Context Navigation

message.c @ 863401f

Visit:

RELEASE/1.0RELEASE/1.2debug-cidebug-ci-sanitisersfaster-cavernloglog-selectstereostereo-2025walls-datawalls-data-hanging-as-warningwarn-only-for-hanging-survey

Last change on this file since 863401f was c94999e, checked in by Olly Betts <olly@…>, 20 years ago

Fixed detection of codepage 1250 on windows.

git-svn-id: file:///home/survex-svn/survex/trunk@2800 4b37db11-9a0c-4f06-9ece-9ab7cdaee568

Property mode set to 100644

File size: 48.9 KB

Line
1	/* message.c
2	* Fairly general purpose message and error routines
3	* Copyright (C) 1993-2003,2004,2005 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18	*/
19
20	/#define DEBUG 1/
21
22	#ifdef HAVE_CONFIG_H
23	# include <config.h>
24	#endif
25
26	#include <stdio.h>
27	#include <stdlib.h>
28	#include <string.h>
29	#include <ctype.h>
30	#include <limits.h>
31	#include <errno.h>
32	#include <locale.h>
33
34	#include "cmdline.h"
35	#include "whichos.h"
36	#include "filename.h"
37	#include "message.h"
38	#include "osdepend.h"
39	#include "filelist.h"
40	#include "debug.h"
41
42	#ifdef AVEN
43	# include "aven.h"
44	#endif
45
46	#ifdef HAVE_SIGNAL
47	# ifdef HAVE_SETJMP_H
48	# include <setjmp.h>
49	static jmp_buf jmpbufSignal;
50	# include <signal.h>
51	# else
52	# undef HAVE_SIGNAL
53	# endif
54	#endif
55
56	#if (OS==WIN32)
57	# define WIN32_LEAN_AND_MEAN
58	# include <windows.h>
59	#elif (OS==MSDOS)
60	# include <dos.h>
61	# ifdef __DJGPP__
62	# include <dpmi.h>
63	# include <go32.h>
64	# include <sys/movedata.h>
65	# endif
66	#elif (OS==RISCOS)
67	# include "oslib/wimpreadsy.h"
68	# include "oslib/territory.h"
69	#elif (OS==UNIX)
70	# include <sys/types.h>
71	# include <sys/stat.h>
72	#endif
73
74	/* For funcs which want to be immune from messing around with different
75	* calling conventions */
76	#ifndef CDECL
77	# define CDECL
78	#endif
79
80	int msg_warnings = 0; /* keep track of how many warnings we've given */
81	int msg_errors = 0; /* and how many (non-fatal) errors */
82
83	/* in case osmalloc() fails before appname_copy is set up */
84	static const char *appname_copy = "anonymous program";
85
86	/* error code for failed osmalloc and osrealloc calls */
87	static void
88	outofmem(OSSIZE_T size)
89	{
90	fatalerror(/Out of memory (couldn't find %lu bytes)./1,
91	(unsigned long)size);
92	}
93
94	#ifdef TOMBSTONES
95	#define TOMBSTONE_SIZE 16
96	static const char tombstone[TOMBSTONE_SIZE] = "012345\xfftombstone";
97	#endif
98
99	/* malloc with error catching if it fails. Also allows us to write special
100	* versions easily eg for DOS EMS or MS Windows.
