1 /* Skeleton for a converison module.
2 Copyright (C) 1998 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
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. */
21 /* This file can be included to provide definitions of several things
22 many modules have in common. It can be customized using the following
25 DEFINE_INIT define the default initializer. This requires the
26 following symbol to be defined.
28 CHARSET_NAME string with official name of the coded character
31 DEFINE_FINI define the default destructor function.
33 MIN_NEEDED_FROM minimal number of bytes needed for the from-charset.
34 MIN_NEEDED_TO likewise for the to-charset.
36 MAX_NEEDED_FROM maximal number of bytes needed for the from-charset.
37 This macro is optional, it defaults to MIN_NEEDED_FROM.
38 MAX_NEEDED_TO likewise for the to-charset.
40 DEFINE_DIRECTION_OBJECTS
41 two objects will be defined to be used when the
42 `gconv' function must only distinguish two
43 directions. This is implied by DEFINE_INIT.
44 If this macro is not defined the following
45 macro must be available.
47 FROM_DIRECTION this macro is supposed to return a value != 0
48 if we convert from the current character set,
49 otherwise it return 0.
51 EMIT_SHIFT_TO_INIT this symbol is optional. If it is defined it
52 defines some code which writes out a sequence
53 of characters which bring the current state into
56 FROM_LOOP name of the function implementing the conversion
57 from the current characters.
58 TO_LOOP likewise for the other direction
60 RESET_STATE in case of an error we must reset the state for
61 the rerun so this macro must be defined for
62 stateful encodings. It takes an argument which
63 is nonzero when saving.
65 RESET_INPUT_BUFFER If the input character sets allow this the macro
66 can be defined to reset the input buffer pointers
67 to cover only those characters up to the error.
69 FUNCTION_NAME if not set the conversion function is named `gconv'.
80 /* The direction objects. */
81 #if DEFINE_DIRECTION_OBJECTS || DEFINE_INIT
82 static int from_object;
85 # ifndef FROM_DIRECTION
86 # define FROM_DIRECTION step->data == &from_object
89 # ifndef FROM_DIRECTION
90 # error "FROM_DIRECTION must be provided if direction objects are not used"
95 /* How many bytes are needed at most for the from-charset. */
96 #ifndef MAX_NEEDED_FROM
97 # define MAX_NEEDED_FROM MIN_NEEDED_FROM
100 /* Same for the to-charset. */
101 #ifndef MAX_NEEDED_TO
102 # define MAX_NEEDED_TO MIN_NEEDED_TO
106 /* For conversions from a fixed width character sets to another fixed width
107 character set we we can define RESET_INPUT_BUFFER is necessary. */
108 #if !defined RESET_INPUT_BUFFER && !defined SAVE_RESET_STATE
109 # if MIN_NEEDED_FROM == MAX_NEEDED_FROM && MIN_NEEDED_TO == MAX_NEEDED_TO
110 /* We have to used these `if's here since the compiler cannot know that
111 (outbuf - outerr) is always divisible by MIN_NEEDED_TO. */
112 # define RESET_INPUT_BUFFER \
113 if (MIN_NEEDED_FROM % MIN_NEEDED_TO == 0) \
114 *inbuf -= (outbuf - outerr) * (MIN_NEEDED_FROM / MIN_NEEDED_TO); \
115 else if (MIN_NEEDED_TO % MIN_NEEDED_FROM == 0) \
116 *inbuf -= (outbuf - outerr) / (MIN_NEEDED_TO / MIN_NEEDED_FROM); \
118 *inbuf -= ((outbuf - outerr) / MIN_NEEDED_TO) * MIN_NEEDED_FROM
123 /* The default init function. It simply matches the name and initializes
124 the step data to point to one of the objects above. */
126 # ifndef CHARSET_NAME
127 # error "CHARSET_NAME not defined"
131 gconv_init (struct gconv_step *step)
133 /* Determine which direction. */
134 if (__strcasestr (step->from_name, CHARSET_NAME) != NULL)
135 step->data = &from_object;
136 else if (__strcasestr (step->to_name, CHARSET_NAME) != NULL)
137 step->data = &to_object;
141 step->min_needed_from = MIN_NEEDED_FROM;
