1 | /* message.c |
---|
2 | * Fairly general purpose message and error routines |
---|
3 | * Copyright (C) 1993-2022 Olly Betts |
---|
4 | * |
---|
5 | * This program is free software; you can redistribute it and/or modify |
---|
6 | * it under the terms of the GNU General Public License as published by |
---|
7 | * the Free Software Foundation; either version 2 of the License, or |
---|
8 | * (at your option) any later version. |
---|
9 | * |
---|
10 | * This program is distributed in the hope that it will be useful, |
---|
11 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
12 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
---|
13 | * GNU General Public License for more details. |
---|
14 | * |
---|
15 | * You should have received a copy of the GNU General Public License |
---|
16 | * along with this program; if not, write to the Free Software |
---|
17 | * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA |
---|
18 | */ |
---|
19 | |
---|
20 | /*#define DEBUG 1*/ |
---|
21 | |
---|
22 | #include <config.h> |
---|
23 | |
---|
24 | #include <stdio.h> |
---|
25 | #include <stdlib.h> |
---|
26 | #include <string.h> |
---|
27 | #include <ctype.h> |
---|
28 | #include <limits.h> |
---|
29 | #include <errno.h> |
---|
30 | #include <locale.h> |
---|
31 | |
---|
32 | #include "cmdline.h" |
---|
33 | #include "whichos.h" |
---|
34 | #include "filename.h" |
---|
35 | #include "message.h" |
---|
36 | #include "osalloc.h" |
---|
37 | #include "filelist.h" |
---|
38 | #include "debug.h" |
---|
39 | #include "str.h" |
---|
40 | |
---|
41 | #ifdef AVEN |
---|
42 | # include "aven.h" |
---|
43 | #endif |
---|
44 | |
---|
45 | #if OS_WIN32 |
---|
46 | # define WIN32_LEAN_AND_MEAN |
---|
47 | # include <windows.h> |
---|
48 | #elif OS_UNIX |
---|
49 | # include <sys/types.h> |
---|
50 | #endif |
---|
51 | |
---|
52 | #include <sys/stat.h> |
---|
53 | |
---|
54 | /* For funcs which want to be immune from messing around with different |
---|
55 | * calling conventions */ |
---|
56 | #ifndef CDECL |
---|
57 | # define CDECL |
---|
58 | #endif |
---|
59 | |
---|
60 | int msg_warnings = 0; /* keep track of how many warnings we've given */ |
---|
61 | int msg_errors = 0; /* and how many (non-fatal) errors */ |
---|
62 | |
---|
63 | /* in case osmalloc() fails before appname_copy is set up */ |
---|
64 | static const char *appname_copy = "anonymous program"; |
---|
65 | |
---|
66 | /* Path to use to look for executables (used by aven to find cavern). */ |
---|
67 | static const char *exe_pth = ""; |
---|
68 | |
---|
69 | /* error code for failed osmalloc and osrealloc calls */ |
---|
70 | static void |
---|
71 | outofmem(OSSIZE_T size) |
---|
72 | { |
---|
73 | /* TRANSLATORS: "%lu" is a placeholder for the number of bytes which Survex |
---|
74 | * was trying to allocate space for. */ |
---|
75 | fatalerror(/*Out of memory (couldn’t find %lu bytes).*/1, |
---|
76 | (unsigned long)size); |
---|
77 | } |
---|
78 | |
---|
79 | #ifdef TOMBSTONES |
---|
80 | #define TOMBSTONE_SIZE 16 |
---|
81 | static const char tombstone[TOMBSTONE_SIZE] = "012345\xfftombstone"; |
---|
82 | #endif |
---|
83 | |
---|
84 | /* malloc with error catching if it fails. Also allows us to write special |
---|
85 | * versions easily eg for MS Windows. |
---|
86 | */ |
---|
87 | void * |
---|
88 | osmalloc(OSSIZE_T size) |
---|
89 | { |
---|
90 | void *p; |
---|
91 | #ifdef TOMBSTONES |
---|
92 | size += TOMBSTONE_SIZE * 2; |
---|
93 | p = malloc(size); |
---|
94 | #else |
---|
95 | p = xosmalloc(size); |
---|
96 | #endif |
---|
97 | if (p == NULL) outofmem(size); |
---|
98 | #ifdef TOMBSTONES |
---|
99 | printf("osmalloc truep=%p truesize=%d\n", p, size); |
---|
100 | memcpy(p, tombstone, TOMBSTONE_SIZE); |
---|
101 | memcpy(p + size - TOMBSTONE_SIZE, tombstone, TOMBSTONE_SIZE); |
---|
102 | *(size_t *)p = size; |
---|
103 | p += TOMBSTONE_SIZE; |
---|
104 | #endif |
---|
105 | return p; |
---|
106 | } |
---|
107 | |
---|
108 | /* realloc with error catching if it fails. */ |
---|
109 | void * |
---|
110 | osrealloc(void *p, OSSIZE_T size) |
---|
111 | { |
---|
112 | /* some pre-ANSI realloc implementations don't cope with a NULL pointer */ |
---|
113 | if (p == NULL) { |
---|
114 | p = xosmalloc(size); |
---|
115 | } else { |
---|
116 | #ifdef TOMBSTONES |
---|
117 | int true_size; |
---|
118 | size += TOMBSTONE_SIZE * 2; |
---|
119 | p -= TOMBSTONE_SIZE; |
---|
120 | true_size = *(size_t *)p; |
---|
121 | printf("osrealloc (in truep=%p truesize=%d)\n", p, true_size); |
---|
122 | if (memcmp(p + sizeof(size_t), tombstone + sizeof(size_t), |
---|
123 | TOMBSTONE_SIZE - sizeof(size_t)) != 0) { |
---|
124 | printf("start tombstone for block %p, size %d corrupted!", |
---|
125 | p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2); |
---|
126 | } |
---|
127 | if (memcmp(p + true_size - TOMBSTONE_SIZE, tombstone, |
---|
128 | TOMBSTONE_SIZE) != 0) { |
---|
129 | printf("end tombstone for block %p, size %d corrupted!", |
---|
130 | p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2); |
---|
131 | } |
---|
132 | p = realloc(p, size); |
---|
133 | if (p == NULL) outofmem(size); |
---|
134 | printf("osrealloc truep=%p truesize=%d\n", p, size); |
---|
135 | memcpy(p, tombstone, TOMBSTONE_SIZE); |
---|
136 | memcpy(p + size - TOMBSTONE_SIZE, tombstone, TOMBSTONE_SIZE); |
---|
137 | *(size_t *)p = size; |
---|
138 | p += TOMBSTONE_SIZE; |
---|
139 | #else |
---|
140 | p = xosrealloc(p, size); |
---|
141 | #endif |
---|
142 | } |
---|
143 | if (p == NULL) outofmem(size); |
---|
144 | return p; |
---|
145 | } |
---|
146 | |
---|
147 | char * |
---|
148 | osstrdup(const char *str) |
---|
149 | { |
---|
150 | char *p; |
---|
151 | OSSIZE_T len; |
---|
152 | len = strlen(str) + 1; |
---|
153 | p = osmalloc(len); |
---|
154 | memcpy(p, str, len); |
---|
155 | return p; |
---|
156 | } |
---|
157 | |
---|
158 | /* osfree is usually just a macro in osalloc.h */ |
---|
159 | #ifdef TOMBSTONES |
---|
160 | void |
---|
161 | osfree(void *p) |
---|
162 | { |
---|
163 | int true_size; |
---|
164 | if (!p) return; |
---|
165 | p -= TOMBSTONE_SIZE; |
---|
166 | true_size = *(size_t *)p; |
---|
167 | printf("osfree truep=%p truesize=%d\n", p, true_size); |
