1 /* Conversion module for UTF-16.
2 Copyright (C) 1999, 2000 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Ulrich Drepper <drepper@cygnus.com>, 1999.
6 The GNU C Library is free software; you can redistribute it and/or
7 modify it under the terms of the GNU Library General Public License as
8 published by the Free Software Foundation; either version 2 of the
9 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 Library General Public License for more details.
16 You should have received a copy of the GNU Library General Public
17 License along with the GNU C Library; see the file COPYING.LIB. If not,
18 write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
19 Boston, MA 02111-1307, USA. */
29 /* This is the Byte Order Mark character (BOM). */
31 /* And in the other byte order. */
35 /* Definitions used in the body of the `gconv' function. */
36 #define FROM_LOOP from_utf16_loop
37 #define TO_LOOP to_utf16_loop
40 #define MIN_NEEDED_FROM 2
41 #define MAX_NEEDED_FROM 4
42 #define MIN_NEEDED_TO 4
43 #define FROM_DIRECTION (dir == from_utf16)
44 #define PREPARE_LOOP \
45 enum direction dir = ((struct utf16_data *) step->__data)->dir; \
46 enum variant var = ((struct utf16_data *) step->__data)->var; \
47 int swap = ((struct utf16_data *) step->__data)->swap; \
48 if (FROM_DIRECTION || var == UTF_16) \
50 if (data->__invocation_counter == 0) \
52 /* We have to find out which byte order the file is encoded in. */ \
53 if (inptr + 2 > inend) \
54 return __GCONV_EMPTY_INPUT; \
56 if (get16u (inptr) == BOM) \
57 /* Simply ignore the BOM character. */ \
59 else if (get16u (inptr) == BOM_OE) \
61 ((struct utf16_data *) step->__data)->swap = 1; \
66 else if (!FROM_DIRECTION && var == UTF_16 && !data->__internal_use \
67 && data->__invocation_counter == 0) \
69 /* Emit the Byte Order Mark. */ \
70 if (__builtin_expect (outbuf + 2 > outend, 0)) \
71 return __GCONV_FULL_OUTPUT; \
73 put16u (outbuf, BOM); \
76 #define EXTRA_LOOP_ARGS , var, swap
79 /* Direction of the transformation. */
104 gconv_init (struct __gconv_step *step)
106 /* Determine which direction. */
107 struct utf16_data *new_data;
108 enum direction dir = illegal_dir;
109 enum variant var = illegal_var;
112 if (__strcasecmp (step->__from_name, "UTF-16") == 0)
117 else if (__strcasecmp (step->__to_name, "UTF-16") == 0)
122 else if (__strcasecmp (step->__from_name, "UTF-16BE") == 0)
127 else if (__strcasecmp (step->__to_name, "UTF-16BE") == 0)
132 else if (__strcasecmp (step->__from_name, "UTF-16LE") == 0)
137 else if (__strcasecmp (step->__to_name, "UTF-16LE") == 0)
143 result = __GCONV_NOCONV;
144 if (__builtin_expect (dir, to_utf16) != illegal_dir)
146 new_data = (struct utf16_data *) malloc (sizeof (struct utf16_data));
148 result = __GCONV_NOMEM;
149 if (new_data != NULL)
153 new_data->swap = ((var == UTF_16LE && BYTE_ORDER == BIG_ENDIAN)
155 && BYTE_ORDER == LITTLE_ENDIAN));
156 step->__data = new_data;
158 if (dir == from_utf16)
160 step->__min_needed_from = MIN_NEEDED_FROM;
161 step->__max_needed_from = MIN_NEEDED_FROM;
162 step->__min_needed_to = MIN_NEEDED_TO;
163 step->__max_needed_to = MIN_NEEDED_TO;
167 step->__min_needed_from = MIN_NEEDED_TO;
168 step->__max_needed_from = MIN_NEEDED_TO;
169 step->__min_needed_to = MIN_NEEDED_FROM;
170 step->__max_needed_to = MIN_NEEDED_FROM;
173 step->__stateful = 0;
184 gconv_end (struct __gconv_step *data)
190 /* Convert from the internal (UCS4-like) format to UTF-16. */
191 #define MIN_NEEDED_INPUT MIN_NEEDED_TO
