Commit | Line | Data |
---|---|---|
234b45d4 | 1 | /* Character set conversion support for GDB. |
1bac305b | 2 | |
0b302171 | 3 | Copyright (C) 2001, 2003, 2007-2012 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" | |
21 | #include "charset.h" | |
22 | #include "gdbcmd.h" | |
23 | #include "gdb_assert.h" | |
6c7a06a3 | 24 | #include "gdb_obstack.h" |
732f6a93 | 25 | #include "gdb_wait.h" |
6c7a06a3 TT |
26 | #include "charset-list.h" |
27 | #include "vec.h" | |
40b5c9fb | 28 | #include "environ.h" |
f870a310 | 29 | #include "arch-utils.h" |
234b45d4 KB |
30 | |
31 | #include <stddef.h> | |
4ef3f3be | 32 | #include "gdb_string.h" |
234b45d4 KB |
33 | #include <ctype.h> |
34 | ||
43484f03 DJ |
35 | #ifdef USE_WIN32API |
36 | #include <windows.h> | |
37 | #endif | |
234b45d4 KB |
38 | \f |
39 | /* How GDB's character set support works | |
40 | ||
6c7a06a3 | 41 | GDB has three global settings: |
234b45d4 KB |
42 | |
43 | - The `current host character set' is the character set GDB should | |
44 | use in talking to the user, and which (hopefully) the user's | |
6c7a06a3 TT |
45 | terminal knows how to display properly. Most users should not |
46 | change this. | |
234b45d4 KB |
47 | |
48 | - The `current target character set' is the character set the | |
49 | program being debugged uses. | |
50 | ||
6c7a06a3 TT |
51 | - The `current target wide character set' is the wide character set |
52 | the program being debugged uses, that is, the encoding used for | |
53 | wchar_t. | |
54 | ||
234b45d4 KB |
55 | There are commands to set each of these, and mechanisms for |
56 | choosing reasonable default values. GDB has a global list of | |
57 | character sets that it can use as its host or target character | |
58 | sets. | |
59 | ||
60 | The header file `charset.h' declares various functions that | |
61 | different pieces of GDB need to perform tasks like: | |
62 | ||
63 | - printing target strings and characters to the user's terminal | |
64 | (mostly target->host conversions), | |
65 | ||
66 | - building target-appropriate representations of strings and | |
67 | characters the user enters in expressions (mostly host->target | |
68 | conversions), | |
69 | ||
6c7a06a3 TT |
70 | and so on. |
71 | ||
72 | To avoid excessive code duplication and maintenance efforts, | |
73 | GDB simply requires a capable iconv function. Users on platforms | |
74 | without a suitable iconv can use the GNU iconv library. */ | |
234b45d4 KB |
75 | |
76 | \f | |
6c7a06a3 | 77 | #ifdef PHONY_ICONV |
234b45d4 | 78 | |
6c7a06a3 TT |
79 | /* Provide a phony iconv that does as little as possible. Also, |
80 | arrange for there to be a single available character set. */ | |
234b45d4 | 81 | |
6c7a06a3 TT |
82 | #undef GDB_DEFAULT_HOST_CHARSET |
83 | #define GDB_DEFAULT_HOST_CHARSET "ISO-8859-1" | |
84 | #define GDB_DEFAULT_TARGET_CHARSET "ISO-8859-1" | |
85 | #define GDB_DEFAULT_TARGET_WIDE_CHARSET "ISO-8859-1" | |
86 | #undef DEFAULT_CHARSET_NAMES | |
87 | #define DEFAULT_CHARSET_NAMES GDB_DEFAULT_HOST_CHARSET , | |
88 | ||
89 | #undef iconv_t | |
90 | #define iconv_t int | |
91 | #undef iconv_open | |
62234ccc | 92 | #define iconv_open phony_iconv_open |
6c7a06a3 | 93 | #undef iconv |
62234ccc | 94 | #define iconv phony_iconv |
6c7a06a3 | 95 | #undef iconv_close |
62234ccc | 96 | #define iconv_close phony_iconv_close |
6c7a06a3 | 97 | |
0dd7fb99 TT |
98 | #undef ICONV_CONST |
99 | #define ICONV_CONST const | |
100 | ||
76208fec | 101 | /* Some systems don't have EILSEQ, so we define it here, but not as |
e726d784 EZ |
102 | EINVAL, because callers of `iconv' want to distinguish EINVAL and |
103 | EILSEQ. This is what iconv.h from libiconv does as well. Note | |
104 | that wchar.h may also define EILSEQ, so this needs to be after we | |
105 | include wchar.h, which happens in defs.h through gdb_wchar.h. */ | |
106 | #ifndef EILSEQ | |
107 | #define EILSEQ ENOENT | |
108 | #endif | |
109 | ||
6c7a06a3 | 110 | iconv_t |
62234ccc | 111 | phony_iconv_open (const char *to, const char *from) |
6c7a06a3 | 112 | { |
b8899f2b | 113 | /* We allow conversions from UTF-32BE, wchar_t, and the host charset. |
6c7a06a3 | 114 | We allow conversions to wchar_t and the host charset. */ |
b8899f2b | 115 | if (strcmp (from, "UTF-32BE") && strcmp (from, "wchar_t") |
6c7a06a3 TT |
116 | && strcmp (from, GDB_DEFAULT_HOST_CHARSET)) |
117 | return -1; | |
118 | if (strcmp (to, "wchar_t") && strcmp (to, GDB_DEFAULT_HOST_CHARSET)) | |
119 | return -1; | |
234b45d4 | 120 | |
b8899f2b | 121 | /* Return 1 if we are converting from UTF-32BE, 0 otherwise. This is |
6c7a06a3 | 122 | used as a flag in calls to iconv. */ |
b8899f2b | 123 | return !strcmp (from, "UTF-32BE"); |
6c7a06a3 | 124 | } |
234b45d4 | 125 | |
6c7a06a3 | 126 | int |
62234ccc | 127 | phony_iconv_close (iconv_t arg) |
6c7a06a3 TT |
128 | { |
129 | return 0; | |
130 | } | |
234b45d4 | 131 | |
6c7a06a3 | 132 | size_t |
62234ccc TT |
133 | phony_iconv (iconv_t utf_flag, const char **inbuf, size_t *inbytesleft, |
134 | char **outbuf, size_t *outbytesleft) | |
6c7a06a3 | 135 | { |