101	*/
102	void Far *
103	osmalloc(OSSIZE_T size)
104	{
105	void Far *p;
106	#ifdef TOMBSTONES
107	size += TOMBSTONE_SIZE * 2;
108	p = malloc(size);
109	#else
110	p = xosmalloc(size);
111	#endif
112	if (p == NULL) outofmem(size);
113	#ifdef TOMBSTONES
114	printf("osmalloc truep=%p truesize=%d\n", p, size);
115	memcpy(p, tombstone, TOMBSTONE_SIZE);
116	memcpy(p + size - TOMBSTONE_SIZE, tombstone, TOMBSTONE_SIZE);
117	(size_t )p = size;
118	p += TOMBSTONE_SIZE;
119	#endif
120	return p;
121	}
122
123	/* realloc with error catching if it fails. */
124	void Far *
125	osrealloc(void *p, OSSIZE_T size)
126	{
127	/* some pre-ANSI realloc implementations don't cope with a NULL pointer */
128	if (p == NULL) {
129	p = xosmalloc(size);
130	} else {
131	#ifdef TOMBSTONES
132	int true_size;
133	size += TOMBSTONE_SIZE * 2;
134	p -= TOMBSTONE_SIZE;
135	true_size = (size_t )p;
136	printf("osrealloc (in truep=%p truesize=%d)\n", p, true_size);
137	if (memcmp(p + sizeof(size_t), tombstone + sizeof(size_t),
138	TOMBSTONE_SIZE - sizeof(size_t)) != 0) {
139	printf("start tombstone for block %p, size %d corrupted!",
140	p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2);
141	}
142	if (memcmp(p + true_size - TOMBSTONE_SIZE, tombstone,
143	TOMBSTONE_SIZE) != 0) {
144	printf("end tombstone for block %p, size %d corrupted!",
145	p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2);
146	}
147	p = realloc(p, size);
148	if (p == NULL) outofmem(size);
149	printf("osrealloc truep=%p truesize=%d\n", p, size);
150	memcpy(p, tombstone, TOMBSTONE_SIZE);
151	memcpy(p + size - TOMBSTONE_SIZE, tombstone, TOMBSTONE_SIZE);
152	(size_t )p = size;
153	p += TOMBSTONE_SIZE;
154	#else
155	p = xosrealloc(p, size);
156	#endif
157	}
158	if (p == NULL) outofmem(size);
159	return p;
160	}
161
162	char Far *
163	osstrdup(const char *str)
164	{
165	char *p;
166	OSSIZE_T len;
167	len = strlen(str) + 1;
168	p = osmalloc(len);
169	memcpy(p, str, len);
170	return p;
171	}
172
173	/* osfree is usually just a macro in osalloc.h */
174	#ifdef TOMBSTONES
175	void
176	osfree(void *p)
177	{
178	int true_size;
179	if (!p) return;
180	p -= TOMBSTONE_SIZE;
181	true_size = (size_t )p;
182	printf("osfree truep=%p truesize=%d\n", p, true_size);
183	if (memcmp(p + sizeof(size_t), tombstone + sizeof(size_t),
184	TOMBSTONE_SIZE - sizeof(size_t)) != 0) {
185	printf("start tombstone for block %p, size %d corrupted!",
186	p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2);
187	}
188	if (memcmp(p + true_size - TOMBSTONE_SIZE, tombstone,
189	TOMBSTONE_SIZE) != 0) {
190	printf("end tombstone for block %p, size %d corrupted!",
191	p + TOMBSTONE_SIZE, true_size - TOMBSTONE_SIZE * 2);
192	}
193	free(p);
194	}
195	#endif
196
197	#ifdef HAVE_SIGNAL
198
199	static int sigReceived;
200
201	/* for systems not using autoconf, assume the signal handler returns void
202	* unless specified elsewhere */
203	#ifndef RETSIGTYPE