---|
168 | if (memcmp(p + sizeof(size_t), tombstone + sizeof(size_t), |
---|
169 | TOMBSTONE_SIZE - sizeof(size_t)) != 0) { |
---|
170 | printf("start tombstone for block %p, size %d corrupted!", |
---|
171 | p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2); |
---|
172 | } |
---|
173 | if (memcmp(p + true_size - TOMBSTONE_SIZE, tombstone, |
---|
174 | TOMBSTONE_SIZE) != 0) { |
---|
175 | printf("end tombstone for block %p, size %d corrupted!", |
---|
176 | p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2); |
---|
177 | } |
---|
178 | free(p); |
---|
179 | } |
---|
180 | #endif |
---|
181 | |
---|
182 | static int |
---|
183 | default_charset(void) |
---|
184 | { |
---|
185 | if (getenv("SURVEX_UTF8")) return CHARSET_UTF8; |
---|
186 | #if OS_WIN32 |
---|
187 | # ifdef AVEN |
---|
188 | # define CODEPAGE GetACP() |
---|
189 | # else |
---|
190 | # define CODEPAGE GetConsoleOutputCP() |
---|
191 | # endif |
---|
192 | switch (CODEPAGE) { |
---|
193 | case 0: return CHARSET_UTF8; |
---|
194 | case 1252: return CHARSET_WINCP1252; |
---|
195 | case 1250: return CHARSET_WINCP1250; |
---|
196 | case 850: return CHARSET_DOSCP850; |
---|
197 | } |
---|
198 | return CHARSET_USASCII; |
---|
199 | #elif OS_UNIX |
---|
200 | #ifdef AVEN |
---|
201 | return CHARSET_UTF8; |
---|
202 | #else |
---|
203 | const char *p = getenv("LC_ALL"); |
---|
204 | if (p == NULL || p[0] == '\0') { |
---|
205 | p = getenv("LC_CTYPE"); |
---|
206 | if (p == NULL || p[0] == '\0') { |
---|
207 | p = getenv("LANG"); |
---|
208 | /* Something (AutoCAD?) on Microsoft Windows sets LANG to a number. */ |
---|
209 | if (p == NULL || !isalpha((unsigned char)p[0])) |
---|
210 | p = msg_lang; |
---|
211 | } |
---|
212 | } |
---|
213 | |
---|
214 | if (p) { |
---|
215 | const char *q = strchr(p, '.'); |
---|
216 | if (q) p = q + 1; |
---|
217 | } |
---|
218 | |
---|
219 | if (p) { |
---|
220 | const char *chset = p; |
---|
221 | size_t name_len; |
---|
222 | |
---|
223 | while (*p != '\0' && *p != '@') p++; |
---|
224 | |
---|
225 | name_len = p - chset; |
---|
226 | |
---|
227 | if (name_len) { |
---|
228 | int only_digit = 1; |
---|
229 | size_t cnt; |
---|
230 | |
---|
231 | for (cnt = 0; cnt < name_len; ++cnt) |
---|
232 | if (isalpha((unsigned char)chset[cnt])) { |
---|
233 | only_digit = 0; |
---|
234 | break; |
---|
235 | } |
---|
236 | |
---|
237 | if (only_digit) goto iso; |
---|
238 | |
---|
239 | switch (tolower((unsigned char)chset[0])) { |
---|
240 | case 'i': |
---|
241 | if (tolower((unsigned char)chset[1]) == 's' && |
---|
242 | tolower((unsigned char)chset[2]) == 'o') { |
---|
243 | chset += 3; |
---|
244 | iso: |
---|
245 | if (strncmp(chset, "8859", 4) == 0) { |
---|
246 | chset += 4; |
---|
247 | while (chset < p && *chset && !isdigit((unsigned char)*chset)) |
---|
248 | chset++; |
---|
249 | switch (atoi(chset)) { |
---|
250 | case 1: return CHARSET_ISO_8859_1; |
---|
251 | case 2: return CHARSET_ISO_8859_2; |
---|
252 | case 15: return CHARSET_ISO_8859_15; |
---|
253 | default: return CHARSET_USASCII; |
---|
254 | } |
---|
255 | } |
---|
256 | } |
---|
257 | break; |
---|
258 | case 'u': |
---|
259 | if (tolower((unsigned char)chset[1]) == 't' && |
---|
260 | tolower((unsigned char)chset[2]) == 'f') { |
---|
261 | chset += 3; |
---|
262 | while (chset < p && *chset && !isdigit((unsigned char)*chset)) |
---|
263 | chset++; |
---|
264 | switch (atoi(chset)) { |
---|
265 | case 8: return CHARSET_UTF8; |
---|
266 | default: return CHARSET_USASCII; |
---|
267 | } |
---|
268 | } |
---|
269 | } |
---|
270 | } |
---|
271 | } |
---|
272 | return CHARSET_USASCII; |
---|
273 | #endif |
---|
274 | #else |
---|
275 | # error Do not know operating system! |
---|
276 | #endif |
---|
277 | } |
---|
278 | |
---|
279 | /* It seems that Swedish and maybe some other scandanavian languages don't |
---|
280 | * transliterate ä to ae - but it seems there may be conflicting views |
---|
281 | * on this... |
---|
282 | */ |
---|
283 | #define umlaut_to_e() 1 |
---|
284 | |
---|
285 | /* values <= 127 already dealt with */ |
---|
286 | static int |
---|
287 | add_unicode(int charset, unsigned char *p, int value) |
---|
288 | { |
---|
289 | #ifdef DEBUG |
---|
290 | fprintf(stderr, "add_unicode(%d, %p, %d)\n", charset, p, value); |
---|
291 | #endif |
---|
292 | if (value == 0) return 0; |
---|
293 | switch (charset) { |
---|
294 | case CHARSET_USASCII: |
---|
295 | if (value < 0x80) { |
---|
296 | *p = value; |
---|
297 | return 1; |
---|
298 | } |
---|
299 | break; |
---|
300 | case CHARSET_ISO_8859_1: |
---|
301 | if (value < 0x100) { |
---|
302 | *p = value; |
---|
303 | return 1; |
---|
304 | } |
---|
305 | break; |
---|
306 | case CHARSET_ISO_8859_2: |
---|
307 | if (value >= 0xa0) { |
---|
308 | int v = 0; |
---|
309 | switch (value) { |
---|
310 | case 0xa0: case 0xa4: case 0xa7: case 0xa8: case 0xad: case 0xb0: |
---|
311 | case 0xb4: case 0xb8: case 0xc1: case 0xc2: case 0xc4: case 0xc7: |
---|
312 | case 0xc9: case 0xcb: case 0xcd: case 0xce: case 0xd3: case 0xd4: |
---|
313 | case 0xd6: case 0xd7: case 0xda: case 0xdc: case 0xdd: case 0xdf: |
---|
314 | case 0xe1: case 0xe2: case 0xe4: case 0xe7: case 0xe9: case 0xeb: |
---|
315 | case 0xed: case 0xee: case 0xf3: case 0xf4: case 0xf6: case 0xf7: |
---|
316 | case 0xfa: case 0xfc: case 0xfd: |
---|
317 | v = value; break; |
---|
318 | case 0x104: v = '\xa1'; break; |
---|
319 | case 0x2d8: v = '\xa2'; break; |
---|
320 | case 0x141: v = '\xa3'; break; |
---|
321 | case 0x13d: v = '\xa5'; break; |
---|
322 | case 0x15a: v = '\xa6'; break; |
---|
323 | case 0x160: v = '\xa9'; break; |
---|
324 | case 0x15e: v = '\xaa'; break; /* Scedil */ |
---|
325 | case 0x164: v = '\xab'; break; |
---|
326 | case 0x179: v = '\xac'; break; |
---|
327 | case 0x17d: v = '\xae'; break; |
---|
328 | case 0x17b: v = '\xaf'; break; |
---|
329 | case 0x105: v = '\xb1'; break; |
---|
330 | case 0x2db: v = '\xb2'; break; |
---|
331 | case 0x142: v = '\xb3'; break; |
---|
332 | case 0x13e: v = '\xb5'; break; |
---|
333 | case 0x15b: v = '\xb6'; break; |
---|
334 | case 0x2c7: v = '\xb7'; break; |
---|
335 | case 0x161: v = '\xb9'; break; |
---|
336 | case 0x15f: v = '\xba'; break; /* scedil */ |
---|