b8899f2b | 136 | if (utf_flag) |
6c7a06a3 TT |
137 | { |
138 | while (*inbytesleft >= 4) | |
139 | { | |
140 | size_t j; | |
141 | unsigned long c = 0; | |
142 | ||
143 | for (j = 0; j < 4; ++j) | |
144 | { | |
145 | c <<= 8; | |
146 | c += (*inbuf)[j] & 0xff; | |
147 | } | |
148 | ||
149 | if (c >= 256) | |
150 | { | |
151 | errno = EILSEQ; | |
152 | return -1; | |
153 | } | |
154 | **outbuf = c & 0xff; | |
155 | ++*outbuf; | |
156 | --*outbytesleft; | |
157 | ||
158 | ++*inbuf; | |
159 | *inbytesleft -= 4; | |
160 | } | |
161 | if (*inbytesleft < 4) | |
162 | { | |
163 | errno = EINVAL; | |
164 | return -1; | |
165 | } | |
166 | } | |
167 | else | |
168 | { | |
169 | /* In all other cases we simply copy input bytes to the | |
170 | output. */ | |
171 | size_t amt = *inbytesleft; | |
c5504eaf | 172 | |
6c7a06a3 TT |
173 | if (amt > *outbytesleft) |
174 | amt = *outbytesleft; | |
175 | memcpy (*outbuf, *inbuf, amt); | |
176 | *inbuf += amt; | |
177 | *outbuf += amt; | |
178 | *inbytesleft -= amt; | |
179 | *outbytesleft -= amt; | |
180 | } | |
234b45d4 | 181 | |
6c7a06a3 TT |
182 | if (*inbytesleft) |
183 | { | |
184 | errno = E2BIG; | |
185 | return -1; | |
186 | } | |
234b45d4 | 187 | |
6c7a06a3 TT |
188 | /* The number of non-reversible conversions -- but they were all |
189 | reversible. */ | |
190 | return 0; | |
191 | } | |
234b45d4 | 192 | |
6c7a06a3 | 193 | #endif |
234b45d4 KB |
194 | |
195 | ||
196 | \f | |
197 | /* The global lists of character sets and translations. */ | |
198 | ||
199 | ||
e33d66ec EZ |
200 | #ifndef GDB_DEFAULT_TARGET_CHARSET |
201 | #define GDB_DEFAULT_TARGET_CHARSET "ISO-8859-1" | |
202 | #endif | |
203 | ||
6c7a06a3 | 204 | #ifndef GDB_DEFAULT_TARGET_WIDE_CHARSET |
b8899f2b | 205 | #define GDB_DEFAULT_TARGET_WIDE_CHARSET "UTF-32" |
6c7a06a3 TT |
206 | #endif |
207 | ||
208 | static const char *auto_host_charset_name = GDB_DEFAULT_HOST_CHARSET; | |
209 | static const char *host_charset_name = "auto"; | |
920d2a44 AC |
210 | static void |
211 | show_host_charset_name (struct ui_file *file, int from_tty, | |
212 | struct cmd_list_element *c, | |
213 | const char *value) | |
214 | { | |
6c7a06a3 TT |
215 | if (!strcmp (value, "auto")) |
216 | fprintf_filtered (file, | |
217 | _("The host character set is \"auto; currently %s\".\n"), | |
218 | auto_host_charset_name); | |
219 | else | |
220 | fprintf_filtered (file, _("The host character set is \"%s\".\n"), value); | |
920d2a44 AC |
221 | } |
222 | ||
f870a310 | 223 | static const char *target_charset_name = "auto"; |
920d2a44 AC |
224 | static void |
225 | show_target_charset_name (struct ui_file *file, int from_tty, | |
226 | struct cmd_list_element *c, const char *value) | |
227 | { | |
f870a310 TT |
228 | if (!strcmp (value, "auto")) |
229 | fprintf_filtered (file, | |
230 | _("The target character set is \"auto; " | |
231 | "currently %s\".\n"), | |
232 | gdbarch_auto_charset (get_current_arch ())); | |
233 | else | |
234 | fprintf_filtered (file, _("The target character set is \"%s\".\n"), | |
235 | value); | |
920d2a44 AC |
236 | } |
237 | ||
f870a310 | 238 | static const char *target_wide_charset_name = "auto"; |
6c7a06a3 | 239 | static void |
aff410f1 MS |
240 | show_target_wide_charset_name (struct ui_file *file, |
241 | int from_tty, | |
242 | struct cmd_list_element *c, | |
243 | const char *value) | |
e33d66ec | 244 | { |
f870a310 TT |
245 | if (!strcmp (value, "auto")) |
246 | fprintf_filtered (file, | |
247 | _("The target wide character set is \"auto; " | |
248 | "currently %s\".\n"), | |
249 | gdbarch_auto_wide_charset (get_current_arch ())); | |
250 | else | |
251 | fprintf_filtered (file, _("The target wide character set is \"%s\".\n"), | |
252 | value); | |
6c7a06a3 | 253 | } |
e33d66ec | 254 | |
6c7a06a3 | 255 | static const char *default_charset_names[] = |
e33d66ec | 256 | { |
6c7a06a3 | 257 | DEFAULT_CHARSET_NAMES |
e33d66ec EZ |
258 | 0 |
259 | }; | |
234b45d4 | 260 | |
6c7a06a3 | 261 | static const char **charset_enum; |
234b45d4 | 262 | |
6c7a06a3 TT |
263 | \f |
264 | /* If the target wide character set has big- or little-endian | |
265 | variants, these are the corresponding names. */ | |
266 | static const char *target_wide_charset_be_name; | |
267 | static const char *target_wide_charset_le_name; | |
234b45d4 | 268 | |
f870a310 TT |
269 | /* The architecture for which the BE- and LE-names are valid. */ |
270 | static struct gdbarch *be_le_arch; | |
271 | ||
272 | /* A helper function which sets the target wide big- and little-endian | |
273 | character set names, if possible. */ | |
234b45d4 | 274 | |
6c7a06a3 | 275 | static void |
f870a310 | 276 | set_be_le_names (struct gdbarch *gdbarch) |
234b45d4 | 277 | { |
6c7a06a3 | 278 | int i, len; |
f870a310 TT |
279 | const char *target_wide; |
280 | ||
281 | if (be_le_arch == gdbarch) | |
282 | return; | |
283 | be_le_arch = gdbarch; | |
234b45d4 | 284 | |
6c7a06a3 TT |
285 | target_wide_charset_le_name = NULL; |
286 | target_wide_charset_be_name = NULL; | |
234b45d4 | 287 | |
f870a310 TT |
288 | target_wide = target_wide_charset_name; |
