Commit | Line | Data |
---|---|---|
234b45d4 | 1 | /* Character set conversion support for GDB. |
1bac305b | 2 | |
42a4f53d | 3 | Copyright (C) 2001-2019 Free Software Foundation, Inc. |
234b45d4 KB |
4 | |
5 | This file is part of GDB. | |
6 | ||
7 | This program is free software; you can redistribute it and/or modify | |
8 | it under the terms of the GNU General Public License as published by | |
a9762ec7 | 9 | the Free Software Foundation; either version 3 of the License, or |
234b45d4 KB |
10 | (at your option) any later version. |
11 | ||
12 | This program is distributed in the hope that it will be useful, | |
13 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
14 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
15 | GNU General Public License for more details. | |
16 | ||
17 | You should have received a copy of the GNU General Public License | |
a9762ec7 | 18 | along with this program. If not, see <http://www.gnu.org/licenses/>. */ |
234b45d4 KB |
19 | |
20 | #include "defs.h" | |
d55e5aa6 TT |
21 | |
22 | /* Standard C includes. */ | |
23 | #include <ctype.h> | |
24 | ||
25 | /* Local non-gdb includes. */ | |
26 | #include "arch-utils.h" | |
6c7a06a3 | 27 | #include "charset-list.h" |
d55e5aa6 | 28 | #include "charset.h" |
0747795c | 29 | #include "common/environ.h" |
0747795c | 30 | #include "common/gdb_vecs.h" |
d55e5aa6 TT |
31 | #include "common/gdb_wait.h" |
32 | #include "common/vec.h" | |
33 | #include "gdb_obstack.h" | |
34 | #include "gdbcmd.h" | |
234b45d4 | 35 | |
43484f03 DJ |
36 | #ifdef USE_WIN32API |
37 | #include <windows.h> | |
38 | #endif | |
234b45d4 KB |
39 | \f |
40 | /* How GDB's character set support works | |
41 | ||
6c7a06a3 | 42 | GDB has three global settings: |
234b45d4 KB |
43 | |
44 | - The `current host character set' is the character set GDB should | |
45 | use in talking to the user, and which (hopefully) the user's | |
6c7a06a3 TT |
46 | terminal knows how to display properly. Most users should not |
47 | change this. | |
234b45d4 KB |
48 | |
49 | - The `current target character set' is the character set the | |
50 | program being debugged uses. | |
51 | ||
6c7a06a3 TT |
52 | - The `current target wide character set' is the wide character set |
53 | the program being debugged uses, that is, the encoding used for | |
54 | wchar_t. | |
55 | ||
234b45d4 KB |
56 | There are commands to set each of these, and mechanisms for |
57 | choosing reasonable default values. GDB has a global list of | |
58 | character sets that it can use as its host or target character | |
59 | sets. | |
60 | ||
61 | The header file `charset.h' declares various functions that | |
62 | different pieces of GDB need to perform tasks like: | |
63 | ||
64 | - printing target strings and characters to the user's terminal | |
65 | (mostly target->host conversions), | |
66 | ||
67 | - building target-appropriate representations of strings and | |
68 | characters the user enters in expressions (mostly host->target | |
69 | conversions), | |
70 | ||
6c7a06a3 TT |
71 | and so on. |
72 | ||
73 | To avoid excessive code duplication and maintenance efforts, | |
74 | GDB simply requires a capable iconv function. Users on platforms | |
75 | without a suitable iconv can use the GNU iconv library. */ | |
234b45d4 KB |
76 | |
77 | \f | |
6c7a06a3 | 78 | #ifdef PHONY_ICONV |
234b45d4 | 79 | |
6c7a06a3 TT |
80 | /* Provide a phony iconv that does as little as possible. Also, |
81 | arrange for there to be a single available character set. */ | |
234b45d4 | 82 | |
6c7a06a3 | 83 | #undef GDB_DEFAULT_HOST_CHARSET |
f74f61cb SL |
84 | #ifdef USE_WIN32API |
85 | # define GDB_DEFAULT_HOST_CHARSET "CP1252" | |
86 | #else | |
87 | # define GDB_DEFAULT_HOST_CHARSET "ISO-8859-1" | |
88 | #endif | |
89 | #define GDB_DEFAULT_TARGET_CHARSET GDB_DEFAULT_HOST_CHARSET | |
90 | #define GDB_DEFAULT_TARGET_WIDE_CHARSET "UTF-32" | |
6c7a06a3 TT |
91 | #undef DEFAULT_CHARSET_NAMES |
92 | #define DEFAULT_CHARSET_NAMES GDB_DEFAULT_HOST_CHARSET , | |
93 | ||
94 | #undef iconv_t | |
95 | #define iconv_t int | |
96 | #undef iconv_open | |
62234ccc | 97 | #define iconv_open phony_iconv_open |
6c7a06a3 | 98 | #undef iconv |
62234ccc | 99 | #define iconv phony_iconv |
6c7a06a3 | 100 | #undef iconv_close |
62234ccc | 101 | #define iconv_close phony_iconv_close |
6c7a06a3 | 102 | |
0dd7fb99 TT |
103 | #undef ICONV_CONST |
104 | #define ICONV_CONST const | |
105 | ||
f74f61cb SL |
106 | /* We allow conversions from UTF-32, wchar_t, and the host charset. |
107 | We allow conversions to wchar_t and the host charset. | |
108 | Return 1 if we are converting from UTF-32BE, 2 if from UTF32-LE, | |
109 | 0 otherwise. This is used as a flag in calls to iconv. */ | |
110 | ||
a95babbf | 111 | static iconv_t |
62234ccc | 112 | phony_iconv_open (const char *to, const char *from) |
6c7a06a3 | 113 | { |
6c7a06a3 TT |
114 | if (strcmp (to, "wchar_t") && strcmp (to, GDB_DEFAULT_HOST_CHARSET)) |
115 | return -1; | |
234b45d4 | 116 | |
f74f61cb SL |
117 | if (!strcmp (from, "UTF-32BE") || !strcmp (from, "UTF-32")) |
118 | return 1; | |
119 | ||
120 | if (!strcmp (from, "UTF-32LE")) | |
121 | return 2; | |
122 | ||
123 | if (strcmp (from, "wchar_t") && strcmp (from, GDB_DEFAULT_HOST_CHARSET)) | |
