| 1 |
/* |
| 2 |
* Copyright (C) 1999-2001, 2008, 2011, 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 Library General Public |
| 7 |
* License as published by the Free Software Foundation; either version 2 |
| 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 |
* Library General Public License for more details. |
| 14 |
* |
| 15 |
* You should have received a copy of the GNU Library General Public |
| 16 |
* License along with the GNU LIBICONV Library; see the file COPYING.LIB. |
| 17 |
* If not, see <http://www.gnu.org/licenses/>. |
| 18 |
*/ |
| 19 |
|
| 20 |
/* |
| 21 |
* UTF-32 |
| 22 |
*/ |
| 23 |
|
| 24 |
/* Specification: Unicode 3.1 Standard Annex #19 */ |
| 25 |
|
| 26 |
/* Here we accept FFFE0000/0000FEFF marks as endianness indicators |
| 27 |
everywhere in the stream, not just at the beginning. (This is contrary |
| 28 |
to what #19 D36c specifies, but it allows concatenation of byte |
| 29 |
sequences to work flawlessly, while disagreeing with #19 behaviour |
| 30 |
only for strings containing U+FEFF characters, which is quite rare.) |
| 31 |
The default is big-endian. */ |
| 32 |
/* The state is 0 if big-endian, 1 if little-endian. */ |
| 33 |
static int |
| 34 |
utf32_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n) |
| 35 |
{ |
| 36 |
state_t state = conv->istate; |
| 37 |
int count = 0; |
| 38 |
for (; n >= 4 && count <= RET_COUNT_MAX && count <= INT_MAX-4;) { |
| 39 |
ucs4_t wc = (state |
| 40 |
? s[0] + (s[1] << 8) + (s[2] << 16) + (s[3] << 24) |
| 41 |
: (s[0] << 24) + (s[1] << 16) + (s[2] << 8) + s[3]); |
| 42 |
if (wc == 0x0000feff) { |
| 43 |
} else if (wc == 0xfffe0000u) { |
| 44 |
state ^= 1; |
| 45 |
} else { |
| 46 |
if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000)) { |
| 47 |
*pwc = wc; |
| 48 |
conv->istate = state; |
| 49 |
return count+4; |
| 50 |
} else { |
| 51 |
conv->istate = state; |
| 52 |
return RET_SHIFT_ILSEQ(count); |
| 53 |
} |
| 54 |
} |
| 55 |
s += 4; n -= 4; count += 4; |
| 56 |
} |
| 57 |
conv->istate = state; |
| 58 |
return RET_TOOFEW(count); |
| 59 |
} |
| 60 |
|
| 61 |
/* We output UTF-32 in big-endian order, with byte-order mark. */ |
| 62 |
/* The state is 0 at the beginning, 1 after the BOM has been written. */ |
| 63 |
static int |
| 64 |
utf32_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n) |
| 65 |
{ |
| 66 |
if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000)) { |
| 67 |
int count = 0; |
| 68 |
if (!conv->ostate) { |
| 69 |
if (n >= 4) { |
| 70 |
r[0] = 0x00; |
| 71 |
r[1] = 0x00; |
| 72 |
r[2] = 0xFE; |
| 73 |
r[3] = 0xFF; |
| 74 |
r += 4; n -= 4; count += 4; |
| 75 |
} else |
| 76 |
return RET_TOOSMALL; |
| 77 |
} |
| 78 |
if (wc < 0x110000) { |
| 79 |
if (n >= 4) { |
| 80 |
r[0] = 0; |
| 81 |
r[1] = (unsigned char) (wc >> 16); |
| 82 |
r[2] = (unsigned char) (wc >> 8); |
| 83 |
r[3] = (unsigned char) wc; |
| 84 |
conv->ostate = 1; |
| 85 |
return count+4; |
| 86 |
} else |
| 87 |
return RET_TOOSMALL; |
| 88 |
} |
| 89 |
} |
| 90 |
return RET_ILUNI; |
| 91 |
} |