289 | if (!strcmp (target_wide, "auto")) | |
290 | target_wide = gdbarch_auto_wide_charset (gdbarch); | |
291 | ||
292 | len = strlen (target_wide); | |
6c7a06a3 TT |
293 | for (i = 0; charset_enum[i]; ++i) |
294 | { | |
f870a310 | 295 | if (strncmp (target_wide, charset_enum[i], len)) |
6c7a06a3 TT |
296 | continue; |
297 | if ((charset_enum[i][len] == 'B' | |
298 | || charset_enum[i][len] == 'L') | |
299 | && charset_enum[i][len + 1] == 'E' | |
300 | && charset_enum[i][len + 2] == '\0') | |
301 | { | |
302 | if (charset_enum[i][len] == 'B') | |
303 | target_wide_charset_be_name = charset_enum[i]; | |
304 | else | |
305 | target_wide_charset_le_name = charset_enum[i]; | |
306 | } | |
307 | } | |
234b45d4 KB |
308 | } |
309 | ||
6c7a06a3 TT |
310 | /* 'Set charset', 'set host-charset', 'set target-charset', 'set |
311 | target-wide-charset', 'set charset' sfunc's. */ | |
234b45d4 KB |
312 | |
313 | static void | |
f870a310 | 314 | validate (struct gdbarch *gdbarch) |
234b45d4 | 315 | { |
6c7a06a3 TT |
316 | iconv_t desc; |
317 | const char *host_cset = host_charset (); | |
f870a310 TT |
318 | const char *target_cset = target_charset (gdbarch); |
319 | const char *target_wide_cset = target_wide_charset_name; | |
c5504eaf | 320 | |
f870a310 TT |
321 | if (!strcmp (target_wide_cset, "auto")) |
322 | target_wide_cset = gdbarch_auto_wide_charset (gdbarch); | |
234b45d4 | 323 | |
f870a310 | 324 | desc = iconv_open (target_wide_cset, host_cset); |
6c7a06a3 | 325 | if (desc == (iconv_t) -1) |
a73c6dcd | 326 | error (_("Cannot convert between character sets `%s' and `%s'"), |
f870a310 | 327 | target_wide_cset, host_cset); |
6c7a06a3 | 328 | iconv_close (desc); |
234b45d4 | 329 | |
f870a310 | 330 | desc = iconv_open (target_cset, host_cset); |
6c7a06a3 | 331 | if (desc == (iconv_t) -1) |
a73c6dcd | 332 | error (_("Cannot convert between character sets `%s' and `%s'"), |
f870a310 | 333 | target_cset, host_cset); |
6c7a06a3 | 334 | iconv_close (desc); |
234b45d4 | 335 | |
f870a310 TT |
336 | /* Clear the cache. */ |
337 | be_le_arch = NULL; | |
234b45d4 KB |
338 | } |
339 | ||
6c7a06a3 TT |
340 | /* This is the sfunc for the 'set charset' command. */ |
341 | static void | |
aff410f1 MS |
342 | set_charset_sfunc (char *charset, int from_tty, |
343 | struct cmd_list_element *c) | |
234b45d4 | 344 | { |
aff410f1 | 345 | /* CAREFUL: set the target charset here as well. */ |
6c7a06a3 | 346 | target_charset_name = host_charset_name; |
f870a310 | 347 | validate (get_current_arch ()); |
234b45d4 KB |
348 | } |
349 | ||
6c7a06a3 TT |
350 | /* 'set host-charset' command sfunc. We need a wrapper here because |
351 | the function needs to have a specific signature. */ | |
352 | static void | |
353 | set_host_charset_sfunc (char *charset, int from_tty, | |
354 | struct cmd_list_element *c) | |
234b45d4 | 355 | { |
f870a310 | 356 | validate (get_current_arch ()); |
234b45d4 KB |
357 | } |
358 | ||
6c7a06a3 TT |
359 | /* Wrapper for the 'set target-charset' command. */ |
360 | static void | |
361 | set_target_charset_sfunc (char *charset, int from_tty, | |
362 | struct cmd_list_element *c) | |
234b45d4 | 363 | { |
f870a310 | 364 | validate (get_current_arch ()); |
234b45d4 KB |
365 | } |
366 | ||
6c7a06a3 TT |
367 | /* Wrapper for the 'set target-wide-charset' command. */ |
368 | static void | |
369 | set_target_wide_charset_sfunc (char *charset, int from_tty, | |
370 | struct cmd_list_element *c) | |
234b45d4 | 371 | { |
f870a310 | 372 | validate (get_current_arch ()); |
234b45d4 KB |
373 | } |
374 | ||
6c7a06a3 TT |
375 | /* sfunc for the 'show charset' command. */ |
376 | static void | |
aff410f1 MS |
377 | show_charset (struct ui_file *file, int from_tty, |
378 | struct cmd_list_element *c, | |
6c7a06a3 | 379 | const char *name) |
234b45d4 | 380 | { |
6c7a06a3 TT |
381 | show_host_charset_name (file, from_tty, c, host_charset_name); |
382 | show_target_charset_name (file, from_tty, c, target_charset_name); | |
aff410f1 MS |
383 | show_target_wide_charset_name (file, from_tty, c, |
384 | target_wide_charset_name); | |
234b45d4 KB |
385 | } |
386 | ||
234b45d4 | 387 | \f |
6c7a06a3 | 388 | /* Accessor functions. */ |
234b45d4 | 389 | |
6c7a06a3 TT |
390 | const char * |
391 | host_charset (void) | |
234b45d4 | 392 | { |
6c7a06a3 TT |
393 | if (!strcmp (host_charset_name, "auto")) |
394 | return auto_host_charset_name; | |
395 | return host_charset_name; | |
234b45d4 KB |
396 | } |
397 | ||
6c7a06a3 | 398 | const char * |
f870a310 | 399 | target_charset (struct gdbarch *gdbarch) |
234b45d4 | 400 | { |
f870a310 TT |
401 | if (!strcmp (target_charset_name, "auto")) |
402 | return gdbarch_auto_charset (gdbarch); | |
6c7a06a3 | 403 | return target_charset_name; |
234b45d4 | 404 | } |
234b45d4 | 405 | |
6c7a06a3 | 406 | const char * |
f870a310 | 407 | target_wide_charset (struct gdbarch *gdbarch) |
234b45d4 | 408 | { |
f870a310 TT |
409 | enum bfd_endian byte_order = gdbarch_byte_order (gdbarch); |
410 | ||
411 | set_be_le_names (gdbarch); | |
e17a4113 | 412 | if (byte_order == BFD_ENDIAN_BIG) |
234b45d4 | 413 | { |
6c7a06a3 TT |