337 | case 0x165: v = '\xbb'; break; |
---|
338 | case 0x17a: v = '\xbc'; break; |
---|
339 | case 0x2dd: v = '\xbd'; break; |
---|
340 | case 0x17e: v = '\xbe'; break; |
---|
341 | case 0x17c: v = '\xbf'; break; |
---|
342 | case 0x154: v = '\xc0'; break; |
---|
343 | case 0x102: v = '\xc3'; break; |
---|
344 | case 0x139: v = '\xc5'; break; |
---|
345 | case 0x106: v = '\xc6'; break; |
---|
346 | case 0x10c: v = '\xc8'; break; |
---|
347 | case 0x118: v = '\xca'; break; |
---|
348 | case 0x11a: v = '\xcc'; break; |
---|
349 | case 0x10e: v = '\xcf'; break; |
---|
350 | case 0x110: v = '\xd0'; break; |
---|
351 | case 0x143: v = '\xd1'; break; |
---|
352 | case 0x147: v = '\xd2'; break; |
---|
353 | case 0x150: v = '\xd5'; break; |
---|
354 | case 0x158: v = '\xd8'; break; |
---|
355 | case 0x16e: v = '\xd9'; break; |
---|
356 | case 0x170: v = '\xdb'; break; |
---|
357 | case 0x162: v = '\xde'; break; /* Ţ */ |
---|
358 | case 0x155: v = '\xe0'; break; |
---|
359 | case 0x103: v = '\xe3'; break; |
---|
360 | case 0x13a: v = '\xe5'; break; |
---|
361 | case 0x107: v = '\xe6'; break; |
---|
362 | case 0x10d: v = '\xe8'; break; |
---|
363 | case 0x119: v = '\xea'; break; |
---|
364 | case 0x11b: v = '\xec'; break; |
---|
365 | case 0x10f: v = '\xef'; break; |
---|
366 | case 0x111: v = '\xf0'; break; |
---|
367 | case 0x144: v = '\xf1'; break; |
---|
368 | case 0x148: v = '\xf2'; break; |
---|
369 | case 0x151: v = '\xf5'; break; |
---|
370 | case 0x159: v = '\xf8'; break; |
---|
371 | case 0x16f: v = '\xf9'; break; |
---|
372 | case 0x171: v = '\xfb'; break; |
---|
373 | case 0x163: v = '\xfe'; break; /* tcedil */ |
---|
374 | case 0x2d9: v = '\xff'; break; |
---|
375 | } |
---|
376 | if (v == 0) break; |
---|
377 | value = v; |
---|
378 | } |
---|
379 | *p = value; |
---|
380 | return 1; |
---|
381 | case CHARSET_ISO_8859_15: |
---|
382 | switch (value) { |
---|
383 | case 0xa4: case 0xa6: case 0xb0: case 0xc4: |
---|
384 | case 0xd0: case 0xd4: case 0xd5: case 0xd6: |
---|
385 | goto donthave; |
---|
386 | case 0x152: value = 0xd4; break; /* Œ */ |
---|
387 | case 0x153: value = 0xd5; break; /* œ */ |
---|
388 | #if 0 |
---|
389 | case 0x0: value = 0xa4; break; /* euro */ |
---|
390 | #endif |
---|
391 | case 0x160: value = 0xa6; break; /* Scaron */ |
---|
392 | case 0x161: value = 0xb0; break; /* scaron */ |
---|
393 | case 0x17d: value = 0xc4; break; /* Zcaron */ |
---|
394 | case 0x17e: value = 0xd0; break; /* zcaron */ |
---|
395 | #if 0 |
---|
396 | case 0x0: value = 0xd6; break; /* Ydiersis */ |
---|
397 | #endif |
---|
398 | } |
---|
399 | if (value < 0x100) { |
---|
400 | *p = value; |
---|
401 | return 1; |
---|
402 | } |
---|
403 | donthave: |
---|
404 | break; |
---|
405 | #if OS_WIN32 |
---|
406 | case CHARSET_WINCP1250: |
---|
407 | /* MS Windows rough equivalent to ISO-8859-2 */ |
---|
408 | if (value >= 0x80) { |
---|
409 | int v = 0; |
---|
410 | /* This mapping is complete - there are 5 unused positions: |
---|
411 | * 0x81 0x83 0x88 0x90 0x98 */ |
---|
412 | switch (value) { |
---|
413 | case 0xa0: case 0xa4: case 0xa6: case 0xa7: case 0xa8: case 0xa9: |
---|
414 | case 0xab: case 0xac: case 0xad: case 0xae: case 0xb0: case 0xb1: |
---|
415 | case 0xb4: case 0xb5: case 0xb6: case 0xb7: case 0xb8: case 0xbb: |
---|
416 | case 0xc1: case 0xc2: case 0xc4: case 0xc7: case 0xc9: case 0xcb: |
---|
417 | case 0xcd: case 0xce: case 0xd3: case 0xd4: case 0xd6: case 0xd7: |
---|
418 | case 0xda: case 0xdc: case 0xdd: case 0xdf: case 0xe1: case 0xe2: |
---|
419 | case 0xe4: case 0xe7: case 0xe9: case 0xeb: case 0xed: case 0xee: |
---|
420 | case 0xf3: case 0xf4: case 0xf6: case 0xf7: case 0xfa: case 0xfc: |
---|
421 | case 0xfd: |
---|
422 | v = value; break; |
---|
423 | case 0x20ac: v = '\x80'; break; |
---|
424 | case 0x201a: v = '\x82'; break; |
---|
425 | case 0x201e: v = '\x84'; break; |
---|
426 | case 0x2026: v = '\x85'; break; |
---|
427 | case 0x2020: v = '\x86'; break; |
---|
428 | case 0x2021: v = '\x87'; break; |
---|
429 | case 0x2030: v = '\x89'; break; |
---|
430 | case 0x0160: v = '\x8a'; break; |
---|
431 | case 0x2039: v = '\x8b'; break; |
---|
432 | case 0x015a: v = '\x8c'; break; |
---|
433 | case 0x0164: v = '\x8d'; break; |
---|
434 | case 0x017d: v = '\x8e'; break; |
---|
435 | case 0x0179: v = '\x8f'; break; |
---|
436 | case 0x2018: v = '\x91'; break; |
---|
437 | case 0x2019: v = '\x92'; break; |
---|
438 | case 0x201c: v = '\x93'; break; |
---|
439 | case 0x201d: v = '\x94'; break; |
---|
440 | case 0x2022: v = '\x95'; break; |
---|
441 | case 0x2013: v = '\x96'; break; |
---|
442 | case 0x2014: v = '\x97'; break; |
---|
443 | case 0x2122: v = '\x99'; break; |
---|
444 | case 0x0161: v = '\x9a'; break; |
---|
445 | case 0x203a: v = '\x9b'; break; |
---|
446 | case 0x015b: v = '\x9c'; break; |
---|
447 | case 0x0165: v = '\x9d'; break; |
---|
448 | case 0x017e: v = '\x9e'; break; |
---|
449 | case 0x017a: v = '\x9f'; break; |
---|
450 | case 0x02c7: v = '\xa1'; break; |
---|
451 | case 0x02d8: v = '\xa2'; break; |
---|
452 | case 0x0141: v = '\xa3'; break; |
---|
453 | case 0x0104: v = '\xa5'; break; |
---|
454 | case 0x015e: v = '\xaa'; break; /* Scedil */ |
---|
455 | case 0x017b: v = '\xaf'; break; |
---|
456 | case 0x02db: v = '\xb2'; break; |
---|
457 | case 0x0142: v = '\xb3'; break; |
---|
458 | case 0x0105: v = '\xb9'; break; |
---|
459 | case 0x015f: v = '\xba'; break; /* scedil */ |
---|
460 | case 0x013d: v = '\xbc'; break; |
---|
461 | case 0x02dd: v = '\xbd'; break; |
---|
462 | case 0x013e: v = '\xbe'; break; |
---|
463 | case 0x017c: v = '\xbf'; break; |
---|
464 | case 0x0154: v = '\xc0'; break; |
---|
465 | case 0x0102: v = '\xc3'; break; |
---|
466 | case 0x0139: v = '\xc5'; break; |
---|
467 | case 0x0106: v = '\xc6'; break; |
---|
468 | case 0x010c: v = '\xc8'; break; |
---|
469 | case 0x0118: v = '\xca'; break; |
---|
470 | case 0x011a: v = '\xcc'; break; |
---|
471 | case 0x010e: v = '\xcf'; break; |
---|
472 | case 0x0110: v = '\xd0'; break; |
---|
473 | case 0x0143: v = '\xd1'; break; |
---|
474 | case 0x0147: v = '\xd2'; break; |
---|
475 | case 0x0150: v = '\xd5'; break; |
---|
476 | case 0x0158: v = '\xd8'; break; |
---|