204	# define RETSIGTYPE void
205	#endif
206
207	static CDECL RETSIGTYPE Far
208	report_sig(int sig)
209	{
210	sigReceived = sig;
211	longjmp(jmpbufSignal, 1);
212	}
213
214	static void
215	init_signals(void)
216	{
217	int en;
218	if (!setjmp(jmpbufSignal)) {
219	signal(SIGABRT, report_sig); /* abnormal termination eg abort() */
220	signal(SIGFPE, report_sig); /* arithmetic error eg /0 or overflow */
221	signal(SIGILL, report_sig); /* illegal function image eg illegal instruction */
222	signal(SIGSEGV, report_sig); /* illegal storage access eg access outside memory limits */
223	# ifdef SIGSTAK /* only on RISC OS AFAIK */
224	signal(SIGSTAK, report_sig); /* stack overflow */
225	# endif
226	return;
227	}
228
229	switch (sigReceived) {
230	case SIGABRT: en = /Abnormal termination/90; break;
231	case SIGFPE: en = /Arithmetic error/91; break;
232	case SIGILL: en = /Illegal instruction/92; break;
233	case SIGSEGV: en = /Bad memory access/94; break;
234	# ifdef SIGSTAK
235	case SIGSTAK: en = /Stack overflow/96; break;
236	# endif
237	default: en = /Unknown signal received/97; break;
238	}
239	fputsnl(msg(en), STDERR);
240
241	/* Any of the signals we catch indicates a bug */
242	fatalerror(/Bug in program detected! Please report this to the authors/11);
243
244	exit(EXIT_FAILURE);
245	}
246	#endif
247
248	static int
249	default_charset(void)
250	{
251	#if (OS==RISCOS)
252	/* RISCOS 3.1 and above CHARSET_RISCOS31 (ISO_8859_1 + extras in 128-159)
253	* RISCOS < 3.1 is ISO_8859_1 */
254	int version;
255	if (xwimpreadsysinfo_version(&version) != NULL) {
256	/* RISC OS 2 or some error (don't care which) */
257	return CHARSET_ISO_8859_1;
258	}
259
260	/* oddly wimp_VERSION_RO3 is RISC OS 3.1 */
261	if (version < wimp_VERSION_RO3) return CHARSET_ISO_8859_1;
262
263	return CHARSET_RISCOS31;
264	#elif (OS==MSDOS)
265	#ifdef __DJGPP__
266	__dpmi_regs r;
267	r.x.ax = 0x6501;
268	r.x.bx = 0xffff;
269	r.x.dx = 0xffff;
270	/* Use DJGPP's transfer buffer (which is at least 2K) */
271	r.x.es = __tb >> 4;
272	r.x.di = __tb & 0x0f;
273	r.x.cx = 2048;
274	/* bit 1 is the carry flag */
275	if (__dpmi_int(0x21, &r) != -1 && !(r.x.flags & 1)) {
276	unsigned short p;
277	dosmemget(__tb + 5, 2, &p);
278	#else
279	union REGS r;
280	struct SREGS s = { 0 };
281
282	unsigned char buf[48];
283	r.x.ax = 0x6501;
284	r.x.bx = 0xffff;
285	r.x.dx = 0xffff;
286	s.es = FP_SEG(buf);
287	r.x.di = FP_OFF(buf);
288	r.x.cx = 48;
289	intdosx(&r, &r, &s);
290	if (!r.x.cflag) {
291	unsigned short p = buf[5] \| (buf[6] << 8);
292	#endif
293	if (p == 437) return CHARSET_DOSCP437;
294	if (p == 850) return CHARSET_DOSCP850;
295	if (p == 912) return CHARSET_ISO_8859_2;
296	}
297	return CHARSET_USASCII;
298	#elif (OS==WIN32)
299	# ifdef AVEN
300	# define CODEPAGE GetACP()
301	# else
302	# define CODEPAGE GetConsoleOutputCP()
303	# endif
304	switch (CODEPAGE) {
305	case 1252: return CHARSET_WINCP1252;