124 | return -1; | |
125 | ||
126 | return 0; | |
6c7a06a3 | 127 | } |
234b45d4 | 128 | |
a95babbf | 129 | static int |
62234ccc | 130 | phony_iconv_close (iconv_t arg) |
6c7a06a3 TT |
131 | { |
132 | return 0; | |
133 | } | |
234b45d4 | 134 | |
a95babbf | 135 | static size_t |
62234ccc TT |
136 | phony_iconv (iconv_t utf_flag, const char **inbuf, size_t *inbytesleft, |
137 | char **outbuf, size_t *outbytesleft) | |
6c7a06a3 | 138 | { |
b8899f2b | 139 | if (utf_flag) |
6c7a06a3 | 140 | { |
f74f61cb SL |
141 | enum bfd_endian endian |
142 | = utf_flag == 1 ? BFD_ENDIAN_BIG : BFD_ENDIAN_LITTLE; | |
6c7a06a3 TT |
143 | while (*inbytesleft >= 4) |
144 | { | |
f74f61cb SL |
145 | unsigned long c |
146 | = extract_unsigned_integer ((const gdb_byte *)*inbuf, 4, endian); | |
6c7a06a3 TT |
147 | |
148 | if (c >= 256) | |
149 | { | |
150 | errno = EILSEQ; | |
151 | return -1; | |
152 | } | |
f74f61cb SL |
153 | if (*outbytesleft < 1) |
154 | { | |
155 | errno = E2BIG; | |
156 | return -1; | |
157 | } | |
6c7a06a3 TT |
158 | **outbuf = c & 0xff; |
159 | ++*outbuf; | |
160 | --*outbytesleft; | |
161 | ||
f74f61cb | 162 | *inbuf += 4; |
6c7a06a3 TT |
163 | *inbytesleft -= 4; |
164 | } | |
f74f61cb | 165 | if (*inbytesleft) |
6c7a06a3 | 166 | { |
f74f61cb | 167 | /* Partial sequence on input. */ |
6c7a06a3 TT |
168 | errno = EINVAL; |
169 | return -1; | |
170 | } | |
171 | } | |
172 | else | |
173 | { | |
174 | /* In all other cases we simply copy input bytes to the | |
175 | output. */ | |
176 | size_t amt = *inbytesleft; | |
c5504eaf | 177 | |
6c7a06a3 TT |
178 | if (amt > *outbytesleft) |
179 | amt = *outbytesleft; | |
180 | memcpy (*outbuf, *inbuf, amt); | |
181 | *inbuf += amt; | |
182 | *outbuf += amt; | |
183 | *inbytesleft -= amt; | |
184 | *outbytesleft -= amt; | |
f74f61cb SL |
185 | if (*inbytesleft) |
186 | { | |
187 | errno = E2BIG; | |
188 | return -1; | |
189 | } | |
6c7a06a3 | 190 | } |
234b45d4 | 191 | |
6c7a06a3 TT |
192 | /* The number of non-reversible conversions -- but they were all |
193 | reversible. */ | |
194 | return 0; | |
195 | } | |
234b45d4 | 196 | |
83030110 PA |
197 | #else /* PHONY_ICONV */ |
198 | ||
199 | /* On systems that don't have EILSEQ, GNU iconv's iconv.h defines it | |
200 | to ENOENT, while gnulib defines it to a different value. Always | |
201 | map ENOENT to gnulib's EILSEQ, leaving callers agnostic. */ | |
202 | ||
203 | static size_t | |
204 | gdb_iconv (iconv_t utf_flag, ICONV_CONST char **inbuf, size_t *inbytesleft, | |
205 | char **outbuf, size_t *outbytesleft) | |
206 | { | |
207 | size_t ret; | |
208 | ||
209 | ret = iconv (utf_flag, inbuf, inbytesleft, outbuf, outbytesleft); | |
210 | if (errno == ENOENT) | |
211 | errno = EILSEQ; | |
212 | return ret; | |
213 | } | |
214 | ||
215 | #undef iconv | |
216 | #define iconv gdb_iconv | |
234b45d4 | 217 | |
83030110 | 218 | #endif /* PHONY_ICONV */ |
234b45d4 KB |
219 | |
220 | \f | |
221 | /* The global lists of character sets and translations. */ | |
222 | ||
223 | ||
e33d66ec EZ |
224 | #ifndef GDB_DEFAULT_TARGET_CHARSET |
225 | #define GDB_DEFAULT_TARGET_CHARSET "ISO-8859-1" | |
226 | #endif | |
227 | ||
6c7a06a3 | 228 | #ifndef GDB_DEFAULT_TARGET_WIDE_CHARSET |
b8899f2b | 229 | #define GDB_DEFAULT_TARGET_WIDE_CHARSET "UTF-32" |
6c7a06a3 TT |
230 | #endif |
231 | ||
232 | static const char *auto_host_charset_name = GDB_DEFAULT_HOST_CHARSET; | |
233 | static const char *host_charset_name = "auto"; | |
920d2a44 AC |
234 | static void |
235 | show_host_charset_name (struct ui_file *file, int from_tty, | |
236 | struct cmd_list_element *c, | |
237 | const char *value) | |
238 | { | |
6c7a06a3 TT |
239 | if (!strcmp (value, "auto")) |
240 | fprintf_filtered (file, | |
241 | _("The host character set is \"auto; currently %s\".\n"), | |
242 | auto_host_charset_name); | |
243 | else | |
244 | fprintf_filtered (file, _("The host character set is \"%s\".\n"), value); | |
920d2a44 AC |
245 | } |
246 | ||
f870a310 | 247 | static const char *target_charset_name = "auto"; |
920d2a44 AC |
248 | static void |
249 | show_target_charset_name (struct ui_file *file, int from_tty, | |
250 | struct cmd_list_element *c, const char *value) | |
251 | { | |
f870a310 TT |
252 | if (!strcmp (value, "auto")) |
253 | fprintf_filtered (file, | |
254 | _("The target character set is \"auto; " | |
255 | "currently %s\".\n"), | |
256 | gdbarch_auto_charset (get_current_arch ())); | |
257 | else | |
258 | fprintf_filtered (file, _("The target character set is \"%s\".\n"), | |
259 | value); | |
920d2a44 AC |
260 | } |
261 | ||
f870a310 | 262 | static const char *target_wide_charset_name = "auto"; |
6c7a06a3 | 263 | static void |
aff410f1 MS |
264 | show_target_wide_charset_name (struct ui_file *file, |
265 | int from_tty, | |
266 | struct cmd_list_element *c, | |
267 | const char *value) | |
e33d66ec | 268 | { |
f870a310 TT |
269 | if (!strcmp (value, "auto")) |
270 | fprintf_filtered (file, | |
271 | _("The target wide character set is \"auto; " | |
272 | "currently %s\".\n"), | |
273 | gdbarch_auto_wide_charset (get_current_arch ())); | |
274 | else | |
275 | fprintf_filtered (file, _("The target wide character set is \"%s\".\n"), | |
276 | value); | |
6c7a06a3 | 277 | } |