414 | if (target_wide_charset_be_name) |
415 | return target_wide_charset_be_name; | |
234b45d4 | 416 | } |
6c7a06a3 | 417 | else |
234b45d4 | 418 | { |
6c7a06a3 TT |
419 | if (target_wide_charset_le_name) |
420 | return target_wide_charset_le_name; | |
234b45d4 KB |
421 | } |
422 | ||
f870a310 TT |
423 | if (!strcmp (target_wide_charset_name, "auto")) |
424 | return gdbarch_auto_wide_charset (gdbarch); | |
425 | ||
6c7a06a3 | 426 | return target_wide_charset_name; |
234b45d4 KB |
427 | } |
428 | ||
234b45d4 | 429 | \f |
6c7a06a3 TT |
430 | /* Host character set management. For the time being, we assume that |
431 | the host character set is some superset of ASCII. */ | |
234b45d4 | 432 | |
6c7a06a3 TT |
433 | char |
434 | host_letter_to_control_character (char c) | |
234b45d4 | 435 | { |
6c7a06a3 TT |
436 | if (c == '?') |
437 | return 0177; | |
438 | return c & 0237; | |
234b45d4 KB |
439 | } |
440 | ||
6c7a06a3 TT |
441 | /* Convert a host character, C, to its hex value. C must already have |
442 | been validated using isxdigit. */ | |
234b45d4 | 443 | |
6c7a06a3 TT |
444 | int |
445 | host_hex_value (char c) | |
234b45d4 | 446 | { |
6c7a06a3 TT |
447 | if (isdigit (c)) |
448 | return c - '0'; | |
449 | if (c >= 'a' && c <= 'f') | |
450 | return 10 + c - 'a'; | |
451 | gdb_assert (c >= 'A' && c <= 'F'); | |
452 | return 10 + c - 'A'; | |
234b45d4 KB |
453 | } |
454 | ||
234b45d4 | 455 | \f |
6c7a06a3 | 456 | /* Public character management functions. */ |
234b45d4 | 457 | |
6c7a06a3 | 458 | /* A cleanup function which is run to close an iconv descriptor. */ |
234b45d4 | 459 | |
6c7a06a3 TT |
460 | static void |
461 | cleanup_iconv (void *p) | |
234b45d4 | 462 | { |
6c7a06a3 TT |
463 | iconv_t *descp = p; |
464 | iconv_close (*descp); | |
234b45d4 KB |
465 | } |
466 | ||
6c7a06a3 TT |
467 | void |
468 | convert_between_encodings (const char *from, const char *to, | |
469 | const gdb_byte *bytes, unsigned int num_bytes, | |
470 | int width, struct obstack *output, | |
471 | enum transliterations translit) | |
472 | { | |
473 | iconv_t desc; | |
474 | struct cleanup *cleanups; | |
475 | size_t inleft; | |
476 | char *inp; | |
477 | unsigned int space_request; | |
478 | ||
479 | /* Often, the host and target charsets will be the same. */ | |
480 | if (!strcmp (from, to)) | |
481 | { | |
482 | obstack_grow (output, bytes, num_bytes); | |
483 | return; | |
484 | } | |
234b45d4 | 485 | |
6c7a06a3 TT |
486 | desc = iconv_open (to, from); |
487 | if (desc == (iconv_t) -1) | |
9b20d036 | 488 | perror_with_name (_("Converting character sets")); |
6c7a06a3 | 489 | cleanups = make_cleanup (cleanup_iconv, &desc); |
234b45d4 | 490 | |
6c7a06a3 TT |
491 | inleft = num_bytes; |
492 | inp = (char *) bytes; | |
234b45d4 | 493 | |
6c7a06a3 | 494 | space_request = num_bytes; |
234b45d4 | 495 | |
6c7a06a3 | 496 | while (inleft > 0) |
234b45d4 | 497 | { |
6c7a06a3 TT |
498 | char *outp; |
499 | size_t outleft, r; | |
500 | int old_size; | |
501 | ||
502 | old_size = obstack_object_size (output); | |
503 | obstack_blank (output, space_request); | |
504 | ||
505 | outp = obstack_base (output) + old_size; | |
506 | outleft = space_request; | |
507 | ||
0dd7fb99 | 508 | r = iconv (desc, (ICONV_CONST char **) &inp, &inleft, &outp, &outleft); |
6c7a06a3 TT |
509 | |
510 | /* Now make sure that the object on the obstack only includes | |
511 | bytes we have converted. */ | |
512 | obstack_blank (output, - (int) outleft); | |
513 | ||
514 | if (r == (size_t) -1) | |
515 | { | |
516 | switch (errno) | |
517 | { | |
518 | case EILSEQ: | |
519 | { | |
520 | int i; | |
521 | ||
522 | /* Invalid input sequence. */ | |
523 | if (translit == translit_none) | |
3e43a32a MS |
524 | error (_("Could not convert character " |
525 | "to `%s' character set"), to); | |
6c7a06a3 TT |
526 | |
527 | /* We emit escape sequence for the bytes, skip them, | |
528 | and try again. */ | |
529 | for (i = 0; i < width; ++i) | |
530 | { | |
531 | char octal[5]; | |
532 | ||
533 | sprintf (octal, "\\%.3o", *inp & 0xff); | |
534 | obstack_grow_str (output, octal); | |
535 | ||
536 | ++inp; | |
537 | --inleft; | |
538 | } | |
539 | } | |
540 | break; | |
541 | ||
542 | case E2BIG: | |
543 | /* We ran out of space in the output buffer. Make it | |
544 | bigger next time around. */ | |
545 | space_request *= 2; | |
546 | break; | |
547 | ||
548 | case EINVAL: | |
549 | /* Incomplete input sequence. FIXME: ought to report this | |
550 | to the caller somehow. */ | |
551 | inleft = 0; | |
552 | break; | |
553 | ||
554 | default: | |
9b20d036 MS |
555 | perror_with_name (_("Internal error while " |
556 | "converting character sets")); | |
6c7a06a3 TT |
557 | } |
558 | } | |
234b45d4 | 559 | } |
234b45d4 | 560 | |
6c7a06a3 | 561 | do_cleanups (cleanups); |
234b45d4 KB |
562 | } |
563 | ||
e33d66ec | 564 | \f |
e33d66ec | 565 | |
6c7a06a3 TT |
566 | /* An iterator that returns host wchar_t's from a target string. */ |
567 | struct wchar_iterator | |
e33d66ec | 568 | { |
6c7a06a3 TT |
569 | /* The underlying iconv descriptor. */ |
570 | iconv_t desc; | |