306	case 1250: return CHARSET_WINCP1250;
307	case 850: return CHARSET_DOSCP850;
308	}
309	return CHARSET_USASCII;
310	#elif (OS==UNIX)
311	#if defined(XCAVEROT) \|\| defined(AVEN)
312	return CHARSET_ISO_8859_1;
313	#else
314	const char *p = getenv("LC_ALL");
315	if (p == NULL \|\| p[0] == '\0') {
316	p = getenv("LC_CTYPE");
317	if (p == NULL \|\| p[0] == '\0') {
318	p = msg_lang;
319	}
320	}
321
322	if (p) {
323	char *q = strchr(p, '.');
324	if (q) p = q + 1;
325	}
326
327	if (p) {
328	const char *chset = p;
329	size_t name_len;
330
331	while (p != '\0' && p != '@') p++;
332
333	name_len = p - chset;
334
335	if (name_len) {
336	int only_digit = 1;
337	size_t cnt;
338
339	for (cnt = 0; cnt < name_len; ++cnt)
340	if (isalpha((unsigned char)chset[cnt])) {
341	only_digit = 0;
342	break;
343	}
344
345	if (only_digit) goto iso;
346
347	switch (tolower(chset[0])) {
348	case 'i':
349	if (tolower(chset[1]) == 's' && tolower(chset[2]) == 'o') {
350	chset += 3;
351	iso:
352	if (strncmp(chset, "8859", 4) == 0) {
353	chset += 4;
354	while (chset < p && chset && !isdigit((unsigned char)chset))
355	chset++;
356	switch (atoi(chset)) {
357	case 1: return CHARSET_ISO_8859_1;
358	case 2: return CHARSET_ISO_8859_2;
359	case 15: return CHARSET_ISO_8859_15;
360	default: return CHARSET_USASCII;
361	}
362	}
363	}
364	break;
365	case 'u':
366	if (tolower(chset[1]) == 't' && tolower(chset[2]) == 'f') {
367	chset += 3;
368	while (chset < p && chset && !isdigit((unsigned char)chset))
369	chset++;
370	switch (atoi(chset)) {
371	case 8: return CHARSET_UTF8;
372	default: return CHARSET_USASCII;
373	}
374	}
375	}
376	}
377	}
378	return CHARSET_USASCII;
379	#endif
380	#else
381	# error Do not know operating system 'OS'
382	#endif
383	}
384
385	/* It seems that Swedish and maybe some other scandanavian languages don't
386	* transliterate ä to ae - but it seems there may be conflicting views
387	* on this...
388	*/
389	#define umlaut_to_e() 1
390
391	/* values <= 127 already dealt with */
392	static int
393	add_unicode(int charset, unsigned char *p, int value)
394	{
395	#ifdef DEBUG
396	fprintf(stderr, "add_unicode(%d, %p, %d)\n", charset, p, value);
397	#endif
398	if (value == 0) return 0;
399	switch (charset) {
400	case CHARSET_USASCII:
401	if (value < 0x80) {
402	*p = value;
403	return 1;
404	}
405	break;
406	case CHARSET_ISO_8859_1:
407	if (value < 0x100) {
408	*p = value;
409	return 1;
410	}
411	break;
412	case CHARSET_ISO_8859_2:
413	if (value >= 0xa0) {
414	int v = 0;
415	switch (value) {
416	case 0xa0: case 0xa4: case 0xa7: case 0xa8: case 0xad: case 0xb0:
417	case 0xb4: case 0xb8: case 0xc1: case 0xc2: case 0xc4: case 0xc7:
418	case 0xc9: case 0xcb: case 0xcd: case 0xce: case 0xd3: case 0xd4:
419	case 0xd6: case 0xd7: case 0xda: case 0xdc: case 0xdd: case 0xdf:
420	case 0xe1: case 0xe2: case 0xe4: case 0xe7: case 0xe9: case 0xeb:
421	case 0xed: case 0xee: case 0xf3: case 0xf4: case 0xf6: case 0xf7:
422	case 0xfa: case 0xfc: case 0xfd:
423	v = value; break;
424	case 0x104: v = '\xa1'; break;
425	case 0x2d8: v = '\xa2'; break;
426	case 0x141: v = '\xa3'; break;
427	case 0x13d: v = '\xa5'; break;
428	case 0x15a: v = '\xa6'; break;
429	case 0x160: v = '\xa9'; break;
430	case 0x15e: v = '\xaa'; break; /* Scedil */
431	case 0x164: v = '\xab'; break;
432	case 0x179: v = '\xac'; break;
433	case 0x17d: v = '\xae'; break;
434	case 0x17b: v = '\xaf'; break;
435	case 0x105: v = '\xb1'; break;
436	case 0x2db: v = '\xb2'; break;
437	case 0x142: v = '\xb3'; break;
438	case 0x13e: v = '\xb5'; break;
439	case 0x15b: v = '\xb6'; break;
440	case 0x2c7: v = '\xb7'; break;
441	case 0x161: v = '\xb9'; break;
442	case 0x15f: v = '\xba'; break; /* scedil */
443	case 0x165: v = '\xbb'; break;
444	case 0x17a: v = '\xbc'; break;
445	case 0x2dd: v = '\xbd'; break;
446	case 0x17e: v = '\xbe'; break;
447	case 0x17c: v = '\xbf'; break;
448	case 0x154: v = '\xc0'; break;
449	case 0x102: v = '\xc3'; break;
450	case 0x139: v = '\xc5'; break;
451	case 0x106: v = '\xc6'; break;
452	case 0x10c: v = '\xc8'; break;
453	case 0x118: v = '\xca'; break;
454	case 0x11a: v = '\xcc'; break;
455	case 0x10e: v = '\xcf'; break;
456	case 0x110: v = '\xd0'; break;
457	case 0x143: v = '\xd1'; break;
458	case 0x147: v = '\xd2'; break;
459	case 0x150: v = '\xd5'; break;
460	case 0x158: v = '\xd8'; break;
461	case 0x16e: v = '\xd9'; break;
462	case 0x170: v = '\xdb'; break;
463	case 0x162: v = '\xde'; break; /* &Tcedil; */
464	case 0x155: v = '\xe0'; break;
465	case 0x103: v = '\xe3'; break;
466	case 0x13a: v = '\xe5'; break;
467	case 0x107: v = '\xe6'; break;
468	case 0x10d: v = '\xe8'; break;
469	case 0x119: v = '\xea'; break;
470	case 0x11b: v = '\xec'; break;
471	case 0x10f: v = '\xef'; break;
472	case 0x111: v = '\xf0'; break;
473	case 0x144: v = '\xf1'; break;
474	case 0x148: v = '\xf2'; break;
475	case 0x151: v = '\xf5'; break;
476	case 0x159: v = '\xf8'; break;
477	case 0x16f: v = '\xf9'; break;
478	case 0x171: v = '\xfb'; break;
479	case 0x163: v = '\xfe'; break; /* tcedil */
480	case 0x2d9: v = '\xff'; break;
481	}
482	if (v == 0) break;
483	value = v;
484	}
485	*p = value;
486	return 1;
487	case CHARSET_ISO_8859_15:
488	switch (value) {
489	case 0xa4: case 0xa6: case 0xb0: case 0xc4:
490	case 0xd0: case 0xd4: case 0xd5: case 0xd6:
491	goto donthave;
492	case 0x152: value = 0xd4; break; /* &OElig; */
493	case 0x153: value = 0xd5; break; /* &oelig; */
494	#if 0
495	case 0x0: value = 0xa4; break; /* euro */
496	#endif
497	case 0x160: value = 0xa6; break; /* Scaron */
498	case 0x161: value = 0xb0; break; /* scaron */
499	case 0x17d: value = 0xc4; break; /* Zcaron */
500	case 0x17e: value = 0xd0; break; /* zcaron */
501	#if 0
502	case 0x0: value = 0xd6; break; /* Ydiersis */
503	#endif
504	}
505	if (value < 0x100) {
506	*p = value;
507	return 1;
508	}
509	donthave:
510	break;