e33d66ec | 278 | |
6c7a06a3 | 279 | static const char *default_charset_names[] = |
e33d66ec | 280 | { |
6c7a06a3 | 281 | DEFAULT_CHARSET_NAMES |
e33d66ec EZ |
282 | 0 |
283 | }; | |
234b45d4 | 284 | |
6c7a06a3 | 285 | static const char **charset_enum; |
234b45d4 | 286 | |
6c7a06a3 TT |
287 | \f |
288 | /* If the target wide character set has big- or little-endian | |
289 | variants, these are the corresponding names. */ | |
290 | static const char *target_wide_charset_be_name; | |
291 | static const char *target_wide_charset_le_name; | |
234b45d4 | 292 | |
f870a310 TT |
293 | /* The architecture for which the BE- and LE-names are valid. */ |
294 | static struct gdbarch *be_le_arch; | |
295 | ||
296 | /* A helper function which sets the target wide big- and little-endian | |
297 | character set names, if possible. */ | |
234b45d4 | 298 | |
6c7a06a3 | 299 | static void |
f870a310 | 300 | set_be_le_names (struct gdbarch *gdbarch) |
234b45d4 | 301 | { |
f870a310 TT |
302 | if (be_le_arch == gdbarch) |
303 | return; | |
304 | be_le_arch = gdbarch; | |
234b45d4 | 305 | |
f74f61cb SL |
306 | #ifdef PHONY_ICONV |
307 | /* Match the wide charset names recognized by phony_iconv_open. */ | |
308 | target_wide_charset_le_name = "UTF-32LE"; | |
309 | target_wide_charset_be_name = "UTF-32BE"; | |
310 | #else | |
15766370 TT |
311 | int i, len; |
312 | const char *target_wide; | |
313 | ||
6c7a06a3 TT |
314 | target_wide_charset_le_name = NULL; |
315 | target_wide_charset_be_name = NULL; | |
234b45d4 | 316 | |
f870a310 TT |
317 | target_wide = target_wide_charset_name; |
318 | if (!strcmp (target_wide, "auto")) | |
319 | target_wide = gdbarch_auto_wide_charset (gdbarch); | |
320 | ||
321 | len = strlen (target_wide); | |
6c7a06a3 TT |
322 | for (i = 0; charset_enum[i]; ++i) |
323 | { | |
f870a310 | 324 | if (strncmp (target_wide, charset_enum[i], len)) |
6c7a06a3 TT |
325 | continue; |
326 | if ((charset_enum[i][len] == 'B' | |
327 | || charset_enum[i][len] == 'L') | |
328 | && charset_enum[i][len + 1] == 'E' | |
329 | && charset_enum[i][len + 2] == '\0') | |
330 | { | |
331 | if (charset_enum[i][len] == 'B') | |
332 | target_wide_charset_be_name = charset_enum[i]; | |
333 | else | |
334 | target_wide_charset_le_name = charset_enum[i]; | |
335 | } | |
336 | } | |
f74f61cb | 337 | # endif /* PHONY_ICONV */ |
234b45d4 KB |
338 | } |
339 | ||
6c7a06a3 TT |
340 | /* 'Set charset', 'set host-charset', 'set target-charset', 'set |
341 | target-wide-charset', 'set charset' sfunc's. */ | |
234b45d4 KB |
342 | |
343 | static void | |
f870a310 | 344 | validate (struct gdbarch *gdbarch) |
234b45d4 | 345 | { |
6c7a06a3 TT |
346 | iconv_t desc; |
347 | const char *host_cset = host_charset (); | |
f870a310 TT |
348 | const char *target_cset = target_charset (gdbarch); |
349 | const char *target_wide_cset = target_wide_charset_name; | |
c5504eaf | 350 | |
f870a310 TT |
351 | if (!strcmp (target_wide_cset, "auto")) |
352 | target_wide_cset = gdbarch_auto_wide_charset (gdbarch); | |
234b45d4 | 353 | |
f870a310 | 354 | desc = iconv_open (target_wide_cset, host_cset); |
6c7a06a3 | 355 | if (desc == (iconv_t) -1) |
a73c6dcd | 356 | error (_("Cannot convert between character sets `%s' and `%s'"), |
f870a310 | 357 | target_wide_cset, host_cset); |
6c7a06a3 | 358 | iconv_close (desc); |
234b45d4 | 359 | |
f870a310 | 360 | desc = iconv_open (target_cset, host_cset); |
6c7a06a3 | 361 | if (desc == (iconv_t) -1) |
a73c6dcd | 362 | error (_("Cannot convert between character sets `%s' and `%s'"), |
f870a310 | 363 | target_cset, host_cset); |
6c7a06a3 | 364 | iconv_close (desc); |
234b45d4 | 365 | |
f870a310 TT |
366 | /* Clear the cache. */ |
367 | be_le_arch = NULL; | |
234b45d4 KB |
368 | } |
369 | ||
6c7a06a3 TT |
370 | /* This is the sfunc for the 'set charset' command. */ |
371 | static void | |
eb4c3f4a | 372 | set_charset_sfunc (const char *charset, int from_tty, |
aff410f1 | 373 | struct cmd_list_element *c) |
234b45d4 | 374 | { |
aff410f1 | 375 | /* CAREFUL: set the target charset here as well. */ |
6c7a06a3 | 376 | target_charset_name = host_charset_name; |
f870a310 | 377 | validate (get_current_arch ()); |
234b45d4 KB |
378 | } |
379 | ||
6c7a06a3 TT |
380 | /* 'set host-charset' command sfunc. We need a wrapper here because |
381 | the function needs to have a specific signature. */ | |
382 | static void | |
eb4c3f4a | 383 | set_host_charset_sfunc (const char *charset, int from_tty, |
6c7a06a3 | 384 | struct cmd_list_element *c) |
234b45d4 | 385 | { |
f870a310 | 386 | validate (get_current_arch ()); |
234b45d4 KB |
387 | } |
388 | ||
6c7a06a3 TT |
389 | /* Wrapper for the 'set target-charset' command. */ |
390 | static void | |
eb4c3f4a | 391 | set_target_charset_sfunc (const char *charset, int from_tty, |
6c7a06a3 | 392 | struct cmd_list_element *c) |
234b45d4 | 393 | { |
f870a310 | 394 | validate (get_current_arch ()); |
234b45d4 KB |
395 | } |
396 | ||
6c7a06a3 TT |
397 | /* Wrapper for the 'set target-wide-charset' command. */ |
398 | static void | |
eb4c3f4a | 399 | set_target_wide_charset_sfunc (const char *charset, int from_tty, |
6c7a06a3 | 400 | struct cmd_list_element *c) |
234b45d4 | 401 | { |
f870a310 | 402 | validate (get_current_arch ()); |
234b45d4 KB |