e33d66ec | 571 | |
6c7a06a3 TT |
572 | /* The input string. This is updated as convert characters. */ |
573 | char *input; | |
574 | /* The number of bytes remaining in the input. */ | |
575 | size_t bytes; | |
e33d66ec | 576 | |
6c7a06a3 TT |
577 | /* The width of an input character. */ |
578 | size_t width; | |
e33d66ec | 579 | |
6c7a06a3 TT |
580 | /* The output buffer and its size. */ |
581 | gdb_wchar_t *out; | |
582 | size_t out_size; | |
583 | }; | |
234b45d4 | 584 | |
6c7a06a3 TT |
585 | /* Create a new iterator. */ |
586 | struct wchar_iterator * | |
aff410f1 MS |
587 | make_wchar_iterator (const gdb_byte *input, size_t bytes, |
588 | const char *charset, size_t width) | |
234b45d4 | 589 | { |
6c7a06a3 TT |
590 | struct wchar_iterator *result; |
591 | iconv_t desc; | |
234b45d4 | 592 | |
732f6a93 | 593 | desc = iconv_open (INTERMEDIATE_ENCODING, charset); |
6c7a06a3 | 594 | if (desc == (iconv_t) -1) |
9b20d036 | 595 | perror_with_name (_("Converting character sets")); |
234b45d4 | 596 | |
6c7a06a3 TT |
597 | result = XNEW (struct wchar_iterator); |
598 | result->desc = desc; | |
599 | result->input = (char *) input; | |
600 | result->bytes = bytes; | |
601 | result->width = width; | |
234b45d4 | 602 | |
6c7a06a3 TT |
603 | result->out = XNEW (gdb_wchar_t); |
604 | result->out_size = 1; | |
234b45d4 | 605 | |
6c7a06a3 | 606 | return result; |
e33d66ec | 607 | } |
234b45d4 | 608 | |
e33d66ec | 609 | static void |
6c7a06a3 | 610 | do_cleanup_iterator (void *p) |
e33d66ec | 611 | { |
6c7a06a3 | 612 | struct wchar_iterator *iter = p; |
234b45d4 | 613 | |
6c7a06a3 TT |
614 | iconv_close (iter->desc); |
615 | xfree (iter->out); | |
616 | xfree (iter); | |
234b45d4 KB |
617 | } |
618 | ||
6c7a06a3 TT |
619 | struct cleanup * |
620 | make_cleanup_wchar_iterator (struct wchar_iterator *iter) | |
e33d66ec | 621 | { |
6c7a06a3 | 622 | return make_cleanup (do_cleanup_iterator, iter); |
e33d66ec | 623 | } |
234b45d4 | 624 | |
6c7a06a3 TT |
625 | int |
626 | wchar_iterate (struct wchar_iterator *iter, | |
627 | enum wchar_iterate_result *out_result, | |
628 | gdb_wchar_t **out_chars, | |
629 | const gdb_byte **ptr, | |
630 | size_t *len) | |
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; | |
640 | while (iter->bytes > 0) | |
e33d66ec | 641 | { |
6c7a06a3 TT |
642 | char *outptr = (char *) &iter->out[0]; |
643 | char *orig_inptr = iter->input; | |
644 | size_t orig_in = iter->bytes; | |
645 | size_t out_avail = out_request * sizeof (gdb_wchar_t); | |
646 | size_t num; | |
0dd7fb99 | 647 | size_t r = iconv (iter->desc, |
aff410f1 MS |
648 | (ICONV_CONST char **) &iter->input, |
649 | &iter->bytes, &outptr, &out_avail); | |
c5504eaf | 650 | |
6c7a06a3 TT |
651 | if (r == (size_t) -1) |
652 | { | |
653 | switch (errno) | |
654 | { | |
655 | case EILSEQ: | |
aff410f1 MS |
656 | /* Invalid input sequence. We still might have |
657 | converted a character; if so, return it. */ | |
1558ab4c JK |
658 | if (out_avail < out_request * sizeof (gdb_wchar_t)) |
659 | break; | |
660 | ||
aff410f1 MS |
661 | /* Otherwise skip the first invalid character, and let |
662 | the caller know about it. */ | |
6c7a06a3 TT |
663 | *out_result = wchar_iterate_invalid; |
664 | *ptr = iter->input; | |
665 | *len = iter->width; | |
666 | iter->input += iter->width; | |
667 | iter->bytes -= iter->width; | |
668 | return 0; | |
669 | ||
670 | case E2BIG: | |
671 | /* We ran out of space. We still might have converted a | |
672 | character; if so, return it. Otherwise, grow the | |
673 | buffer and try again. */ | |
674 | if (out_avail < out_request * sizeof (gdb_wchar_t)) | |
675 | break; | |
676 | ||
677 | ++out_request; | |
678 | if (out_request > iter->out_size) | |
679 | { | |
680 | iter->out_size = out_request; | |
681 | iter->out = xrealloc (iter->out, | |
682 | out_request * sizeof (gdb_wchar_t)); | |
683 | } | |
684 | continue; | |
685 | ||
686 | case EINVAL: | |
687 | /* Incomplete input sequence. Let the caller know, and | |
688 | arrange for future calls to see EOF. */ | |
689 | *out_result = wchar_iterate_incomplete; | |
690 | *ptr = iter->input; | |
691 | *len = iter->bytes; | |
692 | iter->bytes = 0; | |
693 | return 0; | |
694 | ||
695 | default: | |
9b20d036 MS |
696 | perror_with_name (_("Internal error while " |
697 | "converting character sets")); | |
6c7a06a3 TT |
698 | } |
699 | } | |
700 | ||
701 | /* We converted something. */ | |
702 | num = out_request - out_avail / sizeof (gdb_wchar_t); | |
703 | *out_result = wchar_iterate_ok; | |
704 | *out_chars = iter->out; | |
705 | *ptr = orig_inptr; | |
706 | *len = orig_in - iter->bytes; | |
707 | return num; | |
e33d66ec | 708 | } |
6c7a06a3 TT |
709 | |
710 | /* Really done. */ | |
711 | *out_result = wchar_iterate_eof; | |
712 | return -1; | |
234b45d4 KB |
713 | } |
714 | ||
e33d66ec | 715 | \f |
6c7a06a3 | 716 | /* The charset.c module initialization function. */ |
234b45d4 | 717 | |
6c7a06a3 | 718 | extern initialize_file_ftype _initialize_charset; /* -Wmissing-prototype */ |
234b45d4 | 719 | |
6c7a06a3 | 720 | DEF_VEC_P (char_ptr); |
234b45d4 | 721 | |