511	#if (OS==RISCOS)
512	case CHARSET_RISCOS31:
513	/* RISC OS 3.1 (and later) extensions to ISO-8859-1 */
514	switch (value) {
515	case 0x152: value = 0x9a; break; /* &OElig; */
516	case 0x153: value = 0x9b; break; /* &oelig; */
517	#if 0
518	case 0x174: value = 0x81; break; /* &Wcirc; */
519	case 0x175: value = 0x82; break; /* &wcirc; */
520	case 0x176: value = 0x85; break; /* &Ycirc; */
521	case 0x177: value = 0x86; break; /* &ycirc; */
522	#endif
523	}
524	if (value < 0x100) {
525	*p = value;
526	return 1;
527	}
528	break;
529	#elif (OS==WIN32)
530	case CHARSET_WINCP1250:
531	/* MS Windows rough equivalent to ISO-8859-2 */
532	if (value >= 0x80) {
533	int v = 0;
534	switch (value) {
535	case 0xa0: case 0xa4: case 0xa6: case 0xa7: case 0xa8: case 0xa9:
536	case 0xab: case 0xac: case 0xad: case 0xae: case 0xb0: case 0xb1:
537	case 0xb4: case 0xb5: case 0xb6: case 0xb7: case 0xb8: case 0xbb:
538	case 0xc1: case 0xc2: case 0xc4: case 0xc7: case 0xc9: case 0xcb:
539	case 0xcd: case 0xce: case 0xd3: case 0xd4: case 0xd6: case 0xd7:
540	case 0xda: case 0xdc: case 0xdd: case 0xdf: case 0xe1: case 0xe2:
541	case 0xe4: case 0xe7: case 0xe9: case 0xeb: case 0xed: case 0xee:
542	case 0xf3: case 0xf4: case 0xf6: case 0xf7: case 0xfa: case 0xfc:
543	case 0xfd:
544	v = value; break;
545	case 0x20ac: v = '\x80'; break;
546	case 0x201a: v = '\x82'; break;
547	case 0x201e: v = '\x84'; break;
548	case 0x2026: v = '\x85'; break;
549	case 0x2020: v = '\x86'; break;
550	case 0x2021: v = '\x87'; break;
551	case 0x2030: v = '\x89'; break;
552	case 0x0160: v = '\x8a'; break;
553	case 0x2039: v = '\x8b'; break;
554	case 0x015a: v = '\x8c'; break;
555	case 0x0164: v = '\x8d'; break;
556	case 0x017d: v = '\x8e'; break;
557	case 0x0179: v = '\x8f'; break;
558	case 0x2018: v = '\x91'; break;
559	case 0x2019: v = '\x92'; break;
560	case 0x201c: v = '\x93'; break;
561	case 0x201d: v = '\x94'; break;
562	case 0x2022: v = '\x95'; break;
563	case 0x2013: v = '\x96'; break;
564	case 0x2014: v = '\x97'; break;
565	case 0x2122: v = '\x99'; break;
566	case 0x0161: v = '\x9a'; break;
567	case 0x203a: v = '\x9b'; break;
568	case 0x015b: v = '\x9c'; break;
569	case 0x0165: v = '\x9d'; break;
570	case 0x017e: v = '\x9e'; break;
571	case 0x017a: v = '\x9f'; break;
572	case 0x02c7: v = '\xa1'; break;
573	case 0x02d8: v = '\xa2'; break;
574	case 0x0141: v = '\xa3'; break;
575	case 0x0104: v = '\xa5'; break;
576	case 0x015e: v = '\xaa'; break; /* Scedil */
577	case 0x017b: v = '\xaf'; break;
578	case 0x02db: v = '\xb2'; break;
579	case 0x0142: v = '\xb3'; break;
580	case 0x0105: v = '\xb9'; break;
581	case 0x015f: v = '\xba'; break; /* scedil */
582	case 0x013d: v = '\xbc'; break;
583	case 0x02dd: v = '\xbd'; break;
584	case 0x013e: v = '\xbe'; break;
585	case 0x017c: v = '\xbf'; break;
586	case 0x0154: v = '\xc0'; break;
587	case 0x0102: v = '\xc3'; break;
588	case 0x0139: v = '\xc5'; break;