403 | } |
404 | ||
6c7a06a3 TT |
405 | /* sfunc for the 'show charset' command. */ |
406 | static void | |
aff410f1 MS |
407 | show_charset (struct ui_file *file, int from_tty, |
408 | struct cmd_list_element *c, | |
6c7a06a3 | 409 | const char *name) |
234b45d4 | 410 | { |
6c7a06a3 TT |
411 | show_host_charset_name (file, from_tty, c, host_charset_name); |
412 | show_target_charset_name (file, from_tty, c, target_charset_name); | |
aff410f1 MS |
413 | show_target_wide_charset_name (file, from_tty, c, |
414 | target_wide_charset_name); | |
234b45d4 KB |
415 | } |
416 | ||
234b45d4 | 417 | \f |
6c7a06a3 | 418 | /* Accessor functions. */ |
234b45d4 | 419 | |
6c7a06a3 TT |
420 | const char * |
421 | host_charset (void) | |
234b45d4 | 422 | { |
6c7a06a3 TT |
423 | if (!strcmp (host_charset_name, "auto")) |
424 | return auto_host_charset_name; | |
425 | return host_charset_name; | |
234b45d4 KB |
426 | } |
427 | ||
6c7a06a3 | 428 | const char * |
f870a310 | 429 | target_charset (struct gdbarch *gdbarch) |
234b45d4 | 430 | { |
f870a310 TT |
431 | if (!strcmp (target_charset_name, "auto")) |
432 | return gdbarch_auto_charset (gdbarch); | |
6c7a06a3 | 433 | return target_charset_name; |
234b45d4 | 434 | } |
234b45d4 | 435 | |
6c7a06a3 | 436 | const char * |
f870a310 | 437 | target_wide_charset (struct gdbarch *gdbarch) |
234b45d4 | 438 | { |
f870a310 TT |
439 | enum bfd_endian byte_order = gdbarch_byte_order (gdbarch); |
440 | ||
441 | set_be_le_names (gdbarch); | |
e17a4113 | 442 | if (byte_order == BFD_ENDIAN_BIG) |
234b45d4 | 443 | { |
6c7a06a3 TT |
444 | if (target_wide_charset_be_name) |
445 | return target_wide_charset_be_name; | |
234b45d4 | 446 | } |
6c7a06a3 | 447 | else |
234b45d4 | 448 | { |
6c7a06a3 TT |
449 | if (target_wide_charset_le_name) |
450 | return target_wide_charset_le_name; | |
234b45d4 KB |
451 | } |
452 | ||
f870a310 TT |
453 | if (!strcmp (target_wide_charset_name, "auto")) |
454 | return gdbarch_auto_wide_charset (gdbarch); | |
455 | ||
6c7a06a3 | 456 | return target_wide_charset_name; |
234b45d4 KB |
457 | } |
458 | ||
234b45d4 | 459 | \f |
6c7a06a3 TT |
460 | /* Host character set management. For the time being, we assume that |
461 | the host character set is some superset of ASCII. */ | |
234b45d4 | 462 | |
6c7a06a3 TT |
463 | char |
464 | host_letter_to_control_character (char c) | |
234b45d4 | 465 | { |
6c7a06a3 TT |
466 | if (c == '?') |
467 | return 0177; | |
468 | return c & 0237; | |
234b45d4 KB |
469 | } |
470 | ||
6c7a06a3 TT |
471 | /* Convert a host character, C, to its hex value. C must already have |
472 | been validated using isxdigit. */ | |
234b45d4 | 473 | |
6c7a06a3 TT |
474 | int |
475 | host_hex_value (char c) | |
234b45d4 | 476 | { |
6c7a06a3 TT |
477 | if (isdigit (c)) |
478 | return c - '0'; | |
479 | if (c >= 'a' && c <= 'f') | |
480 | return 10 + c - 'a'; | |
481 | gdb_assert (c >= 'A' && c <= 'F'); | |
482 | return 10 + c - 'A'; | |
234b45d4 KB |
483 | } |
484 | ||
234b45d4 | 485 | \f |
6c7a06a3 | 486 | /* Public character management functions. */ |
234b45d4 | 487 | |
80a3b8c5 | 488 | class iconv_wrapper |
234b45d4 | 489 | { |
80a3b8c5 TT |
490 | public: |
491 | ||
492 | iconv_wrapper (const char *to, const char *from) | |
493 | { | |
494 | m_desc = iconv_open (to, from); | |
495 | if (m_desc == (iconv_t) -1) | |
496 | perror_with_name (_("Converting character sets")); | |
497 | } | |
498 | ||
499 | ~iconv_wrapper () | |
500 | { | |
501 | iconv_close (m_desc); | |
502 | } | |
503 | ||
504 | size_t convert (ICONV_CONST char **inp, size_t *inleft, char **outp, | |
505 | size_t *outleft) | |
506 | { | |
507 | return iconv (m_desc, inp, inleft, outp, outleft); | |
508 | } | |
509 | ||
510 | private: | |
511 | ||
512 | iconv_t m_desc; | |
513 | }; | |
234b45d4 | 514 | |
6c7a06a3 TT |
515 | void |
516 | convert_between_encodings (const char *from, const char *to, | |
517 | const gdb_byte *bytes, unsigned int num_bytes, | |
518 | int width, struct obstack *output, | |
519 | enum transliterations translit) | |
520 | { | |
6c7a06a3 | 521 | size_t inleft; |
39086a0e | 522 | ICONV_CONST char *inp; |
6c7a06a3 TT |
523 | unsigned int space_request; |
524 | ||
525 | /* Often, the host and target charsets will be the same. */ | |
526 | if (!strcmp (from, to)) | |
527 | { | |
528 | obstack_grow (output, bytes, num_bytes); | |
529 | return; | |
530 | } | |
234b45d4 | 531 | |
80a3b8c5 | 532 | iconv_wrapper desc (to, from); |
234b45d4 | 533 | |
6c7a06a3 | 534 | inleft = num_bytes; |
39086a0e | 535 | inp = (ICONV_CONST char *) bytes; |
234b45d4 | 536 | |
6c7a06a3 | 537 | space_request = num_bytes; |
234b45d4 | 538 | |
6c7a06a3 | 539 | while (inleft > 0) |
234b45d4 | 540 | { |
6c7a06a3 TT |
541 | char *outp; |
542 | size_t outleft, r; | |
543 | int old_size; | |
544 | ||
545 | old_size = obstack_object_size (output); | |
546 | obstack_blank (output, space_request); | |
547 | ||
241fd515 | 548 | outp = (char *) obstack_base (output) + old_size; |
6c7a06a3 TT |
549 | outleft = space_request; |
550 | ||
80a3b8c5 | 551 | r = desc.convert (&inp, &inleft, &outp, &outleft); |
6c7a06a3 TT |
552 | |
553 | /* Now make sure that the object on the obstack only includes | |
554 | bytes we have converted. */ | |