477 | case 0x016e: v = '\xd9'; break; |
---|
478 | case 0x0170: v = '\xdb'; break; |
---|
479 | case 0x0162: v = '\xde'; break; /* Ţ */ |
---|
480 | case 0x0155: v = '\xe0'; break; |
---|
481 | case 0x0103: v = '\xe3'; break; |
---|
482 | case 0x013a: v = '\xe5'; break; |
---|
483 | case 0x0107: v = '\xe6'; break; |
---|
484 | case 0x010d: v = '\xe8'; break; |
---|
485 | case 0x0119: v = '\xea'; break; |
---|
486 | case 0x011b: v = '\xec'; break; |
---|
487 | case 0x010f: v = '\xef'; break; |
---|
488 | case 0x0111: v = '\xf0'; break; |
---|
489 | case 0x0144: v = '\xf1'; break; |
---|
490 | case 0x0148: v = '\xf2'; break; |
---|
491 | case 0x0151: v = '\xf5'; break; |
---|
492 | case 0x0159: v = '\xf8'; break; |
---|
493 | case 0x016f: v = '\xf9'; break; |
---|
494 | case 0x0171: v = '\xfb'; break; |
---|
495 | case 0x0163: v = '\xfe'; break; /* tcedil */ |
---|
496 | case 0x02d9: v = '\xff'; break; |
---|
497 | } |
---|
498 | if (v == 0) break; |
---|
499 | value = v; |
---|
500 | } |
---|
501 | *p = value; |
---|
502 | return 1; |
---|
503 | case CHARSET_WINCP1252: |
---|
504 | /* MS Windows extensions to ISO-8859-1 */ |
---|
505 | /* This mapping is complete - there are 5 unused positions: |
---|
506 | * 0x81 0x8d 0x8f 0x90 0x9d */ |
---|
507 | switch (value) { |
---|
508 | case 0x2026: value = 0x85; break; /* hellip */ |
---|
509 | case 0x0160: value = 0x8a; break; /* Scaron */ |
---|
510 | case 0x0152: value = 0x8c; break; /* OElig */ |
---|
511 | case 0x017d: value = 0x8e; break; /* Zcaron */ |
---|
512 | case 0x2019: value = 0x92; break; /* rsquo */ |
---|
513 | case 0x201c: value = 0x93; break; /* ldquo */ |
---|
514 | case 0x201d: value = 0x94; break; /* rdquo */ |
---|
515 | case 0x0161: value = 0x9a; break; /* scaron */ |
---|
516 | case 0x0153: value = 0x9c; break; /* oelig */ |
---|
517 | case 0x017e: value = 0x9e; break; /* zcaron */ |
---|
518 | #if 0 |
---|
519 | /* there are a few other obscure ones we don't currently need */ |
---|
520 | case 0x20ac: value = 0x80; break; /* euro */ |
---|
521 | case 0x201a: value = 0x82; break; /* sbquo */ |
---|
522 | case 0x0192: value = 0x83; break; /* fnof */ |
---|
523 | case 0x201e: value = 0x84; break; /* bdquo */ |
---|
524 | case 0x2020: value = 0x86; break; /* dagger */ |
---|
525 | case 0x2021: value = 0x87; break; /* Dagger */ |
---|
526 | case 0x02c6: value = 0x88; break; /* circ */ |
---|
527 | case 0x2030: value = 0x89; break; /* permil */ |
---|
528 | case 0x2039: value = 0x8b; break; /* lsaquo */ |
---|
529 | case 0x2018: value = 0x91; break; /* lsquo */ |
---|
530 | case 0x2022: value = 0x95; break; /* bull */ |
---|
531 | case 0x2013: value = 0x96; break; /* ndash */ |
---|
532 | case 0x2014: value = 0x97; break; /* mdash */ |
---|
533 | case 0x02dc: value = 0x98; break; /* tilde */ |
---|
534 | case 0x2122: value = 0x99; break; /* trade */ |
---|
535 | case 0x203a: value = 0x9b; break; /* rsaquo */ |
---|
536 | case 0x0178: value = 0x9f; break; /* Yuml */ |
---|
537 | #endif |
---|
538 | } |
---|
539 | if (value < 0x100) { |
---|
540 | *p = value; |
---|
541 | return 1; |
---|
542 | } |
---|
543 | break; |
---|
544 | #endif |
---|
545 | #if OS_WIN32 |
---|
546 | case CHARSET_DOSCP850: { |
---|
547 | static const unsigned char uni2dostab[] = { |
---|
548 | 255, 173, 189, 156, 207, 190, 221, 245, |
---|
549 | 249, 184, 166, 174, 170, 240, 169, 238, |
---|
550 | 248, 241, 253, 252, 239, 230, 244, 250, |
---|
551 | 247, 251, 167, 175, 172, 171, 243, 168, |
---|
552 | 183, 181, 182, 199, 142, 143, 146, 128, |
---|
553 | 212, 144, 210, 211, 222, 214, 215, 216, |
---|
554 | 209, 165, 227, 224, 226, 229, 153, 158, |
---|
555 | 157, 235, 233, 234, 154, 237, 232, 225, |
---|
556 | 133, 160, 131, 198, 132, 134, 145, 135, |
---|
557 | 138, 130, 136, 137, 141, 161, 140, 139, |
---|
558 | 208, 164, 149, 162, 147, 228, 148, 246, |
---|
559 | 155, 151, 163, 150, 129, 236, 231, 152 |
---|
560 | }; |
---|
561 | if (value >= 160 && value < 256) { |
---|
562 | *p = (int)uni2dostab[value - 160]; |
---|
563 | return 1; |
---|
564 | } |
---|
565 | #if 0 |
---|
566 | if (value == 305) { /* LATIN SMALL LETTER DOTLESS I */ |
---|
567 | *p = 213; /* "Modified CP850" has the Euro sign here. */ |
---|
568 | return 1; |
---|
569 | } |
---|
570 | if (value == 402) { /* LATIN SMALL LETTER F WITH HOOK */ |
---|
571 | *p = 159; |
---|
572 | return 1; |
---|
573 | } |
---|
574 | #endif |
---|
575 | break; |
---|
576 | } |
---|
577 | #endif |
---|
578 | } |
---|
579 | /* Transliterate characters we can't represent */ |
---|
580 | #ifdef DEBUG |
---|
581 | fprintf(stderr, "transliterate “%c” 0x%x\n", value, value); |
---|
582 | #endif |
---|
583 | switch (value) { |
---|
584 | case 160: |
---|
585 | *p = ' '; return 1; |
---|
586 | case 161 /* ¡ */: |
---|
587 | *p = '!'; return 1; |
---|
588 | case 176 /* ° */: |
---|
589 | *p = 'd'; p[1] = 'g'; return 2; |
---|
590 | case 191 /* ¿ */: |
---|
591 | *p = '?'; return 1; |
---|
592 | case 192 /* À */: case 193 /* Á */: case 194 /* Â */: case 195 /* Ã */: |
---|
593 | *p = 'A'; return 1; |
---|
594 | case 197 /* Å */: |
---|
595 | p[1] = *p = 'A'; return 2; |
---|
596 | case 196 /* Ä */: /* Ä */ |
---|
597 | *p = 'A'; |
---|
598 | if (!umlaut_to_e()) return 1; |
---|
599 | p[1] = 'E'; return 2; |
---|
600 | case 198 /* Æ */: |
---|
601 | *p = 'A'; p[1] = 'E'; return 2; |
---|
602 | case 199 /* Ç */: case 268: /* Č */ |
---|
603 | *p = 'C'; return 1; |
---|
604 | case 270: /* Ď */ |
---|
605 | *p = 'D'; return 1; |
---|
606 | case 200 /* È */: case 201 /* É */: case 202 /* Ê */: case 203 /* Ë */: |
---|
607 | *p = 'E'; return 1; |
---|
608 | case 204 /* Ì */: case 205 /* Í */: case 206 /* Î */: case 207 /* Ï */: |
---|
609 | *p = 'I'; return 1; |
---|
610 | case 208 /* Ð */: case 222 /* Þ */: |
---|
611 | *p = 'T'; p[1] = 'H'; return 2; |
---|
612 | case 315: /* Ĺ */ |
---|
613 | case 317: /* Ľ */ |
---|
614 | *p = 'L'; return 1; |
---|
615 | case 209 /* Ñ */: |
---|
616 | *p = 'N'; return 1; |
---|
617 | case 210 /* Ò */: case 211 /* Ó */: case 212 /* Ô */: case 213 /* Õ */: |
---|
618 | *p = 'O'; return 1; |
---|
619 | case 214 /* Ö */: /* Ö */ case 0x152: /* Œ */ |
---|
620 | *p = 'O'; p[1] = 'E'; return 2; |
---|
621 | case 352: /* Š */ |