6c7a06a3 | 722 | static VEC (char_ptr) *charsets; |
234b45d4 | 723 | |
6c7a06a3 | 724 | #ifdef PHONY_ICONV |
234b45d4 | 725 | |
6c7a06a3 TT |
726 | static void |
727 | find_charset_names (void) | |
234b45d4 | 728 | { |
6c7a06a3 TT |
729 | VEC_safe_push (char_ptr, charsets, GDB_DEFAULT_HOST_CHARSET); |
730 | VEC_safe_push (char_ptr, charsets, NULL); | |
234b45d4 KB |
731 | } |
732 | ||
6c7a06a3 | 733 | #else /* PHONY_ICONV */ |
fc3b640d TT |
734 | |
735 | /* Sometimes, libiconv redefines iconvlist as libiconvlist -- but | |
736 | provides different symbols in the static and dynamic libraries. | |
737 | So, configure may see libiconvlist but not iconvlist. But, calling | |
738 | iconvlist is the right thing to do and will work. Hence we do a | |
739 | check here but unconditionally call iconvlist below. */ | |
740 | #if defined (HAVE_ICONVLIST) || defined (HAVE_LIBICONVLIST) | |
234b45d4 | 741 | |
6c7a06a3 TT |
742 | /* A helper function that adds some character sets to the vector of |
743 | all character sets. This is a callback function for iconvlist. */ | |
744 | ||
745 | static int | |
746 | add_one (unsigned int count, const char *const *names, void *data) | |
234b45d4 | 747 | { |
6c7a06a3 | 748 | unsigned int i; |
234b45d4 | 749 | |
6c7a06a3 TT |
750 | for (i = 0; i < count; ++i) |
751 | VEC_safe_push (char_ptr, charsets, xstrdup (names[i])); | |
234b45d4 | 752 | |
6c7a06a3 | 753 | return 0; |
234b45d4 KB |
754 | } |
755 | ||
6c7a06a3 TT |
756 | static void |
757 | find_charset_names (void) | |
234b45d4 | 758 | { |
6c7a06a3 TT |
759 | iconvlist (add_one, NULL); |
760 | VEC_safe_push (char_ptr, charsets, NULL); | |
234b45d4 KB |
761 | } |
762 | ||
6c7a06a3 | 763 | #else |
234b45d4 | 764 | |
40b5c9fb DE |
765 | /* Return non-zero if LINE (output from iconv) should be ignored. |
766 | Older iconv programs (e.g. 2.2.2) include the human readable | |
767 | introduction even when stdout is not a tty. Newer versions omit | |
768 | the intro if stdout is not a tty. */ | |
769 | ||
770 | static int | |
771 | ignore_line_p (const char *line) | |
772 | { | |
773 | /* This table is used to filter the output. If this text appears | |
774 | anywhere in the line, it is ignored (strstr is used). */ | |
775 | static const char * const ignore_lines[] = | |
776 | { | |
777 | "The following", | |
778 | "not necessarily", | |
779 | "the FROM and TO", | |
780 | "listed with several", | |
781 | NULL | |
782 | }; | |
783 | int i; | |
784 | ||
785 | for (i = 0; ignore_lines[i] != NULL; ++i) | |
786 | { | |
787 | if (strstr (line, ignore_lines[i]) != NULL) | |
788 | return 1; | |
789 | } | |
790 | ||
791 | return 0; | |
792 | } | |
793 | ||
6c7a06a3 TT |
794 | static void |
795 | find_charset_names (void) | |
234b45d4 | 796 | { |
732f6a93 TT |
797 | struct pex_obj *child; |
798 | char *args[3]; | |
799 | int err, status; | |
800 | int fail = 1; | |
478aac75 | 801 | int flags; |
40b5c9fb | 802 | struct gdb_environ *iconv_env; |
478aac75 | 803 | char *iconv_program; |
40b5c9fb | 804 | |
aff410f1 MS |
805 | /* Older iconvs, e.g. 2.2.2, don't omit the intro text if stdout is |
806 | not a tty. We need to recognize it and ignore it. This text is | |
807 | subject to translation, so force LANGUAGE=C. */ | |
40b5c9fb DE |
808 | iconv_env = make_environ (); |
809 | init_environ (iconv_env); | |
810 | set_in_environ (iconv_env, "LANGUAGE", "C"); | |
811 | set_in_environ (iconv_env, "LC_ALL", "C"); | |
732f6a93 | 812 | |
40618926 | 813 | child = pex_init (PEX_USE_PIPES, "iconv", NULL); |
732f6a93 | 814 | |
478aac75 DE |
815 | #ifdef ICONV_BIN |
816 | { | |
817 | char *iconv_dir = relocate_gdb_directory (ICONV_BIN, | |
818 | ICONV_BIN_RELOCATABLE); | |
819 | iconv_program = concat (iconv_dir, SLASH_STRING, "iconv", NULL); | |
820 | xfree (iconv_dir); | |
821 | } | |
822 | #else | |
823 | iconv_program = xstrdup ("iconv"); | |
824 | #endif | |
825 | args[0] = iconv_program; | |
732f6a93 TT |
826 | args[1] = "-l"; |
827 | args[2] = NULL; | |
478aac75 DE |
828 | flags = PEX_STDERR_TO_STDOUT; |
829 | #ifndef ICONV_BIN | |
830 | flags |= PEX_SEARCH; | |
831 | #endif | |
732f6a93 | 832 | /* Note that we simply ignore errors here. */ |
478aac75 DE |
833 | if (!pex_run_in_environment (child, flags, |
834 | args[0], args, environ_vector (iconv_env), | |
40b5c9fb | 835 | NULL, NULL, &err)) |
732f6a93 TT |
836 | { |
837 | FILE *in = pex_read_output (child, 0); | |
838 | ||
839 | /* POSIX says that iconv -l uses an unspecified format. We | |
840 | parse the glibc and libiconv formats; feel free to add others | |
841 | as needed. */ | |
40b5c9fb | 842 | |
732f6a93 TT |
843 | while (!feof (in)) |
844 | { | |
845 | /* The size of buf is chosen arbitrarily. */ | |
846 | char buf[1024]; | |
847 | char *start, *r; | |
8ea13695 | 848 | int len; |
732f6a93 TT |
849 | |
850 | r = fgets (buf, sizeof (buf), in); | |
851 | if (!r) | |
852 | break; | |
853 | len = strlen (r); | |
854 | if (len <= 3) | |
855 | continue; | |
40b5c9fb DE |
856 | if (ignore_line_p (r)) |
857 | continue; | |
858 | ||
732f6a93 TT |
859 | /* Strip off the newline. */ |
