Update.
[glibc.git] / iconvdata / ibm939.c
blobaae08ba3d10326b9a7d96de792a10551002b77f9
1 /* Conversion to and from IBM939.
2 Copyright (C) 2000-2002 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Masahide Washizawa <washi@yamato.ibm.co.jp>, 2000.
6 The GNU C Library is free software; you can redistribute it and/or
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.
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
14 Lesser General Public License for more details.
16 You should have received a copy of the GNU Lesser General Public
17 License along with the GNU C Library; if not, write to the Free
18 Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
19 02111-1307 USA. */
21 #include <dlfcn.h>
22 #include <stdint.h>
23 #include <wchar.h>
24 #include <byteswap.h>
25 #include "ibm939.h"
27 /* The shift sequences for this charset (it does not use ESC). */
28 #define SI 0x0F /* Shift In, host code to turn DBCS off. */
29 #define SO 0x0E /* Shift Out, host code to turn DBCS on. */
31 /* Definitions used in the body of the `gconv' function. */
32 #define CHARSET_NAME "IBM939//"
33 #define FROM_LOOP from_ibm939
34 #define TO_LOOP to_ibm939
35 #define FROM_LOOP_MIN_NEEDED_FROM 1
36 #define FROM_LOOP_MAX_NEEDED_FROM 2
37 #define FROM_LOOP_MIN_NEEDED_TO 4
38 #define FROM_LOOP_MAX_NEEDED_TO 4
39 #define TO_LOOP_MIN_NEEDED_FROM 4
40 #define TO_LOOP_MAX_NEEDED_FROM 4
41 #define TO_LOOP_MIN_NEEDED_TO 1
42 #define TO_LOOP_MAX_NEEDED_TO 3
43 #define PREPARE_LOOP \
44 int save_curcs; \
45 int *curcsp = &data->__statep->__count;
46 #define EXTRA_LOOP_ARGS , curcsp
48 /* Definitions of initialization and destructor function. */
49 #define DEFINE_INIT 1
50 #define DEFINE_FINI 1
53 /* Since this is a stateful encoding we have to provide code which resets
54 the output state to the initial state. This has to be done during the
55 flushing. */
56 #define EMIT_SHIFT_TO_INIT \
57 if ((data->__statep->__count & ~7) != sb) \
58 { \
59 if (FROM_DIRECTION) \
60 data->__statep->__count &= 7; \
61 else \
62 { \
63 /* We are not in the initial state. To switch back we have \
64 to emit `SI'. */ \
65 if (__builtin_expect (outbuf >= outend, 0)) \
66 /* We don't have enough room in the output buffer. */ \
67 status = __GCONV_FULL_OUTPUT; \
68 else \
69 { \
70 /* Write out the shift sequence. */ \
71 *outbuf++ = SI; \
72 data->__statep->__count &= 7; \
73 } \
74 } \
78 /* Since we might have to reset input pointer we must be able to save
79 and retore the state. */
80 #define SAVE_RESET_STATE(Save) \
81 if (Save) \
82 save_curcs = *curcsp; \
83 else \
84 *curcsp = save_curcs
87 /* Current codeset type. */
88 enum
90 sb = 0,
91 db = 64
94 /* First, define the conversion function from IBM-939 to UCS4. */
95 #define MIN_NEEDED_INPUT FROM_LOOP_MIN_NEEDED_FROM
96 #define MAX_NEEDED_INPUT FROM_LOOP_MAX_NEEDED_FROM
97 #define MIN_NEEDED_OUTPUT FROM_LOOP_MIN_NEEDED_TO
98 #define MAX_NEEDED_OUTPUT FROM_LOOP_MAX_NEEDED_TO
99 #define LOOPFCT FROM_LOOP
100 #define BODY \
102 uint32_t ch = *inptr; \
103 uint32_t res; \
105 if (__builtin_expect (ch, 0) == SO) \
107 /* Shift OUT, change to DBCS converter. */ \
108 if (curcs == db) \
110 result = __GCONV_ILLEGAL_INPUT; \
111 break; \
113 curcs = db; \
114 ++inptr; \
115 continue; \
117 else if (__builtin_expect (ch, 0) == SI) \
119 /* Shift IN, change to SBCS converter. */ \
120 if (curcs == sb) \
122 result = __GCONV_ILLEGAL_INPUT; \
123 break; \
125 curcs = sb; \
126 ++inptr; \
127 continue; \
130 if (curcs == sb) \
132 /* Use the IBM939 table for single byte. */ \