89eb3c54 | 555 | obstack_blank_fast (output, -(ssize_t) outleft); |
6c7a06a3 TT |
556 | |
557 | if (r == (size_t) -1) | |
558 | { | |
559 | switch (errno) | |
560 | { | |
561 | case EILSEQ: | |
562 | { | |
563 | int i; | |
564 | ||
565 | /* Invalid input sequence. */ | |
566 | if (translit == translit_none) | |
3e43a32a MS |
567 | error (_("Could not convert character " |
568 | "to `%s' character set"), to); | |
6c7a06a3 TT |
569 | |
570 | /* We emit escape sequence for the bytes, skip them, | |
571 | and try again. */ | |
572 | for (i = 0; i < width; ++i) | |
573 | { | |
574 | char octal[5]; | |
575 | ||
08850b56 | 576 | xsnprintf (octal, sizeof (octal), "\\%.3o", *inp & 0xff); |
6c7a06a3 TT |
577 | obstack_grow_str (output, octal); |
578 | ||
579 | ++inp; | |
580 | --inleft; | |
581 | } | |
582 | } | |
583 | break; | |
584 | ||
585 | case E2BIG: | |
586 | /* We ran out of space in the output buffer. Make it | |
587 | bigger next time around. */ | |
588 | space_request *= 2; | |
589 | break; | |
590 | ||
591 | case EINVAL: | |
592 | /* Incomplete input sequence. FIXME: ought to report this | |
593 | to the caller somehow. */ | |
594 | inleft = 0; | |
595 | break; | |
596 | ||
597 | default: | |
9b20d036 MS |
598 | perror_with_name (_("Internal error while " |
599 | "converting character sets")); | |
6c7a06a3 TT |
600 | } |
601 | } | |
234b45d4 | 602 | } |
234b45d4 KB |
603 | } |
604 | ||
e33d66ec | 605 | \f |
e33d66ec | 606 | |
6c7a06a3 | 607 | /* Create a new iterator. */ |
cda6c55b TT |
608 | wchar_iterator::wchar_iterator (const gdb_byte *input, size_t bytes, |
609 | const char *charset, size_t width) | |
610 | : m_input (input), | |
611 | m_bytes (bytes), | |
612 | m_width (width), | |
613 | m_out (1) | |
234b45d4 | 614 | { |
cda6c55b TT |
615 | m_desc = iconv_open (INTERMEDIATE_ENCODING, charset); |
616 | if (m_desc == (iconv_t) -1) | |
9b20d036 | 617 | perror_with_name (_("Converting character sets")); |
234b45d4 KB |
618 | } |
619 | ||
cda6c55b | 620 | wchar_iterator::~wchar_iterator () |
e33d66ec | 621 | { |
cda6c55b TT |
622 | if (m_desc != (iconv_t) -1) |
623 | iconv_close (m_desc); | |
e33d66ec | 624 | } |
234b45d4 | 625 | |
6c7a06a3 | 626 | int |
cda6c55b TT |
627 | wchar_iterator::iterate (enum wchar_iterate_result *out_result, |
628 | gdb_wchar_t **out_chars, | |
629 | const gdb_byte **ptr, | |
630 | size_t *len) | |
6c7a06a3 TT |
631 | { |
632 | size_t out_request; | |
633 | ||
634 | /* Try to convert some characters. At first we try to convert just | |
635 | a single character. The reason for this is that iconv does not | |
636 | necessarily update its outgoing arguments when it encounters an | |
637 | invalid input sequence -- but we want to reliably report this to | |
638 | our caller so it can emit an escape sequence. */ | |
639 | out_request = 1; | |
cda6c55b | 640 | while (m_bytes > 0) |
e33d66ec | 641 | { |
cda6c55b TT |
642 | ICONV_CONST char *inptr = (ICONV_CONST char *) m_input; |
643 | char *outptr = (char *) m_out.data (); | |
644 | const gdb_byte *orig_inptr = m_input; | |
645 | size_t orig_in = m_bytes; | |
6c7a06a3 TT |
646 | size_t out_avail = out_request * sizeof (gdb_wchar_t); |
647 | size_t num; | |
cda6c55b | 648 | size_t r = iconv (m_desc, &inptr, &m_bytes, &outptr, &out_avail); |
39086a0e | 649 | |
cda6c55b | 650 | m_input = (gdb_byte *) inptr; |
c5504eaf | 651 | |
6c7a06a3 TT |
652 | if (r == (size_t) -1) |
653 | { | |
654 | switch (errno) | |
655 | { | |
656 | case EILSEQ: | |
aff410f1 MS |
657 | /* Invalid input sequence. We still might have |
658 | converted a character; if so, return it. */ | |
1558ab4c JK |
659 | if (out_avail < out_request * sizeof (gdb_wchar_t)) |
660 | break; | |
661 | ||
aff410f1 MS |
662 | /* Otherwise skip the first invalid character, and let |
663 | the caller know about it. */ | |
6c7a06a3 | 664 | *out_result = wchar_iterate_invalid; |
cda6c55b TT |
665 | *ptr = m_input; |
666 | *len = m_width; | |
667 | m_input += m_width; | |
668 | m_bytes -= m_width; | |
6c7a06a3 TT |
669 | return 0; |
670 | ||
671 | case E2BIG: | |
672 | /* We ran out of space. We still might have converted a | |
673 | character; if so, return it. Otherwise, grow the | |
674 | buffer and try again. */ | |
675 | if (out_avail < out_request * sizeof (gdb_wchar_t)) | |
676 | break; | |
677 | ||
678 | ++out_request; | |
cda6c55b | 679 | if (out_request > m_out.size ()) |
d5722aa2 | 680 | m_out.resize (out_request); |
6c7a06a3 TT |
681 | continue; |
682 | ||
683 | case EINVAL: | |
684 | /* Incomplete input sequence. Let the caller know, and | |
685 | arrange for future calls to see EOF. */ | |
686 | *out_result = wchar_iterate_incomplete; | |
cda6c55b TT |
687 | *ptr = m_input; |
688 | *len = m_bytes; | |
689 | m_bytes = 0; | |
6c7a06a3 TT |
690 | return 0; |
691 | ||
692 | default: | |
9b20d036 MS |
693 | perror_with_name (_("Internal error while " |
694 | "converting character sets")); | |
6c7a06a3 TT |
695 | } |
696 | } | |
697 | ||
698 | /* We converted something. */ | |
699 | num = out_request - out_avail / sizeof (gdb_wchar_t); | |
700 | *out_result = wchar_iterate_ok; | |
cda6c55b | 701 | *out_chars = m_out.data (); |
6c7a06a3 | 702 | *ptr = orig_inptr; |
cda6c55b | 703 | *len = orig_in - m_bytes; |