860 | --len; | |
861 | /* Strip off one or two '/'s. glibc will print lines like | |
862 | "8859_7//", but also "10646-1:1993/UCS4/". */ | |
863 | if (buf[len - 1] == '/') | |
864 | --len; | |
865 | if (buf[len - 1] == '/') | |
866 | --len; | |
867 | buf[len] = '\0'; | |
868 | ||
869 | /* libiconv will print multiple entries per line, separated | |
aff410f1 MS |
870 | by spaces. Older iconvs will print multiple entries per |
871 | line, indented by two spaces, and separated by ", " | |
40b5c9fb | 872 | (i.e. the human readable form). */ |
732f6a93 TT |
873 | start = buf; |
874 | while (1) | |
875 | { | |
876 | int keep_going; | |
877 | char *p; | |
878 | ||
40b5c9fb DE |
879 | /* Skip leading blanks. */ |
880 | for (p = start; *p && *p == ' '; ++p) | |
881 | ; | |
882 | start = p; | |
883 | /* Find the next space, comma, or end-of-line. */ | |
884 | for ( ; *p && *p != ' ' && *p != ','; ++p) | |
732f6a93 TT |
885 | ; |
886 | /* Ignore an empty result. */ | |
887 | if (p == start) | |
888 | break; | |
889 | keep_going = *p; | |
890 | *p = '\0'; | |
891 | VEC_safe_push (char_ptr, charsets, xstrdup (start)); | |
892 | if (!keep_going) | |
893 | break; | |
894 | /* Skip any extra spaces. */ | |
895 | for (start = p + 1; *start && *start == ' '; ++start) | |
896 | ; | |
897 | } | |
898 | } | |
234b45d4 | 899 | |
732f6a93 TT |
900 | if (pex_get_status (child, 1, &status) |
901 | && WIFEXITED (status) && !WEXITSTATUS (status)) | |
902 | fail = 0; | |
234b45d4 | 903 | |
6c7a06a3 | 904 | } |
234b45d4 | 905 | |
478aac75 | 906 | xfree (iconv_program); |
732f6a93 | 907 | pex_free (child); |
40b5c9fb | 908 | free_environ (iconv_env); |
234b45d4 | 909 | |
732f6a93 TT |
910 | if (fail) |
911 | { | |
912 | /* Some error occurred, so drop the vector. */ | |
913 | int ix; | |
914 | char *elt; | |
915 | for (ix = 0; VEC_iterate (char_ptr, charsets, ix, elt); ++ix) | |
916 | xfree (elt); | |
917 | VEC_truncate (char_ptr, charsets, 0); | |
918 | } | |
919 | else | |
920 | VEC_safe_push (char_ptr, charsets, NULL); | |
6c7a06a3 | 921 | } |
234b45d4 | 922 | |
fc3b640d | 923 | #endif /* HAVE_ICONVLIST || HAVE_LIBICONVLIST */ |
6c7a06a3 | 924 | #endif /* PHONY_ICONV */ |
234b45d4 | 925 | |
f870a310 TT |
926 | /* The "auto" target charset used by default_auto_charset. */ |
927 | static const char *auto_target_charset_name = GDB_DEFAULT_TARGET_CHARSET; | |
928 | ||
929 | const char * | |
930 | default_auto_charset (void) | |
931 | { | |
932 | return auto_target_charset_name; | |
933 | } | |
934 | ||
935 | const char * | |
936 | default_auto_wide_charset (void) | |
937 | { | |
938 | return GDB_DEFAULT_TARGET_WIDE_CHARSET; | |
939 | } | |
940 | ||
bcb28afc PM |
941 | |
942 | #ifdef USE_INTERMEDIATE_ENCODING_FUNCTION | |
943 | /* Macro used for UTF or UCS endianness suffix. */ | |
944 | #if WORDS_BIGENDIAN | |
945 | #define ENDIAN_SUFFIX "BE" | |
946 | #else | |
947 | #define ENDIAN_SUFFIX "LE" | |
948 | #endif | |
949 | ||
950 | /* The code below serves to generate a compile time error if | |
951 | gdb_wchar_t type is not of size 2 nor 4, despite the fact that | |
952 | macro __STDC_ISO_10646__ is defined. | |
953 | This is better than a gdb_assert call, because GDB cannot handle | |
954 | strings correctly if this size is different. */ | |
955 | ||
956 | extern char your_gdb_wchar_t_is_bogus[(sizeof (gdb_wchar_t) == 2 | |
957 | || sizeof (gdb_wchar_t) == 4) | |
958 | ? 1 : -1]; | |
959 | ||
960 | /* intermediate_encoding returns the charset unsed internally by | |
961 | GDB to convert between target and host encodings. As the test above | |
962 | compiled, sizeof (gdb_wchar_t) is either 2 or 4 bytes. | |
963 | UTF-16/32 is tested first, UCS-2/4 is tested as a second option, | |
964 | otherwise an error is generated. */ | |
965 | ||
966 | const char * | |
967 | intermediate_encoding (void) | |
968 | { | |
969 | iconv_t desc; | |
970 | static const char *stored_result = NULL; | |
971 | char *result; | |
972 | int i; | |
973 | ||
974 | if (stored_result) | |
975 | return stored_result; | |
976 | result = xstrprintf ("UTF-%d%s", (int) (sizeof (gdb_wchar_t) * 8), | |
977 | ENDIAN_SUFFIX); | |
978 | /* Check that the name is supported by iconv_open. */ | |
979 | desc = iconv_open (result, host_charset ()); | |
980 | if (desc != (iconv_t) -1) | |
981 | { | |
982 | iconv_close (desc); | |
983 | stored_result = result; | |
984 | return result; | |
985 | } | |
986 | /* Not valid, free the allocated memory. */ | |
987 | xfree (result); | |
988 | /* Second try, with UCS-2 type. */ | |
989 | result = xstrprintf ("UCS-%d%s", (int) sizeof (gdb_wchar_t), | |
990 | ENDIAN_SUFFIX); | |
991 | /* Check that the name is supported by iconv_open. */ | |
992 | desc = iconv_open (result, host_charset ()); | |
993 | if (desc != (iconv_t) -1) | |
994 | { | |
995 | iconv_close (desc); | |
996 | stored_result = result; | |
997 | return result; | |
998 | } | |
999 | /* Not valid, free the allocated memory. */ | |
1000 | xfree (result); | |
1001 | /* No valid charset found, generate error here. */ | |
1002 | error (_("Unable to find a vaild charset for string conversions")); | |
