]>
Commit | Line | Data |
---|---|---|
40b4c81d | 1 | /* Mapping tables for EUC-CN handling. |
e438a468 | 2 | Copyright (C) 1998, 1999, 2000-2002 Free Software Foundation, Inc. |
40b4c81d UD |
3 | This file is part of the GNU C Library. |
4 | Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998. | |
5 | ||
6 | The GNU C Library is free software; you can redistribute it and/or | |
41bdb6e2 AJ |
7 | modify it under the terms of the GNU Lesser General Public |
8 | License as published by the Free Software Foundation; either | |
9 | version 2.1 of the License, or (at your option) any later version. | |
40b4c81d UD |
10 | |
11 | The GNU C Library is distributed in the hope that it will be useful, | |
12 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
13 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
41bdb6e2 | 14 | Lesser General Public License for more details. |
40b4c81d | 15 | |
41bdb6e2 | 16 | You should have received a copy of the GNU Lesser General Public |
59ba27a6 PE |
17 | License along with the GNU C Library; if not, see |
18 | <http://www.gnu.org/licenses/>. */ | |
40b4c81d | 19 | |
55985355 | 20 | #include <dlfcn.h> |
40b4c81d | 21 | #include <gb2312.h> |
8619129f | 22 | #include <stdint.h> |
40b4c81d | 23 | |
8619129f | 24 | /* Definitions used in the body of the `gconv' function. */ |
9b26f5c4 | 25 | #define CHARSET_NAME "EUC-CN//" |
8619129f UD |
26 | #define FROM_LOOP from_euc_cn |
27 | #define TO_LOOP to_euc_cn | |
28 | #define DEFINE_INIT 1 | |
29 | #define DEFINE_FINI 1 | |
30 | #define MIN_NEEDED_FROM 1 | |
31 | #define MAX_NEEDED_FROM 2 | |
32 | #define MIN_NEEDED_TO 4 | |
33 | ||
34 | ||
ffa156af | 35 | /* First define the conversion function from EUC-CN to UCS4. */ |
8619129f UD |
36 | #define MIN_NEEDED_INPUT MIN_NEEDED_FROM |
37 | #define MAX_NEEDED_INPUT MAX_NEEDED_FROM | |
38 | #define MIN_NEEDED_OUTPUT MIN_NEEDED_TO | |
39 | #define LOOPFCT FROM_LOOP | |
40 | #define BODY \ | |
41 | { \ | |
42 | uint32_t ch = *inptr; \ | |
43 | \ | |
44 | if (ch <= 0x7f) \ | |
45 | ++inptr; \ | |
46 | else \ | |
db2d05f9 UD |
47 | if ((__builtin_expect (ch <= 0xa0, 0) && ch != 0x8e && ch != 0x8f) \ |
48 | || __builtin_expect (ch > 0xfe, 0)) \ | |
8619129f UD |
49 | { \ |
50 | /* This is illegal. */ \ | |
e438a468 | 51 | STANDARD_FROM_LOOP_ERR_HANDLER (1); \ |
8619129f UD |
52 | } \ |
53 | else \ | |
54 | { \ | |
55 | /* Two or more byte character. First test whether the \ | |
f9ad060c | 56 | next byte is also available. */ \ |
bc900b11 | 57 | const unsigned char *endp; \ |
8619129f | 58 | \ |
55985355 | 59 | if (__builtin_expect (inptr + 1 >= inend, 0)) \ |
8619129f UD |
60 | { \ |
61 | /* The second character is not available. Store \ | |
62 | the intermediate result. */ \ | |
d64b6ad0 | 63 | result = __GCONV_INCOMPLETE_INPUT; \ |
8619129f UD |
64 | break; \ |
65 | } \ | |
66 | \ | |
67 | ch = inptr[1]; \ | |
68 | \ | |
69 | /* All second bytes of a multibyte character must be >= 0xa1. */ \ | |
db2d05f9 | 70 | if (__builtin_expect (ch < 0xa1, 0)) \ |
e438a468 | 71 | STANDARD_FROM_LOOP_ERR_HANDLER (1); \ |
8619129f UD |
72 | \ |
73 | /* This is code set 1: GB 2312-80. */ \ | |
74 | endp = inptr; \ | |
75 | \ | |
76 | ch = gb2312_to_ucs4 (&endp, 2, 0x80); \ | |
e438a468 | 77 | if (__builtin_expect (ch == __UNKNOWN_10646_CHAR, 0)) \ |
8619129f UD |
78 | { \ |
79 | /* This is an illegal character. */ \ | |
e438a468 | 80 | STANDARD_FROM_LOOP_ERR_HANDLER (2); \ |
8619129f UD |
81 | } \ |
82 | \ | |
83 | inptr += 2; \ | |
84 | } \ | |
85 | \ | |
77e1d15a UD |
86 | put32 (outptr, ch); \ |
87 | outptr += 4; \ | |
8619129f | 88 | } |
55985355 | 89 | #define LOOP_NEED_FLAGS |
f9ad060c UD |
90 | #define ONEBYTE_BODY \ |
91 | { \ | |
92 | if (c < 0x80) \ | |
93 | return c; \ | |
94 | else \ | |
95 | return WEOF; \ | |
96 | } | |
8619129f UD |
97 | #include <iconv/loop.c> |
98 | ||
99 | ||
100 | /* Next, define the other direction. */ | |
101 | #define MIN_NEEDED_INPUT MIN_NEEDED_TO | |
102 | #define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM | |
103 | #define MAX_NEEDED_OUTPUT MAX_NEEDED_FROM | |
104 | #define LOOPFCT TO_LOOP | |
105 | #define BODY \ | |
106 | { \ | |
77e1d15a | 107 | uint32_t ch = get32 (inptr); \ |
8619129f UD |
108 | \ |
109 | if (ch <= L'\x7f') \ | |
110 | /* It's plain ASCII. */ \ | |
111 | *outptr++ = (unsigned char) ch; \ | |
112 | else \ | |
113 | { \ | |
114 | size_t found; \ | |
115 | \ | |
55985355 UD |
116 | found = ucs4_to_gb2312 (ch, outptr, outend - outptr); \ |
117 | if (__builtin_expect (found, 1) != 0) \ | |
8619129f | 118 | { \ |
89301d68 | 119 | if (__builtin_expect (found, 0) == __UNKNOWN_10646_CHAR) \ |
8619129f | 120 | { \ |
601d2942 UD |
121 | UNICODE_TAG_HANDLER (ch, 4); \ |
122 | \ | |
8619129f | 123 | /* Illegal character. */ \ |
e438a468 | 124 | STANDARD_TO_LOOP_ERR_HANDLER (4); \ |
8619129f UD |
125 | } \ |
126 | \ | |
127 | /* It's a GB 2312 character, adjust it for EUC-CN. */ \ | |
128 | *outptr++ += 0x80; \ | |
129 | *outptr++ += 0x80; \ | |
130 | } \ | |
131 | else \ | |
132 | { \ | |
133 | /* We ran out of space. */ \ | |
d64b6ad0 | 134 | result = __GCONV_FULL_OUTPUT; \ |
8619129f UD |
135 | break; \ |
136 | } \ | |
137 | } \ | |
138 | inptr += 4; \ | |
139 | } | |
55985355 | 140 | #define LOOP_NEED_FLAGS |
8619129f UD |
141 | #include <iconv/loop.c> |
142 | ||
143 | ||
144 | /* Now define the toplevel functions. */ | |
145 | #include <iconv/skeleton.c> |