133 res = __ibm939sb_to_ucs4[ch]; \
134 if (__builtin_expect (res == L'\0', 0) && ch != '\0') \
136 /* This is an illegal character. */ \
137 STANDARD_FROM_LOOP_ERR_HANDLER (1); \
139 else \
141 put32 (outptr, res); \
142 outptr += 4; \
144 ++inptr; \
146 else \
148 /* Use the IBM939 table for double byte. */ \
149 const struct gap *rp2 = __ibm939db_to_ucs4_idx; \
151 assert (curcs == db); \
153 if (__builtin_expect (inptr + 1 >= inend, 0)) \
155 /* The second character is not available. Store the \
156 intermediate result. */ \
157 result = __GCONV_INCOMPLETE_INPUT; \
158 break; \
161 ch = (ch * 0x100) + inptr[1]; \
162 while (ch > rp2->end) \
163 ++rp2; \
165 if (__builtin_expect (rp2 == NULL, 0) \
166 || __builtin_expect (ch < rp2->start, 0) \
167 || (res = __ibm939db_to_ucs4[ch + rp2->idx], \
168 __builtin_expect (res, L'\1') == L'\0' && ch != '\0')) \
170 /* This is an illegal character. */ \
171 STANDARD_FROM_LOOP_ERR_HANDLER (2); \
173 else \
175 put32 (outptr, res); \
176 outptr += 4; \
178 inptr += 2; \
181 #define LOOP_NEED_FLAGS
182 #define EXTRA_LOOP_DECLS , int *curcsp
183 #define INIT_PARAMS int curcs = *curcsp & ~7
184 #define UPDATE_PARAMS *curcsp = curcs
185 #include <iconv/loop.c>
187 /* Next, define the other direction */
188 #define MIN_NEEDED_INPUT TO_LOOP_MIN_NEEDED_FROM
189 #define MAX_NEEDED_INPUT TO_LOOP_MAX_NEEDED_FROM
190 #define MIN_NEEDED_OUTPUT TO_LOOP_MIN_NEEDED_TO
191 #define MAX_NEEDED_OUTPUT TO_LOOP_MAX_NEEDED_TO
192 #define LOOPFCT TO_LOOP
193 #define BODY \
195 uint32_t ch = get32 (inptr); \
196 const struct gap *rp1 = __ucs4_to_ibm939sb_idx; \
197 const struct gap *rp2 = __ucs4_to_ibm939db_idx; \
198 const char *cp; \
200 if (__builtin_expect (ch >= 0xffff, 0)) \
202 UNICODE_TAG_HANDLER (ch, 4); \
203 rp1 = NULL; \
204 rp2 = NULL; \
206 else \
207 while (ch > rp1->end) \
208 ++rp1; \
210 /* Use the UCS4 table for single byte. */ \
211 if (__builtin_expect (rp1 == NULL, 0) \
212 || __builtin_expect (ch < rp1->start, 0) \
213 || (cp = __ucs4_to_ibm939sb[ch + rp1->idx], \
214 __builtin_expect (cp[0], L'\1') == L'\0' && ch != '\0')) \
216 /* Use the UCS4 table for double byte. */ \
217 while (ch > rp2->end) \
218 ++rp2; \
220 if (__builtin_expect (rp2 == NULL, 0) \
221 || __builtin_expect (ch < rp2->start, 0) \
222 || (cp = __ucs4_to_ibm939db[ch + rp2->idx], \
223 __builtin_expect (cp[0], L'\1')==L'\0' && ch != '\0')) \
225 /* This is an illegal character. */ \
226 STANDARD_TO_LOOP_ERR_HANDLER (4); \
228 else \
230 if (curcs == sb) \
232 if (__builtin_expect (outptr + 1 > outend, 0)) \
234 result = __GCONV_FULL_OUTPUT; \
235 break; \
237 *outptr++ = SO; \
238 curcs = db; \
241 if (__builtin_expect (outptr + 2 > outend, 0)) \
243 result = __GCONV_FULL_OUTPUT; \
244 break; \
246 *outptr++ = cp[0]; \
247 *outptr++ = cp[1]; \
250 else \
252 if (curcs == db) \
254 if (__builtin_expect (outptr + 1 > outend, 0)) \
256 result = __GCONV_FULL_OUTPUT; \
257 break; \
259 *outptr++ = SI; \
262 if (__builtin_expect (outptr + 1 > outend, 0)) \
264 result = __GCONV_FULL_OUTPUT; \
265 break; \
267 if (ch == 0x7e) \
268 *outptr++ = 0xa0; \
269 else if (ch == 0x5c) \
270 *outptr++ = 0xb2; \
271 else \
272 *outptr++ = cp[0]; \
273 curcs = sb; \
276 /* Now that we wrote the output increment the input pointer. */ \
277 inptr += 4; \
279 #define LOOP_NEED_FLAGS
280 #define EXTRA_LOOP_DECLS , int *curcsp
281 #define INIT_PARAMS int curcs = *curcsp & ~7
282 #define UPDATE_PARAMS *curcsp = curcs
283 #include <iconv/loop.c>
285 /* Now define the toplevel functions. */
286 #include <iconv/skeleton.c>