---|
622 | case 0x15e: /* Ş */ |
---|
623 | *p = 'S'; return 1; |
---|
624 | case 0x162: /* Ţ */ |
---|
625 | case 0x164: /* Ť */ |
---|
626 | *p = 'T'; return 1; |
---|
627 | case 217 /* Ù */: case 218 /* Ú */: case 219 /* Û */: |
---|
628 | *p = 'U'; return 1; |
---|
629 | case 220 /* Ü */: /* Ü */ |
---|
630 | *p = 'U'; p[1] = 'E'; return 2; |
---|
631 | case 221 /* Ý */: |
---|
632 | *p = 'Y'; return 1; |
---|
633 | case 381: /* Ž */ |
---|
634 | *p = 'Z'; return 1; |
---|
635 | case 223 /* ß */: |
---|
636 | p[1] = *p = 's'; return 2; |
---|
637 | case 224 /* à */: case 225 /* á */: case 226 /* â */: case 227 /* ã */: |
---|
638 | case 259: /* ă */ |
---|
639 | *p = 'a'; return 1; |
---|
640 | case 228 /* ä */: /* ä */ case 230 /* æ */: |
---|
641 | *p = 'a'; p[1] = 'e'; return 2; |
---|
642 | case 229 /* å */: |
---|
643 | p[1] = *p = 'a'; return 2; |
---|
644 | case 231 /* ç */: case 269 /* č */: |
---|
645 | *p = 'c'; return 1; |
---|
646 | case 271: /* ď */ |
---|
647 | *p = 'd'; return 1; |
---|
648 | case 232 /* è */: case 233 /* é */: case 234 /* ê */: case 235 /* ë */: |
---|
649 | case 283 /* ě */: |
---|
650 | *p = 'e'; return 1; |
---|
651 | case 236 /* ì */: case 237 /* í */: case 238 /* î */: case 239 /* ï */: |
---|
652 | *p = 'i'; return 1; |
---|
653 | case 316 /* ĺ */: |
---|
654 | case 318 /* ľ */: |
---|
655 | *p = 'l'; return 1; |
---|
656 | case 241 /* ñ */: case 328 /* ň */: |
---|
657 | *p = 'n'; return 1; |
---|
658 | case 345: /* ř */ |
---|
659 | *p = 'r'; return 1; |
---|
660 | case 353: /* š */ |
---|
661 | case 0x15f: /* ş */ |
---|
662 | *p = 's'; return 1; |
---|
663 | case 357: /* ť */ |
---|
664 | case 0x163: /* ţ */ |
---|
665 | *p = 't'; return 1; |
---|
666 | case 240 /* ð */: case 254 /* þ */: |
---|
667 | *p = 't'; p[1] = 'h'; return 2; |
---|
668 | case 242 /* ò */: case 243 /* ó */: case 244 /* ô */: case 245 /* õ */: |
---|
669 | *p = 'o'; return 1; |
---|
670 | case 246 /* ö */: /* ö */ case 0x153: /* œ */ |
---|
671 | *p = 'o'; p[1] = 'e'; return 2; |
---|
672 | case 249 /* ù */: case 250 /* ú */: case 251 /* û */: |
---|
673 | case 367 /* ů */: |
---|
674 | *p = 'u'; return 1; |
---|
675 | case 252 /* ü */: /* ü */ |
---|
676 | *p = 'u'; p[1] = 'e'; return 2; |
---|
677 | case 253 /* ý */: case 255 /* ÿ */: |
---|
678 | *p = 'y'; return 1; |
---|
679 | case 382: /* ž */ |
---|
680 | *p = 'z'; return 1; |
---|
681 | case 0x2019: /* ‘ */ |
---|
682 | *p = '\''; return 1; |
---|
683 | case 171: /* « */ case 187: /* » */ |
---|
684 | case 0x201c: /* “ */ case 0x201d: /* ” */ |
---|
685 | *p = '"'; return 1; |
---|
686 | case 0x2026: /* … */ |
---|
687 | *p = '.'; p[1] = '.'; p[2] = '.'; return 3; |
---|
688 | case 0x2192: /* → */ |
---|
689 | *p = '-'; p[1] = '>'; return 2; |
---|
690 | case 0x1d4d: /* gradient symbol */ |
---|
691 | *p = 'g'; p[1] = 'r'; p[2] = 'd'; return 3; |
---|
692 | case 0x221e: /* infinity symbol */ |
---|
693 | *p = 'i'; p[1] = 'n'; p[2] = 'f'; return 3; |
---|
694 | } |
---|
695 | #ifdef DEBUG |
---|
696 | /* 169 is reported (copyright symbol), but there isn't a good <= 2 ASCII |
---|
697 | * character transliteration for that, so we handle that elsewhere. */ |
---|
698 | fprintf(stderr, "failed to transliterate codepoint %d\n", value); |
---|
699 | #endif |
---|
700 | return 0; |
---|
701 | } |
---|
702 | |
---|
703 | #if OS_UNIX && defined DATADIR && defined PACKAGE |
---|
704 | /* Under Unix, we compile in the configured path */ |
---|
705 | static const char *pth_cfg_files = DATADIR "/" PACKAGE; |
---|
706 | #else |
---|
707 | /* On other platforms, we fall back on looking in the current directory */ |
---|
708 | static const char *pth_cfg_files = ""; |
---|
709 | #endif |
---|
710 | |
---|
711 | static int num_msgs = 0; |
---|
712 | static char **msg_array = NULL; |
---|
713 | |
---|
714 | static bool msg_lang_explicit = false; |
---|
715 | const char *msg_lang = NULL; |
---|
716 | const char *msg_lang2 = NULL; |
---|
717 | |
---|
718 | static char ** |
---|
719 | parse_msgs(int n, unsigned char *p, int charset_code) { |
---|
720 | int i; |
---|
721 | |
---|
722 | char **msgs = osmalloc(n * sizeof(char *)); |
---|
723 | |
---|
724 | for (i = 0; i < n; i++) { |
---|
725 | unsigned char *to = p; |
---|
726 | int ch; |
---|
727 | msgs[i] = (char *)p; |
---|
728 | |
---|
729 | /* If we want UTF8 anyway, we just need to find the start of each |
---|
730 | * message */ |
---|
731 | if (charset_code == CHARSET_UTF8) { |
---|
732 | p += strlen((char *)p) + 1; |
---|
733 | continue; |
---|
734 | } |
---|
735 | |
---|
736 | while ((ch = *p++) != 0) { |
---|
737 | /* A byte in the range 0x80-0xbf or 0xf0-0xff isn't valid in |
---|
738 | * this state, (0xf0-0xfd mean values > 0xffff) so treat as |
---|
739 | * literal and try to resync so we cope better when fed |
---|
740 | * non-utf-8 data. Similarly we abandon a multibyte sequence |
---|
741 | * if we hit an invalid character. */ |
---|
742 | if (ch >= 0xc0 && ch < 0xf0) { |
---|
743 | int ch1 = *p; |
---|
744 | if ((ch1 & 0xc0) != 0x80) goto resync; |
---|
745 | |
---|
746 | if (ch < 0xe0) { |
---|
747 | /* 2 byte sequence */ |
---|
748 | ch = ((ch & 0x1f) << 6) | (ch1 & 0x3f); |
---|
749 | p++; |
---|
750 | } else { |
---|
751 | /* 3 byte sequence */ |
---|
752 | int ch2 = p[1]; |
---|
753 | if ((ch2 & 0xc0) != 0x80) goto resync; |
---|
754 | ch = ((ch & 0x1f) << 12) | ((ch1 & 0x3f) << 6) | (ch2 & 0x3f); |
---|
755 | p += 2; |
---|
756 | } |
---|
757 | } |
---|
758 | |
---|
759 | resync: |
---|
760 | |
---|
761 | if (ch < 127) { |
---|
762 | *to++ = (char)ch; |
---|
763 | } else { |
---|
764 | /* We assume an N byte UTF-8 code never transliterates to more |
---|
765 | * than N characters (so we can't transliterate © to (C) or |
---|
766 | * ® to (R) for example) */ |
---|
767 | to += add_unicode(charset_code, to, ch); |
---|
768 | } |
---|
769 | } |
---|
770 | *to++ = '\0'; |
---|
771 | } |
---|
772 | return msgs; |
---|
773 | } |
---|
774 | |
---|
775 | /* This is the name of the default language, which can be set like so: |
---|
776 | * ./configure --enable-defaultlang=fr |
---|
777 | */ |
---|
778 | #ifdef DEFAULTLANG |
---|
779 | /* No point extracting these errors as they won't get used if file opens */ |