1003 | } | |
1004 | ||
1005 | #endif /* USE_INTERMEDIATE_ENCODING_FUNCTION */ | |
1006 | ||
234b45d4 KB |
1007 | void |
1008 | _initialize_charset (void) | |
1009 | { | |
f870a310 | 1010 | /* The first element is always "auto". */ |
732f6a93 | 1011 | VEC_safe_push (char_ptr, charsets, xstrdup ("auto")); |
6c7a06a3 TT |
1012 | find_charset_names (); |
1013 | ||
1014 | if (VEC_length (char_ptr, charsets) > 1) | |
1015 | charset_enum = (const char **) VEC_address (char_ptr, charsets); | |
1016 | else | |
1017 | charset_enum = default_charset_names; | |
1018 | ||
1019 | #ifndef PHONY_ICONV | |
1020 | #ifdef HAVE_LANGINFO_CODESET | |
f870a310 TT |
1021 | /* The result of nl_langinfo may be overwritten later. This may |
1022 | leak a little memory, if the user later changes the host charset, | |
1023 | but that doesn't matter much. */ | |
1024 | auto_host_charset_name = xstrdup (nl_langinfo (CODESET)); | |
aff410f1 MS |
1025 | /* Solaris will return `646' here -- but the Solaris iconv then does |
1026 | not accept this. Darwin (and maybe FreeBSD) may return "" here, | |
06be6983 TG |
1027 | which GNU libiconv doesn't like (infinite loop). */ |
1028 | if (!strcmp (auto_host_charset_name, "646") || !*auto_host_charset_name) | |
58720494 | 1029 | auto_host_charset_name = "ASCII"; |
f870a310 TT |
1030 | auto_target_charset_name = auto_host_charset_name; |
1031 | #elif defined (USE_WIN32API) | |
1032 | { | |
3e43a32a MS |
1033 | /* "CP" + x<=5 digits + paranoia. */ |
1034 | static char w32_host_default_charset[16]; | |
f870a310 TT |
1035 | |
1036 | snprintf (w32_host_default_charset, sizeof w32_host_default_charset, | |
1037 | "CP%d", GetACP()); | |
1038 | auto_host_charset_name = w32_host_default_charset; | |
1039 | auto_target_charset_name = auto_host_charset_name; | |
1040 | } | |
6c7a06a3 TT |
1041 | #endif |
1042 | #endif | |
e33d66ec | 1043 | |
7ab04401 | 1044 | add_setshow_enum_cmd ("charset", class_support, |
f870a310 | 1045 | charset_enum, &host_charset_name, _("\ |
7ab04401 AC |
1046 | Set the host and target character sets."), _("\ |
1047 | Show the host and target character sets."), _("\ | |
3d263c1d BI |
1048 | The `host character set' is the one used by the system GDB is running on.\n\ |
1049 | The `target character set' is the one used by the program being debugged.\n\ | |
1050 | You may only use supersets of ASCII for your host character set; GDB does\n\ | |
1051 | not support any others.\n\ | |
1052 | To see a list of the character sets GDB supports, type `set charset <TAB>'."), | |
7ab04401 AC |
1053 | /* Note that the sfunc below needs to set |
1054 | target_charset_name, because the 'set | |
1055 | charset' command sets two variables. */ | |
1056 | set_charset_sfunc, | |
1057 | show_charset, | |
1058 | &setlist, &showlist); | |
1059 | ||
1060 | add_setshow_enum_cmd ("host-charset", class_support, | |
6c7a06a3 | 1061 | charset_enum, &host_charset_name, _("\ |
7ab04401 AC |
1062 | Set the host character set."), _("\ |
1063 | Show the host character set."), _("\ | |
3d263c1d BI |
1064 | The `host character set' is the one used by the system GDB is running on.\n\ |
1065 | You may only use supersets of ASCII for your host character set; GDB does\n\ | |
ac74f770 MS |
1066 | not support any others.\n\ |
1067 | To see a list of the character sets GDB supports, type `set host-charset <TAB>'."), | |
7ab04401 | 1068 | set_host_charset_sfunc, |
920d2a44 | 1069 | show_host_charset_name, |
7ab04401 AC |
1070 | &setlist, &showlist); |
1071 | ||
1072 | add_setshow_enum_cmd ("target-charset", class_support, | |
f870a310 | 1073 | charset_enum, &target_charset_name, _("\ |
7ab04401 AC |
1074 | Set the target character set."), _("\ |
1075 | Show the target character set."), _("\ | |
3d263c1d BI |
1076 | The `target character set' is the one used by the program being debugged.\n\ |
1077 | GDB translates characters and strings between the host and target\n\ | |
b670013c | 1078 | character sets as needed.\n\ |
ac74f770 | 1079 | To see a list of the character sets GDB supports, type `set target-charset'<TAB>"), |
7ab04401 | 1080 | set_target_charset_sfunc, |
920d2a44 | 1081 | show_target_charset_name, |
7ab04401 | 1082 | &setlist, &showlist); |
6c7a06a3 TT |
1083 | |
1084 | add_setshow_enum_cmd ("target-wide-charset", class_support, | |
f870a310 | 1085 | charset_enum, &target_wide_charset_name, |
6c7a06a3 TT |
1086 | _("\ |
1087 | Set the target wide character set."), _("\ | |
1088 | Show the target wide character set."), _("\ | |
3e43a32a MS |
1089 | The `target wide character set' is the one used by the program being debugged.\ |
1090 | \nIn particular it is the encoding used by `wchar_t'.\n\ | |
6c7a06a3 TT |
1091 | GDB translates characters and strings between the host and target\n\ |
1092 | character sets as needed.\n\ | |
1093 | To see a list of the character sets GDB supports, type\n\ | |
1094 | `set target-wide-charset'<TAB>"), | |
1095 | set_target_wide_charset_sfunc, | |
1096 | show_target_wide_charset_name, | |
1097 | &setlist, &showlist); | |
234b45d4 | 1098 | } |