192 #define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM
193 #define MAX_NEEDED_OUTPUT MAX_NEEDED_FROM
194 #define LOOPFCT TO_LOOP
197 uint32_t c = get32 (inptr); \
201 if (__builtin_expect (c, 0) >= 0x10000) \
203 if (__builtin_expect (c, 0) >= 0x110000) \
205 if (step_data->__trans.__trans_fct != NULL) \
207 result = DL_CALL_FCT (step_data->__trans.__trans_fct, \
208 (step, step_data, *inptrp, &inptr, \
209 inend, &outbuf, irreversible)); \
210 if (result != __GCONV_OK) \
213 else if (! ignore_errors_p ()) \
215 /* This is an illegal character. */ \
216 result = __GCONV_ILLEGAL_INPUT; \
227 /* Generate a surrogate character. */ \
228 if (__builtin_expect (outptr + 4 > outend, 0)) \
230 /* Overflow in the output buffer. */ \
231 result = __GCONV_FULL_OUTPUT; \
235 put16 (outptr, bswap_16 (0xd7c0 + (c >> 10))); \
237 put16 (outptr, bswap_16 (0xdc00 + (c & 0x3ff))); \
240 put16 (outptr, bswap_16 (c)); \
244 if (__builtin_expect (c, 0) >= 0x10000) \
246 if (__builtin_expect (c, 0) >= 0x110000) \
248 if (step_data->__trans.__trans_fct != NULL) \
250 result = DL_CALL_FCT (step_data->__trans.__trans_fct, \
251 (step, step_data, *inptrp, &inptr, \
252 inend, &outbuf, irreversible)); \
253 if (result != __GCONV_OK) \
256 else if (! ignore_errors_p ()) \
258 /* This is an illegal character. */ \
259 result = __GCONV_ILLEGAL_INPUT; \
270 /* Generate a surrogate character. */ \
271 if (__builtin_expect (outptr + 4 > outend, 0)) \
273 /* Overflow in the output buffer. */ \
274 result = __GCONV_FULL_OUTPUT; \
278 put16 (outptr, 0xd7c0 + (c >> 10)); \
280 put16 (outptr, 0xdc00 + (c & 0x3ff)); \
288 #define LOOP_NEED_FLAGS
289 #define EXTRA_LOOP_DECLS \
290 , enum variant var, int swap
291 #include <iconv/loop.c>
294 /* Convert from UTF-16 to the internal (UCS4-like) format. */
295 #define MIN_NEEDED_INPUT MIN_NEEDED_FROM
296 #define MAX_NEEDED_INPUT MAX_NEEDED_FROM
297 #define MIN_NEEDED_OUTPUT MIN_NEEDED_TO
298 #define LOOPFCT FROM_LOOP
301 uint16_t u1 = get16 (inptr); \
305 u1 = bswap_16 (u1); \
307 if (__builtin_expect (u1, 0) < 0xd800 || u1 > 0xdfff) \
309 /* No surrogate. */ \
310 put32 (outptr, u1); \
317 /* It's a surrogate character. At least the first word says \
319 if (__builtin_expect (inptr + 4 > inend, 0)) \
321 /* We don't have enough input for another complete input \
323 result = __GCONV_INCOMPLETE_INPUT; \
328 u2 = bswap_16 (get16 (inptr)); \
329 if (__builtin_expect (u2, 0xdc00) < 0xdc00 \
330 || __builtin_expect (u2, 0xdc00) >= 0xdfff) \
332 /* This is no valid second word for a surrogate. */ \
333 if (! ignore_errors_p ()) \
336 result = __GCONV_ILLEGAL_INPUT; \
344 put32 (outptr, ((u1 - 0xd7c0) << 10) + (u2 - 0xdc00)); \
350 if (__builtin_expect (u1, 0) < 0xd800 || u1 > 0xdfff) \
352 /* No surrogate. */ \
353 put32 (outptr, u1); \
360 /* It's a surrogate character. At least the first word says \
362 if (__builtin_expect (inptr + 4 > inend, 0)) \
364 /* We don't have enough input for another complete input \
366 result = __GCONV_INCOMPLETE_INPUT; \
371 u2 = get16 (inptr); \
372 if (__builtin_expect (u2, 0xdc00) < 0xdc00 \
373 || __builtin_expect (u2, 0xdc00) >= 0xdfff) \
375 /* This is no valid second word for a surrogate. */ \
376 if (! ignore_errors_p ()) \
379 result = __GCONV_ILLEGAL_INPUT; \
387 put32 (outptr, ((u1 - 0xd7c0) << 10) + (u2 - 0xdc00)); \
393 #define LOOP_NEED_FLAGS
394 #define EXTRA_LOOP_DECLS \
395 , enum variant var, int swap
396 #include <iconv/loop.c>
399 /* Now define the toplevel functions. */
400 #include <iconv/skeleton.c>