---|
780 | # include "../lib/defaultlang.h" |
---|
781 | #else |
---|
782 | #define N_DONTEXTRACTMSGS 5 |
---|
783 | static unsigned char dontextractmsgs[] = |
---|
784 | "Can't open message file \"%s\" using path \"%s\"\0"/*1000*/ |
---|
785 | "Problem with message file \"%s\"\0"/*1001*/ |
---|
786 | "I don't understand this message file version\0"/*1002*/ |
---|
787 | "Message file truncated?\0"/*1003*/ |
---|
788 | "Out of memory (couldn't find %lu bytes).\0"/*1004*/; |
---|
789 | #endif |
---|
790 | |
---|
791 | static char **dontextract = NULL; |
---|
792 | |
---|
793 | static void |
---|
794 | parse_msg_file(int charset_code) |
---|
795 | { |
---|
796 | FILE *fh; |
---|
797 | unsigned char header[20]; |
---|
798 | int i; |
---|
799 | unsigned len; |
---|
800 | unsigned char *p; |
---|
801 | char *fnm, *s; |
---|
802 | int n; |
---|
803 | |
---|
804 | #ifdef DEBUG |
---|
805 | fprintf(stderr, "parse_msg_file(%d)\n", charset_code); |
---|
806 | #endif |
---|
807 | |
---|
808 | /* sort out messages we need to print if we can't open the message file */ |
---|
809 | dontextract = parse_msgs(N_DONTEXTRACTMSGS, dontextractmsgs, charset_code); |
---|
810 | |
---|
811 | fnm = osstrdup(msg_lang); |
---|
812 | /* trim off charset from stuff like "de_DE.iso8859_1" */ |
---|
813 | s = strchr(fnm, '.'); |
---|
814 | if (s) *s = '\0'; |
---|
815 | /* trim off any "@<something>" modifier. */ |
---|
816 | s = strchr(fnm, '@'); |
---|
817 | if (s) *s = '\0'; |
---|
818 | |
---|
819 | fh = fopenWithPthAndExt(pth_cfg_files, fnm, EXT_SVX_MSG, "rb", NULL); |
---|
820 | |
---|
821 | if (!fh && strlen(fnm) > 3 && fnm[2] == '_') { |
---|
822 | /* e.g. if 'en_GB' is unknown, see if we know 'en' */ |
---|
823 | fnm[2] = '\0'; |
---|
824 | fh = fopenWithPthAndExt(pth_cfg_files, fnm, EXT_SVX_MSG, "rb", NULL); |
---|
825 | if (!fh) fnm[2] = '_'; /* for error reporting */ |
---|
826 | } |
---|
827 | |
---|
828 | if (!fh && !msg_lang_explicit) { |
---|
829 | /* If msg_lang wasn't specified using environment variable SURVEXLANG, |
---|
830 | * then default to 'en' if we don't find messages for language msg_lang. |
---|
831 | */ |
---|
832 | if (fnm[0] && fnm[1]) { |
---|
833 | strcpy(fnm, "en"); |
---|
834 | } else { |
---|
835 | osfree(fnm); |
---|
836 | fnm = osstrdup("en"); |
---|
837 | } |
---|
838 | fh = fopenWithPthAndExt(pth_cfg_files, fnm, EXT_SVX_MSG, "rb", NULL); |
---|
839 | } |
---|
840 | |
---|
841 | if (!fh) { |
---|
842 | fatalerror(/*Can't open message file “%s” using path “%s”*/1000, |
---|
843 | fnm, pth_cfg_files); |
---|
844 | } |
---|
845 | |
---|
846 | if (fread(header, 1, 20, fh) < 20 || |
---|
847 | memcmp(header, "Svx\nMsg\r\n\xfe\xff", 12) != 0) { |
---|
848 | fatalerror(/*Problem with message file “%s”*/1001, fnm); |
---|
849 | } |
---|
850 | |
---|
851 | if (header[12] != 0) |
---|
852 | fatalerror(/*I don't understand this message file version*/1002); |
---|
853 | |
---|
854 | n = (header[14] << 8) | header[15]; |
---|
855 | |
---|
856 | len = 0; |
---|
857 | for (i = 16; i < 20; i++) len = (len << 8) | header[i]; |
---|
858 | |
---|
859 | p = osmalloc(len); |
---|
860 | if (fread(p, 1, len, fh) < len) |
---|
861 | fatalerror(/*Message file truncated?*/1003); |
---|
862 | |
---|
863 | fclose(fh); |
---|
864 | |
---|
865 | #ifdef DEBUG |
---|
866 | fprintf(stderr, "fnm = “%s”, n = %d, len = %d\n", fnm, n, len); |
---|
867 | #endif |
---|
868 | osfree(fnm); |
---|
869 | |
---|
870 | msg_array = parse_msgs(n, p, charset_code); |
---|
871 | num_msgs = n; |
---|
872 | } |
---|
873 | |
---|
874 | const char * |
---|
875 | msg_cfgpth(void) |
---|
876 | { |
---|
877 | return pth_cfg_files; |
---|
878 | } |
---|
879 | |
---|
880 | const char * |
---|
881 | msg_exepth(void) |
---|
882 | { |
---|
883 | return exe_pth; |
---|
884 | } |
---|
885 | |
---|
886 | const char * |
---|
887 | msg_appname(void) |
---|
888 | { |
---|
889 | return appname_copy; |
---|
890 | } |
---|
891 | |
---|
892 | void |
---|
893 | (msg_init)(char * const *argv) |
---|
894 | { |
---|
895 | char *p; |
---|
896 | SVX_ASSERT(argv); |
---|
897 | |
---|
898 | /* Point to argv[0] itself so we report a more helpful error if the |
---|
899 | * code to work out the clean appname generates a signal */ |
---|
900 | appname_copy = argv[0]; |
---|
901 | #if OS_UNIX |
---|
902 | /* use name as-is on Unix - programs run from path get name as supplied */ |
---|
903 | appname_copy = osstrdup(argv[0]); |
---|
904 | #else |
---|
905 | /* use the lower-cased leafname on other platforms */ |
---|
906 | p = leaf_from_fnm(argv[0]); |
---|
907 | appname_copy = p; |
---|
908 | while (*p) { |
---|
909 | *p = tolower((unsigned char)*p); |
---|
910 | ++p; |
---|
911 | } |
---|
912 | #endif |
---|
913 | |
---|
914 | /* shortcut --version so you can check the version number even when the |
---|
915 | * correct message file can't be found... */ |
---|
916 | if (argv[1] && strcmp(argv[1], "--version") == 0) { |
---|
917 | cmdline_version(); |
---|
918 | exit(0); |
---|
919 | } |
---|
920 | if (argv[0]) { |
---|
921 | exe_pth = path_from_fnm(argv[0]); |
---|
922 | #if OS_UNIX && defined DATADIR && defined PACKAGE |
---|
923 | bool free_pth = false; |
---|
924 | char *pth = getenv("srcdir"); |
---|
925 | if (!pth || !pth[0]) { |
---|
926 | pth = path_from_fnm(argv[0]); |
---|
927 | free_pth = true; |
---|
928 | } |
---|
929 | if (pth[0]) { |
---|
930 | struct stat buf; |
---|
931 | #if OS_UNIX_MACOS |
---|
932 | # ifndef AVEN |
---|
933 | /* On macOS the programs may be installed anywhere, with the |
---|
934 | * share directory and the binaries in the same directory. */ |
---|
935 | p = use_path(pth, "share/survex/en.msg"); |
---|
936 | if (stat(p, &buf) == 0 && S_ISREG(buf.st_mode)) { |
---|
937 | pth_cfg_files = use_path(pth, "share/survex"); |
---|
938 | goto macos_got_msg; |
---|
939 | } |
---|
940 | osfree(p); |
---|
941 | # endif |
---|
942 | /* In the diskimage package, this case is used for aven, and for |
---|
943 | * the hardlinked copies of cavern and extend alongside the aven |
---|
944 | * binary, which are the ones which aven runs. |
---|
945 | */ |
---|
946 | p = use_path(pth, "../Resources/en.msg"); |
---|
