc >> 12) & 0x0f; r[2] = (i < 10 ? '0'+i : 'a'-10+i);
 108       i = (wc >> 8) & 0x0f;  r[3] = (i < 10 ? '0'+i : 'a'-10+i);
 109       i = (wc >> 4) & 0x0f;  r[4] = (i < 10 ? '0'+i : 'a'-10+i);
 110       i = wc & 0x0f;         r[5] = (i < 10 ? '0'+i : 'a'-10+i);
 111       return 6;
 112     } else
 113       return RET_TOOSMALL;
 114   } else if (wc < 0x110000) {
 115     if (n >= 12) {
 116       ucs4_t wc1 = 0xd800 + ((wc - 0x10000) >> 10);
 117       ucs4_t wc2 = 0xdc00 + ((wc - 0x10000) & 0x3ff);
 118       unsigned int i;
 119       r[0] = '\\';
 120       r[1] = 'u';
 121       i = (wc1 >> 12) & 0x0f; r[2] = (i < 10 ? '0'+i : 'a'-10+i);
 122       i = (wc1 >> 8) & 0x0f;  r[3] = (i < 10 ? '0'+i : 'a'-10+i);
 123       i = (wc1 >> 4) & 0x0f;  r[4] = (i < 10 ? '0'+i : 'a'-10+i);
 124       i = wc1 & 0x0f;         r[5] = (i < 10 ? '0'+i : 'a'-10+i);
 125       r[6] = '\\';
 126       r[7] = 'u';
 127       i = (wc2 >> 12) & 0x0f; r[8] = (i < 10 ? '0'+i : 'a'-10+i);
 128       i = (wc2 >> 8) & 0x0f;  r[9] = (i < 10 ? '0'+i : 'a'-10+i);
 129       i = (wc2 >> 4) & 0x0f; r[10] = (i < 10 ? '0'+i : 'a'-10+i);
 130       i = wc2 & 0x0f;        r[11] = (i < 10 ? '0'+i : 'a'-10+i);
 131       return 12;
 132     } else
 133       return RET_TOOSMALL;
 134   }
 135   return RET_ILUNI;
 136 }