6c7a06a3 | 704 | return num; |
e33d66ec | 705 | } |
6c7a06a3 TT |
706 | |
707 | /* Really done. */ | |
708 | *out_result = wchar_iterate_eof; | |
709 | return -1; | |
234b45d4 KB |
710 | } |
711 | ||
ccb2231c SM |
712 | struct charset_vector |
713 | { | |
714 | ~charset_vector () | |
715 | { | |
716 | clear (); | |
717 | } | |
718 | ||
719 | void clear () | |
720 | { | |
721 | for (char *c : charsets) | |
722 | xfree (c); | |
234b45d4 | 723 | |
ccb2231c SM |
724 | charsets.clear (); |
725 | } | |
726 | ||
727 | std::vector<char *> charsets; | |
728 | }; | |
729 | ||
730 | static charset_vector charsets; | |
234b45d4 | 731 | |
6c7a06a3 | 732 | #ifdef PHONY_ICONV |
234b45d4 | 733 | |
6c7a06a3 TT |
734 | static void |
735 | find_charset_names (void) | |
234b45d4 | 736 | { |
ccb2231c SM |
737 | charsets.charsets.push_back (xstrdup (GDB_DEFAULT_HOST_CHARSET)); |
738 | charsets.charsets.push_back (NULL); | |
234b45d4 KB |
739 | } |
740 | ||
6c7a06a3 | 741 | #else /* PHONY_ICONV */ |
fc3b640d TT |
742 | |
743 | /* Sometimes, libiconv redefines iconvlist as libiconvlist -- but | |
744 | provides different symbols in the static and dynamic libraries. | |
745 | So, configure may see libiconvlist but not iconvlist. But, calling | |
746 | iconvlist is the right thing to do and will work. Hence we do a | |
747 | check here but unconditionally call iconvlist below. */ | |
748 | #if defined (HAVE_ICONVLIST) || defined (HAVE_LIBICONVLIST) | |
234b45d4 | 749 | |
6c7a06a3 TT |
750 | /* A helper function that adds some character sets to the vector of |
751 | all character sets. This is a callback function for iconvlist. */ | |
752 | ||
753 | static int | |
754 | add_one (unsigned int count, const char *const *names, void *data) | |
234b45d4 | 755 | { |
6c7a06a3 | 756 | unsigned int i; |
234b45d4 | 757 | |
6c7a06a3 | 758 | for (i = 0; i < count; ++i) |
ccb2231c | 759 | charsets.charsets.push_back (xstrdup (names[i])); |
234b45d4 | 760 | |
6c7a06a3 | 761 | return 0; |
234b45d4 KB |
762 | } |
763 | ||
6c7a06a3 TT |
764 | static void |
765 | find_charset_names (void) | |
234b45d4 | 766 | { |
6c7a06a3 | 767 | iconvlist (add_one, NULL); |
ccb2231c SM |
768 | |
769 | charsets.charsets.push_back (NULL); | |
234b45d4 KB |
770 | } |
771 | ||
6c7a06a3 | 772 | #else |
234b45d4 | 773 | |
40b5c9fb DE |
774 | /* Return non-zero if LINE (output from iconv) should be ignored. |
775 | Older iconv programs (e.g. 2.2.2) include the human readable | |
776 | introduction even when stdout is not a tty. Newer versions omit | |
777 | the intro if stdout is not a tty. */ | |
778 | ||
779 | static int | |
780 | ignore_line_p (const char *line) | |
781 | { | |
782 | /* This table is used to filter the output. If this text appears | |
783 | anywhere in the line, it is ignored (strstr is used). */ | |
784 | static const char * const ignore_lines[] = | |
785 | { | |
786 | "The following", | |
787 | "not necessarily", | |
788 | "the FROM and TO", | |
789 | "listed with several", | |
790 | NULL | |
791 | }; | |
792 | int i; | |
793 | ||
794 | for (i = 0; ignore_lines[i] != NULL; ++i) | |
795 | { | |
796 | if (strstr (line, ignore_lines[i]) != NULL) | |
797 | return 1; | |
798 | } | |
799 | ||
800 | return 0; | |
801 | } | |
802 | ||
6c7a06a3 TT |
803 | static void |
804 | find_charset_names (void) | |
234b45d4 | 805 | { |
732f6a93 | 806 | struct pex_obj *child; |
a121b7c1 | 807 | const char *args[3]; |
732f6a93 TT |
808 | int err, status; |
809 | int fail = 1; | |
478aac75 | 810 | int flags; |
9a6c7d9c | 811 | gdb_environ iconv_env = gdb_environ::from_host_environ (); |
478aac75 | 812 | char *iconv_program; |
40b5c9fb | 813 | |
aff410f1 MS |
814 | /* Older iconvs, e.g. 2.2.2, don't omit the intro text if stdout is |
815 | not a tty. We need to recognize it and ignore it. This text is | |
816 | subject to translation, so force LANGUAGE=C. */ | |
9a6c7d9c SDJ |
817 | iconv_env.set ("LANGUAGE", "C"); |
818 | iconv_env.set ("LC_ALL", "C"); | |
732f6a93 | 819 | |
40618926 | 820 | child = pex_init (PEX_USE_PIPES, "iconv", NULL); |
732f6a93 | 821 | |
478aac75 DE |
822 | #ifdef ICONV_BIN |
823 | { | |
824 | char *iconv_dir = relocate_gdb_directory (ICONV_BIN, | |
825 | ICONV_BIN_RELOCATABLE); | |
826 | iconv_program = concat (iconv_dir, SLASH_STRING, "iconv", NULL); | |
827 | xfree (iconv_dir); | |
828 | } | |
829 | #else | |
830 | iconv_program = xstrdup ("iconv"); | |
831 | #endif | |
832 | args[0] = iconv_program; | |
732f6a93 TT |
833 | args[1] = "-l"; |
834 | args[2] = NULL; | |
478aac75 DE |
835 | flags = PEX_STDERR_TO_STDOUT; |
836 | #ifndef ICONV_BIN | |
837 | flags |= PEX_SEARCH; | |
838 | #endif | |
732f6a93 | 839 | /* Note that we simply ignore errors here. */ |
478aac75 | 840 | if (!pex_run_in_environment (child, flags, |
a121b7c1 | 841 | args[0], const_cast<char **> (args), |
9a6c7d9c | 842 | iconv_env.envp (), |
40b5c9fb | 843 | NULL, NULL, &err)) |
732f6a93 TT |
844 | { |
845 | FILE *in = pex_read_output (child, 0); | |
846 | ||
847 | /* POSIX says that iconv -l uses an unspecified format. We | |
848 | parse the glibc and libiconv formats; feel free to add others | |
849 | as needed. */ | |
40b5c9fb | 850 | |
1d6b2d2b | 851 | while (in != NULL && !feof (in)) |
732f6a93 TT |
852 | { |