947 | if (stat(p, &buf) == 0 && S_ISREG(buf.st_mode)) { |
---|
948 | pth_cfg_files = use_path(pth, "../Resources"); |
---|
949 | goto macos_got_msg; |
---|
950 | } |
---|
951 | osfree(p); |
---|
952 | #endif |
---|
953 | /* If we're run with an explicit path, check if "../lib/en.msg" |
---|
954 | * from the program's path exists, and if so look there for |
---|
955 | * support files - this allows us to test binaries in the build |
---|
956 | * tree easily. */ |
---|
957 | p = use_path(pth, "../lib/en.msg"); |
---|
958 | if (stat(p, &buf) == 0) { |
---|
959 | #ifdef S_ISREG |
---|
960 | /* POSIX way */ |
---|
961 | if (S_ISREG(buf.st_mode)) { |
---|
962 | pth_cfg_files = use_path(pth, "../lib"); |
---|
963 | } |
---|
964 | #else |
---|
965 | /* BSD way */ |
---|
966 | if ((buf.st_mode & S_IFMT) == S_IFREG) { |
---|
967 | pth_cfg_files = use_path(pth, "../lib"); |
---|
968 | } |
---|
969 | #endif |
---|
970 | } |
---|
971 | #if defined(__GNUC__) && defined(__APPLE_CC__) |
---|
972 | macos_got_msg: |
---|
973 | #endif |
---|
974 | osfree(p); |
---|
975 | } |
---|
976 | |
---|
977 | if (free_pth) osfree(pth); |
---|
978 | #elif OS_WIN32 |
---|
979 | DWORD len = 256; |
---|
980 | char *buf = NULL, *modname; |
---|
981 | while (1) { |
---|
982 | DWORD got; |
---|
983 | buf = osrealloc(buf, len); |
---|
984 | got = GetModuleFileName(NULL, buf, len); |
---|
985 | if (got < len) break; |
---|
986 | len += len; |
---|
987 | } |
---|
988 | modname = buf; |
---|
989 | /* Strange Win32 nastiness - strip prefix "\\?\" if present */ |
---|
990 | if (strncmp(modname, "\\\\?\\", 4) == 0) modname += 4; |
---|
991 | pth_cfg_files = path_from_fnm(modname); |
---|
992 | osfree(buf); |
---|
993 | #else |
---|
994 | /* Get the path to the support files from argv[0] */ |
---|
995 | pth_cfg_files = path_from_fnm(argv[0]); |
---|
996 | #endif |
---|
997 | } |
---|
998 | |
---|
999 | msg_lang = getenv("SURVEXLANG"); |
---|
1000 | #ifdef DEBUG |
---|
1001 | fprintf(stderr, "msg_lang = %p (= \"%s\")\n", msg_lang, msg_lang?msg_lang:"(null)"); |
---|
1002 | #endif |
---|
1003 | |
---|
1004 | msg_lang_explicit = true; |
---|
1005 | if (!msg_lang || !*msg_lang) { |
---|
1006 | msg_lang_explicit = false; |
---|
1007 | msg_lang = getenv("LC_ALL"); |
---|
1008 | } |
---|
1009 | if (!msg_lang || !*msg_lang) { |
---|
1010 | msg_lang = getenv("LC_MESSAGES"); |
---|
1011 | if (!msg_lang || !*msg_lang) { |
---|
1012 | msg_lang = getenv("LANG"); |
---|
1013 | /* Something (AutoCAD?) on Microsoft Windows sets LANG to a number. */ |
---|
1014 | if (msg_lang && !isalpha(msg_lang[0])) msg_lang = NULL; |
---|
1015 | } |
---|
1016 | if (!msg_lang || !*msg_lang) { |
---|
1017 | #if OS_WIN32 |
---|
1018 | LCID locid; |
---|
1019 | #endif |
---|
1020 | #ifdef DEFAULTLANG |
---|
1021 | msg_lang = STRING(DEFAULTLANG); |
---|
1022 | #else |
---|
1023 | msg_lang = "en"; |
---|
1024 | #endif |
---|
1025 | #if OS_WIN32 |
---|
1026 | /* GetUserDefaultUILanguage() requires Microsoft Windows 2000 or |
---|
1027 | * newer, but we don't support anything earlier than Vista. |
---|
1028 | */ |
---|
1029 | locid = GetUserDefaultUILanguage(); |
---|
1030 | if (locid) { |
---|
1031 | WORD langid = LANGIDFROMLCID(locid); |
---|
1032 | switch (PRIMARYLANGID(langid)) { |
---|
1033 | case LANG_BULGARIAN: |
---|
1034 | msg_lang = "bg"; |
---|
1035 | break; |
---|
1036 | /* older mingw compilers don't seem to supply this value */ |
---|
1037 | #ifndef LANG_CATALAN |
---|
1038 | # define LANG_CATALAN 0x03 |
---|
1039 | #endif |
---|
1040 | case LANG_CATALAN: |
---|
1041 | msg_lang = "ca"; |
---|
1042 | break; |
---|
1043 | case LANG_CHINESE: |
---|
1044 | msg_lang = "zh_CN"; |
---|
1045 | break; |
---|
1046 | case LANG_ENGLISH: |
---|
1047 | if (SUBLANGID(langid) == SUBLANG_ENGLISH_US) |
---|
1048 | msg_lang = "en_US"; |
---|
1049 | else |
---|
1050 | msg_lang = "en"; |
---|
1051 | break; |
---|
1052 | case LANG_FRENCH: |
---|
1053 | msg_lang = "fr"; |
---|
1054 | break; |
---|
1055 | case LANG_GERMAN: |
---|
1056 | switch (SUBLANGID(langid)) { |
---|
1057 | case SUBLANG_GERMAN_SWISS: |
---|
1058 | msg_lang = "de_CH"; |
---|
1059 | break; |
---|
1060 | default: |
---|
1061 | msg_lang = "de"; |
---|
1062 | } |
---|
1063 | break; |
---|
1064 | case LANG_GREEK: |
---|
1065 | msg_lang = "el"; |
---|
1066 | break; |
---|
1067 | case LANG_HUNGARIAN: |
---|
1068 | msg_lang = "hu"; |
---|
1069 | break; |
---|
1070 | case LANG_INDONESIAN: |
---|
1071 | msg_lang = "id"; |
---|
1072 | break; |
---|
1073 | case LANG_ITALIAN: |
---|
1074 | msg_lang = "it"; |
---|
1075 | break; |
---|
1076 | case LANG_POLISH: |
---|
1077 | msg_lang = "pl"; |
---|
1078 | break; |
---|
1079 | case LANG_PORTUGUESE: |
---|
1080 | if (SUBLANGID(langid) == SUBLANG_PORTUGUESE_BRAZILIAN) |
---|
1081 | msg_lang = "pt_BR"; |
---|
1082 | else |
---|
1083 | msg_lang = "pt"; |
---|
1084 | break; |
---|
1085 | case LANG_ROMANIAN: |
---|
1086 | msg_lang = "ro"; |
---|
1087 | break; |
---|
1088 | case LANG_RUSSIAN: |
---|
1089 | msg_lang = "ru"; |
---|
1090 | break; |
---|
1091 | case LANG_SLOVAK: |
---|
1092 | msg_lang = "sk"; |
---|
1093 | break; |
---|
1094 | case LANG_SPANISH: |
---|
1095 | msg_lang = "es"; |
---|
1096 | break; |
---|
1097 | } |
---|
1098 | } |
---|
1099 | #endif |
---|
1100 | } |
---|
1101 | } |
---|
1102 | #ifdef DEBUG |
---|
1103 | fprintf(stderr, "msg_lang = %p (= \"%s\")\n", msg_lang, msg_lang?msg_lang:"(null)"); |
---|
1104 | #endif |
---|
1105 | |
---|
1106 | /* On Mandrake LANG defaults to C */ |
---|
1107 | if (strcmp(msg_lang, "C") == 0) msg_lang = "en"; |
---|
1108 | |
---|
1109 | { /* If msg_lang has a country code, snip it out to give msg_lang2. */ |
---|
1110 | size_t b = 0; |
---|
1111 | while (isalpha((unsigned char)msg_lang[b])) { |
---|
1112 | ++b; |
---|
1113 | } |
---|
1114 | if (msg_lang[b] == '_') { |
---|
1115 | char * tmp; |
---|
1116 | size_t e = b + 1; |
---|
1117 | while (isalpha((unsigned char)msg_lang[e])) { |
---|
1118 | ++e; |
---|
1119 | } |
---|
1120 | tmp = osstrdup(msg_lang); |
---|
1121 | memmove(tmp + b, tmp + e, strlen(tmp + e) + 1); |
---|
1122 | msg_lang2 = tmp; |
---|
1123 | } |
---|
1124 | } |
---|
1125 | |
---|
1126 | #ifdef LC_MESSAGES |
---|
1127 | /* try to setlocale() appropriately too */ |