142 step->max_needed_from = MAX_NEEDED_FROM;
143 step->min_needed_to = MIN_NEEDED_TO;
144 step->max_needed_to = MAX_NEEDED_TO;
157 /* The default destructor function does nothing in the moment and so
158 be define it at all. But we still provide the macro just in case
159 we need it some day. */
164 /* This is the actual conversion function. */
165 #ifndef FUNCTION_NAME
166 # define FUNCTION_NAME gconv
170 FUNCTION_NAME (struct gconv_step *step, struct gconv_step_data *data,
171 const char **inbuf, const char *inbufend, size_t *written,
174 struct gconv_step *next_step = step + 1;
175 struct gconv_step_data *next_data = data + 1;
176 gconv_fct fct = next_step->fct;
179 /* If the function is called with no input this means we have to reset
180 to the initial state. The possibly partly converted input is
184 /* Call the steps down the chain if there are any. */
189 #ifdef EMIT_SHIFT_TO_INIT
194 if (status == GCONV_OK)
196 /* Give the modules below the same chance. */
197 status = (*fct) (next_step, next_data, NULL, NULL, written, 1);
202 /* This variable is used to count the number of characters we
203 actually converted. */
204 size_t converted = 0;
205 size_t last_converted;
207 /* We preserve the initial values of the pointer variables. */
208 const char *inptr = *inbuf;
209 char *outbuf = data->outbuf;
210 char *outend = data->outbufend;
215 /* Remember the start value for this round. */
217 /* The outbuf buffer is empty. */
220 /* Save the state. */
221 last_converted = converted;
222 #ifdef SAVE_RESET_STATE
223 SAVE_RESET_STATE (1);
227 /* Run the conversion loop. */
228 status = FROM_LOOP ((const unsigned char **) inbuf,
229 (const unsigned char *) inbufend,
230 (unsigned char **) &outbuf,
231 (unsigned char *) outend,
232 data->statep, step->data, &converted);
234 /* Run the conversion loop. */
235 status = TO_LOOP ((const unsigned char **) inbuf,
236 (const unsigned char *) inbufend,
237 (unsigned char **) &outbuf,
238 (unsigned char *) outend,
239 data->statep, step->data, &converted);
241 /* If this is the last step leave the loop, there is nothgin
245 /* Store information about how many bytes are available. */
246 data->outbuf = outbuf;
250 /* Write out all output which was produced. */
253 const char *outerr = data->outbuf;
256 result = (*fct) (next_step, next_data, &outerr, outbuf,
259 if (result != GCONV_EMPTY_INPUT)
261 if (outerr != outbuf)
263 #ifdef RESET_INPUT_BUFFER
266 /* We have a problem with the in on of the functions
267 below. Undo the conversion upto the error point. */
270 /* Reload the pointers. */
274 /* Reset the state. */
275 converted = last_converted;
276 # ifdef SAVE_RESET_STATE
277 SAVE_RESET_STATE (0);
281 /* Run the conversion loop. */
282 nstatus = FROM_LOOP ((const unsigned char **) inbuf,
283 (const unsigned char *) inbufend,
284 (unsigned char **) &outbuf,
285 (unsigned char *) outerr,
286 data->statep, step->data,
289 /* Run the conversion loop. */
290 nstatus = TO_LOOP ((const unsigned char **) inbuf,
291 (const unsigned char *) inbufend,
292 (unsigned char **) &outbuf,
293 (unsigned char *) outerr,
294 data->statep, step->data,
297 /* We must run out of output buffer space in this
299 assert (outbuf == outerr);
300 assert (nstatus == GCONV_FULL_OUTPUT);
301 #endif /* reset input buffer */
304 /* Change the status. */
308 /* All the output is consumed, we can make another run
309 if everything was ok. */
310 if (status == GCONV_FULL_OUTPUT)
314 while (status == GCONV_OK);
316 /* Remember how many characters we converted. */
317 *written += converted;
326 #undef MIN_NEEDED_FROM
328 #undef MAX_NEEDED_FROM
330 #undef DEFINE_DIRECTION_OBJECTS
331 #undef FROM_DIRECTION
332 #undef EMIT_SHIFT_TO_INIT
336 #undef RESET_INPUT_BUFFER