853 | /* The size of buf is chosen arbitrarily. */ | |
854 | char buf[1024]; | |
855 | char *start, *r; | |
8ea13695 | 856 | int len; |
732f6a93 TT |
857 | |
858 | r = fgets (buf, sizeof (buf), in); | |
859 | if (!r) | |
860 | break; | |
861 | len = strlen (r); | |
862 | if (len <= 3) | |
863 | continue; | |
40b5c9fb DE |
864 | if (ignore_line_p (r)) |
865 | continue; | |
866 | ||
732f6a93 TT |
867 | /* Strip off the newline. */ |
868 | --len; | |
869 | /* Strip off one or two '/'s. glibc will print lines like | |
870 | "8859_7//", but also "10646-1:1993/UCS4/". */ | |
871 | if (buf[len - 1] == '/') | |
872 | --len; | |
873 | if (buf[len - 1] == '/') | |
874 | --len; | |
875 | buf[len] = '\0'; | |
876 | ||
877 | /* libiconv will print multiple entries per line, separated | |
aff410f1 MS |
878 | by spaces. Older iconvs will print multiple entries per |
879 | line, indented by two spaces, and separated by ", " | |
40b5c9fb | 880 | (i.e. the human readable form). */ |
732f6a93 TT |
881 | start = buf; |
882 | while (1) | |
883 | { | |
884 | int keep_going; | |
885 | char *p; | |
886 | ||
40b5c9fb DE |
887 | /* Skip leading blanks. */ |
888 | for (p = start; *p && *p == ' '; ++p) | |
889 | ; | |
890 | start = p; | |
891 | /* Find the next space, comma, or end-of-line. */ | |
892 | for ( ; *p && *p != ' ' && *p != ','; ++p) | |
732f6a93 TT |
893 | ; |
894 | /* Ignore an empty result. */ | |
895 | if (p == start) | |
896 | break; | |
897 | keep_going = *p; | |
898 | *p = '\0'; | |
ccb2231c | 899 | charsets.charsets.push_back (xstrdup (start)); |
732f6a93 TT |
900 | if (!keep_going) |
901 | break; | |
902 | /* Skip any extra spaces. */ | |
903 | for (start = p + 1; *start && *start == ' '; ++start) | |
904 | ; | |
905 | } | |
906 | } | |
234b45d4 | 907 | |
732f6a93 TT |
908 | if (pex_get_status (child, 1, &status) |
909 | && WIFEXITED (status) && !WEXITSTATUS (status)) | |
910 | fail = 0; | |
234b45d4 | 911 | |
6c7a06a3 | 912 | } |
234b45d4 | 913 | |
478aac75 | 914 | xfree (iconv_program); |
732f6a93 | 915 | pex_free (child); |
234b45d4 | 916 | |
732f6a93 TT |
917 | if (fail) |
918 | { | |
919 | /* Some error occurred, so drop the vector. */ | |
ccb2231c | 920 | charsets.clear (); |
732f6a93 TT |
921 | } |
922 | else | |
ccb2231c | 923 | charsets.charsets.push_back (NULL); |
6c7a06a3 | 924 | } |
234b45d4 | 925 | |
fc3b640d | 926 | #endif /* HAVE_ICONVLIST || HAVE_LIBICONVLIST */ |
6c7a06a3 | 927 | #endif /* PHONY_ICONV */ |
234b45d4 | 928 | |
f870a310 TT |
929 | /* The "auto" target charset used by default_auto_charset. */ |
930 | static const char *auto_target_charset_name = GDB_DEFAULT_TARGET_CHARSET; | |
931 | ||
932 | const char * | |
933 | default_auto_charset (void) | |
934 | { | |
935 | return auto_target_charset_name; | |
936 | } | |
937 | ||
938 | const char * | |
939 | default_auto_wide_charset (void) | |
940 | { | |
941 | return GDB_DEFAULT_TARGET_WIDE_CHARSET; | |
942 | } | |
943 | ||
bcb28afc PM |
944 | |
945 | #ifdef USE_INTERMEDIATE_ENCODING_FUNCTION | |
946 | /* Macro used for UTF or UCS endianness suffix. */ | |
947 | #if WORDS_BIGENDIAN | |
948 | #define ENDIAN_SUFFIX "BE" | |
949 | #else | |
950 | #define ENDIAN_SUFFIX "LE" | |
951 | #endif | |
952 | ||
953 | /* The code below serves to generate a compile time error if | |
954 | gdb_wchar_t type is not of size 2 nor 4, despite the fact that | |
955 | macro __STDC_ISO_10646__ is defined. | |
956 | This is better than a gdb_assert call, because GDB cannot handle | |
957 | strings correctly if this size is different. */ | |
958 | ||
959 | extern char your_gdb_wchar_t_is_bogus[(sizeof (gdb_wchar_t) == 2 | |
960 | || sizeof (gdb_wchar_t) == 4) | |
961 | ? 1 : -1]; | |
962 | ||
ee34b3f9 | 963 | /* intermediate_encoding returns the charset used internally by |
bcb28afc PM |
964 | GDB to convert between target and host encodings. As the test above |
965 | compiled, sizeof (gdb_wchar_t) is either 2 or 4 bytes. | |
966 | UTF-16/32 is tested first, UCS-2/4 is tested as a second option, | |
967 | otherwise an error is generated. */ | |
968 | ||
969 | const char * | |
970 | intermediate_encoding (void) | |
971 | { | |
972 | iconv_t desc; | |
973 | static const char *stored_result = NULL; | |
974 | char *result; | |
bcb28afc PM |
975 | |
976 | if (stored_result) | |
977 | return stored_result; | |
978 | result = xstrprintf ("UTF-%d%s", (int) (sizeof (gdb_wchar_t) * 8), | |
979 | ENDIAN_SUFFIX); | |
980 | /* Check that the name is supported by iconv_open. */ | |
981 | desc = iconv_open (result, host_charset ()); | |
982 | if (desc != (iconv_t) -1) | |
983 | { | |
984 | iconv_close (desc); | |
985 | stored_result = result; | |
986 | return result; | |
987 | } | |
988 | /* Not valid, free the allocated memory. */ | |
989 | xfree (result); | |
990 | /* Second try, with UCS-2 type. */ | |
991 | result = xstrprintf ("UCS-%d%s", (int) sizeof (gdb_wchar_t), | |
992 | ENDIAN_SUFFIX); | |
993 | /* Check that the name is supported by iconv_open. */ | |
994 | desc = iconv_open (result, host_charset ()); | |
995 | if (desc != (iconv_t) -1) | |
996 | { | |
997 | iconv_close (desc); | |
998 | stored_result = result; | |
999 | return result; | |
1000 | } | |
1001 | /* Not valid, free the allocated memory. */ | |