---|
1128 | if (!setlocale(LC_MESSAGES, msg_lang)) { |
---|
1129 | if (msg_lang2) { |
---|
1130 | (void)setlocale(LC_MESSAGES, msg_lang2); |
---|
1131 | } |
---|
1132 | } |
---|
1133 | #endif |
---|
1134 | |
---|
1135 | select_charset(default_charset()); |
---|
1136 | } |
---|
1137 | |
---|
1138 | #ifndef AVEN |
---|
1139 | /* Return message if messages available, else a fallback value. */ |
---|
1140 | static const char * |
---|
1141 | msg_opt(int en, const char * fallback) |
---|
1142 | { |
---|
1143 | /* NB can't use SVX_ASSERT here! */ |
---|
1144 | if (!msg_array || en <= 0 || en >= num_msgs) { |
---|
1145 | return fallback; |
---|
1146 | } |
---|
1147 | |
---|
1148 | return msg_array[en]; |
---|
1149 | } |
---|
1150 | #endif |
---|
1151 | |
---|
1152 | const char * |
---|
1153 | msg(int en) |
---|
1154 | { |
---|
1155 | /* NB can't use SVX_ASSERT here! */ |
---|
1156 | if (dontextract && en >= 1000 && en < 1000 + N_DONTEXTRACTMSGS) |
---|
1157 | return dontextract[en - 1000]; |
---|
1158 | if (!msg_array) { |
---|
1159 | if (en != 1) { |
---|
1160 | fprintf(STDERR, "Message %d requested before fully initialised\n", en); |
---|
1161 | return "Message requested before fully initialised\n"; |
---|
1162 | } |
---|
1163 | /* this should be the only other message which can be requested before |
---|
1164 | * the message file is opened and read... */ |
---|
1165 | if (!dontextract) return "Out of memory (couldn't find %lu bytes)."; |
---|
1166 | return dontextract[(/*Out of memory (couldn't find %lu bytes).*/1004) |
---|
1167 | - 1000]; |
---|
1168 | } |
---|
1169 | |
---|
1170 | if (en < 0 || en >= num_msgs) { |
---|
1171 | fprintf(STDERR, "Message %d out of range\n", en); |
---|
1172 | return "Message out of range\n"; |
---|
1173 | } |
---|
1174 | |
---|
1175 | if (en == 0) { |
---|
1176 | const char *p = msg_array[0]; |
---|
1177 | if (!*p) p = "(C)"; |
---|
1178 | return p; |
---|
1179 | } |
---|
1180 | |
---|
1181 | return msg_array[en]; |
---|
1182 | } |
---|
1183 | |
---|
1184 | void |
---|
1185 | v_report(int severity, const char *fnm, int line, int col, int en, va_list ap) |
---|
1186 | { |
---|
1187 | #ifdef AVEN |
---|
1188 | (void)col; |
---|
1189 | aven_v_report(severity, fnm, line, en, ap); |
---|
1190 | #else |
---|
1191 | const char * level; |
---|
1192 | if (fnm) { |
---|
1193 | fputs(fnm, STDERR); |
---|
1194 | if (line) fprintf(STDERR, ":%d", line); |
---|
1195 | if (col > 0) fprintf(STDERR, ":%d", col); |
---|
1196 | } else { |
---|
1197 | fputs(appname_copy, STDERR); |
---|
1198 | } |
---|
1199 | fputs(": ", STDERR); |
---|
1200 | |
---|
1201 | switch (severity) { |
---|
1202 | case DIAG_INFO: |
---|
1203 | /* TRANSLATORS: Indicates a informational message e.g.: |
---|
1204 | * "spoon.svx:12: info: Declination: [...]" */ |
---|
1205 | level = msg_opt(/*info*/485, "info"); |
---|
1206 | break; |
---|
1207 | case DIAG_WARN: |
---|
1208 | /* TRANSLATORS: Indicates a warning message e.g.: |
---|
1209 | * "spoon.svx:12: warning: *prefix is deprecated" */ |
---|
1210 | level = msg_opt(/*warning*/4, "warning"); |
---|
1211 | break; |
---|
1212 | default: |
---|
1213 | /* TRANSLATORS: Indicates an error message e.g.: |
---|
1214 | * "spoon.svx:13:4: error: Field may not be omitted" */ |
---|
1215 | level = msg_opt(/*error*/93, "error"); |
---|
1216 | break; |
---|
1217 | } |
---|
1218 | fputs(level, STDERR); |
---|
1219 | fputs(": ", STDERR); |
---|
1220 | |
---|
1221 | vfprintf(STDERR, msg(en), ap); |
---|
1222 | fputnl(STDERR); |
---|
1223 | #endif |
---|
1224 | |
---|
1225 | switch (severity) { |
---|
1226 | case DIAG_WARN: |
---|
1227 | msg_warnings++; |
---|
1228 | break; |
---|
1229 | case DIAG_ERR: |
---|
1230 | msg_errors++; |
---|
1231 | if (msg_errors == 50) |
---|
1232 | fatalerror_in_file(fnm, 0, /*Too many errors - giving up*/19); |
---|
1233 | break; |
---|
1234 | case DIAG_FATAL: |
---|
1235 | exit(EXIT_FAILURE); |
---|
1236 | } |
---|
1237 | } |
---|
1238 | |
---|
1239 | void |
---|
1240 | diag(int severity, int en, ...) |
---|
1241 | { |
---|
1242 | va_list ap; |
---|
1243 | va_start(ap, en); |
---|
1244 | v_report(severity, NULL, 0, 0, en, ap); |
---|
1245 | va_end(ap); |
---|
1246 | } |
---|
1247 | |
---|
1248 | void |
---|
1249 | diag_in_file(int severity, const char *fnm, int line, int en, ...) |
---|
1250 | { |
---|
1251 | va_list ap; |
---|
1252 | va_start(ap, en); |
---|
1253 | v_report(severity, fnm, line, 0, en, ap); |
---|
1254 | va_end(ap); |
---|
1255 | } |
---|
1256 | |
---|
1257 | /* Code to support switching character set at runtime (e.g. for a printer |
---|
1258 | * driver to support different character sets on screen and on the printer) |
---|
1259 | */ |
---|
1260 | typedef struct charset_li { |
---|
1261 | struct charset_li *next; |
---|
1262 | int code; |
---|
1263 | char **msg_array; |
---|
1264 | } charset_li; |
---|
1265 | |
---|
1266 | static charset_li *charset_head = NULL; |
---|
1267 | |
---|
1268 | static int charset = CHARSET_BAD; |
---|
1269 | |
---|
1270 | int |
---|
1271 | select_charset(int charset_code) |
---|
1272 | { |
---|
1273 | int old_charset = charset; |
---|
1274 | charset_li *p; |
---|
1275 | |
---|
1276 | #ifdef DEBUG |
---|
1277 | fprintf(stderr, "select_charset(%d), old charset = %d\n", charset_code, |
---|
1278 | charset); |
---|
1279 | #endif |
---|
1280 | |
---|
1281 | charset = charset_code; |
---|
1282 | |
---|
1283 | /* check if we've already parsed messages for new charset */ |
---|
1284 | for (p = charset_head; p; p = p->next) { |
---|
1285 | #ifdef DEBUG |
---|
1286 | printf("%p: code %d msg_array %p\n", p, p->code, p->msg_array); |
---|
1287 | #endif |
---|
1288 | if (p->code == charset) { |
---|
1289 | msg_array = p->msg_array; |
---|
1290 | return old_charset; |
---|
1291 | } |
---|
1292 | } |
---|
1293 | |
---|
1294 | /* nope, got to reparse message file */ |
---|
1295 | parse_msg_file(charset_code); |
---|
1296 | |
---|
1297 | /* add to list */ |
---|
1298 | p = osnew(charset_li); |
---|
1299 | p->code = charset; |
---|
1300 | p->msg_array = msg_array; |
---|
1301 | p->next = charset_head; |
---|
1302 | charset_head = p; |
---|
1303 | |
---|
1304 | return old_charset; |
---|
1305 | } |
---|