1002 | xfree (result); | |
1003 | /* No valid charset found, generate error here. */ | |
1004 | error (_("Unable to find a vaild charset for string conversions")); | |
1005 | } | |
1006 | ||
1007 | #endif /* USE_INTERMEDIATE_ENCODING_FUNCTION */ | |
1008 | ||
234b45d4 KB |
1009 | void |
1010 | _initialize_charset (void) | |
1011 | { | |
f870a310 | 1012 | /* The first element is always "auto". */ |
ccb2231c | 1013 | charsets.charsets.push_back (xstrdup ("auto")); |
6c7a06a3 TT |
1014 | find_charset_names (); |
1015 | ||
ccb2231c SM |
1016 | if (charsets.charsets.size () > 1) |
1017 | charset_enum = (const char **) charsets.charsets.data (); | |
6c7a06a3 TT |
1018 | else |
1019 | charset_enum = default_charset_names; | |
1020 | ||
1021 | #ifndef PHONY_ICONV | |
1022 | #ifdef HAVE_LANGINFO_CODESET | |
f870a310 TT |
1023 | /* The result of nl_langinfo may be overwritten later. This may |
1024 | leak a little memory, if the user later changes the host charset, | |
1025 | but that doesn't matter much. */ | |
1026 | auto_host_charset_name = xstrdup (nl_langinfo (CODESET)); | |
aff410f1 MS |
1027 | /* Solaris will return `646' here -- but the Solaris iconv then does |
1028 | not accept this. Darwin (and maybe FreeBSD) may return "" here, | |
06be6983 TG |
1029 | which GNU libiconv doesn't like (infinite loop). */ |
1030 | if (!strcmp (auto_host_charset_name, "646") || !*auto_host_charset_name) | |
58720494 | 1031 | auto_host_charset_name = "ASCII"; |
f870a310 TT |
1032 | auto_target_charset_name = auto_host_charset_name; |
1033 | #elif defined (USE_WIN32API) | |
1034 | { | |
3e43a32a MS |
1035 | /* "CP" + x<=5 digits + paranoia. */ |
1036 | static char w32_host_default_charset[16]; | |
f870a310 TT |
1037 | |
1038 | snprintf (w32_host_default_charset, sizeof w32_host_default_charset, | |
1039 | "CP%d", GetACP()); | |
1040 | auto_host_charset_name = w32_host_default_charset; | |
1041 | auto_target_charset_name = auto_host_charset_name; | |
1042 | } | |
6c7a06a3 TT |
1043 | #endif |
1044 | #endif | |
e33d66ec | 1045 | |
7ab04401 | 1046 | add_setshow_enum_cmd ("charset", class_support, |
f870a310 | 1047 | charset_enum, &host_charset_name, _("\ |
7ab04401 AC |
1048 | Set the host and target character sets."), _("\ |
1049 | Show the host and target character sets."), _("\ | |
3d263c1d BI |
1050 | The `host character set' is the one used by the system GDB is running on.\n\ |
1051 | The `target character set' is the one used by the program being debugged.\n\ | |
1052 | You may only use supersets of ASCII for your host character set; GDB does\n\ | |
1053 | not support any others.\n\ | |
1054 | To see a list of the character sets GDB supports, type `set charset <TAB>'."), | |
7ab04401 AC |
1055 | /* Note that the sfunc below needs to set |
1056 | target_charset_name, because the 'set | |
1057 | charset' command sets two variables. */ | |
1058 | set_charset_sfunc, | |
1059 | show_charset, | |
1060 | &setlist, &showlist); | |
1061 | ||
1062 | add_setshow_enum_cmd ("host-charset", class_support, | |
6c7a06a3 | 1063 | charset_enum, &host_charset_name, _("\ |
7ab04401 AC |
1064 | Set the host character set."), _("\ |
1065 | Show the host character set."), _("\ | |
3d263c1d BI |
1066 | The `host character set' is the one used by the system GDB is running on.\n\ |
1067 | You may only use supersets of ASCII for your host character set; GDB does\n\ | |
ac74f770 MS |
1068 | not support any others.\n\ |
1069 | To see a list of the character sets GDB supports, type `set host-charset <TAB>'."), | |
7ab04401 | 1070 | set_host_charset_sfunc, |
920d2a44 | 1071 | show_host_charset_name, |
7ab04401 AC |
1072 | &setlist, &showlist); |
1073 | ||
1074 | add_setshow_enum_cmd ("target-charset", class_support, | |
f870a310 | 1075 | charset_enum, &target_charset_name, _("\ |
7ab04401 AC |
1076 | Set the target character set."), _("\ |
1077 | Show the target character set."), _("\ | |
3d263c1d BI |
1078 | The `target character set' is the one used by the program being debugged.\n\ |
1079 | GDB translates characters and strings between the host and target\n\ | |
b670013c | 1080 | character sets as needed.\n\ |
ac74f770 | 1081 | To see a list of the character sets GDB supports, type `set target-charset'<TAB>"), |
7ab04401 | 1082 | set_target_charset_sfunc, |
920d2a44 | 1083 | show_target_charset_name, |
7ab04401 | 1084 | &setlist, &showlist); |
6c7a06a3 TT |
1085 | |
1086 | add_setshow_enum_cmd ("target-wide-charset", class_support, | |
f870a310 | 1087 | charset_enum, &target_wide_charset_name, |
6c7a06a3 TT |
1088 | _("\ |
1089 | Set the target wide character set."), _("\ | |
1090 | Show the target wide character set."), _("\ | |
3e43a32a MS |
1091 | The `target wide character set' is the one used by the program being debugged.\ |
1092 | \nIn particular it is the encoding used by `wchar_t'.\n\ | |
6c7a06a3 TT |
1093 | GDB translates characters and strings between the host and target\n\ |
1094 | character sets as needed.\n\ | |
1095 | To see a list of the character sets GDB supports, type\n\ | |
1096 | `set target-wide-charset'<TAB>"), | |
1097 | set_target_wide_charset_sfunc, | |
1098 | show_target_wide_charset_name, | |
1099 | &setlist, &showlist); | |
234b45d4 | 1100 | } |