589	case 0x0106: v = '\xc6'; break;
590	case 0x010c: v = '\xc8'; break;
591	case 0x0118: v = '\xca'; break;
592	case 0x011a: v = '\xcc'; break;
593	case 0x010e: v = '\xcf'; break;
594	case 0x0110: v = '\xd0'; break;
595	case 0x0143: v = '\xd1'; break;
596	case 0x0147: v = '\xd2'; break;
597	case 0x0150: v = '\xd5'; break;
598	case 0x0158: v = '\xd8'; break;
599	case 0x016e: v = '\xd9'; break;
600	case 0x0170: v = '\xdb'; break;
601	case 0x0162: v = '\xde'; break; /* &Tcedil; */
602	case 0x0155: v = '\xe0'; break;
603	case 0x0103: v = '\xe3'; break;
604	case 0x013a: v = '\xe5'; break;
605	case 0x0107: v = '\xe6'; break;
606	case 0x010d: v = '\xe8'; break;
607	case 0x0119: v = '\xea'; break;
608	case 0x011b: v = '\xec'; break;
609	case 0x010f: v = '\xef'; break;
610	case 0x0111: v = '\xf0'; break;
611	case 0x0144: v = '\xf1'; break;
612	case 0x0148: v = '\xf2'; break;
613	case 0x0151: v = '\xf5'; break;
614	case 0x0159: v = '\xf8'; break;
615	case 0x016f: v = '\xf9'; break;
616	case 0x0171: v = '\xfb'; break;
617	case 0x0163: v = '\xfe'; break; /* tcedil */
618	case 0x02d9: v = '\xff'; break;
619	}
620	if (v == 0) break;
621	value = v;
622	}
623	*p = value;
624	return 1;
625	case CHARSET_WINCP1252:
626	/* MS Windows extensions to ISO-8859-1 */
627	switch (value) {
628	case 0x152: value = 0x8c; break; /* &OElig; */
629	case 0x153: value = 0x9c; break; /* &oelig; */
630	#if 0
631	/* there are a few other obscure ones we don't currently need */
632	#endif
633	}
634	if (value < 0x100) {
635	*p = value;
636	return 1;
637	}
638	break;
639	#endif
640	#if (OS==MSDOS)
641	case CHARSET_DOSCP437: {
642	unsigned char uni2dostab[] = {
643	255, 173, 155, 156, 0, 157, 0, 0,
644	0, 0, 166, 174, 170, 0, 0, 0,
645	248, 241, 253, 0, 0, 230, 0, 250,
646	0, 0, 167, 175, 172, 171, 0, 168,
647	0, 0, 0, 0, 142, 143, 146, 128,
648	0, 144, 0, 0, 0, 0, 0, 0,
649	0, 165, 0, 0, 0, 0, 153, 0,
650	0, 0, 0, 0, 154, 0, 0, 225,
651	133, 160, 131, 0, 132, 134, 145, 135,
652	138, 130, 136, 137, 141, 161, 140, 139,
653	0, 164, 149, 162, 147, 0, 148, 246,
654	0, 151, 163, 150, 129, 0, 0, 152,
655	};
656	if (value >= 160 && value < 256) {
657	int ch = (int)uni2dostab[value - 160];
658	if (!ch) break;
659	*p = ch;
660	return 1;
661	}
662	#if 0
663	switch (value) {
664	case 8359: p = 158; return 1; / PESETA SIGN */
665	case 402: p = 159; return 1; / LATIN SMALL LETTER F WITH HOOK */
666	case 8976: p = 169; return 1; / REVERSED NOT SIGN */
667	case 945: p = 224; return 1; / GREEK SMALL LETTER ALPHA */
668	case 915: p = 226; return 1; / GREEK CAPITAL LETTER GAMMA */
669	case 960: p = 227; return 1; / GREEK SMALL LETTER PI */
670	case 931: p = 228; return 1; / GREEK CAPITAL LETTER SIGMA */
671	case 963: p = 229; return 1; / GREEK SMALL LETTER SIGMA */
672	case 964: p = 231; return 1; / GREEK SMALL LETTER TAU */
673	case 934: p = 232; return 1; / GREEK CAPITAL LETTER PHI */
674	case 920: p = 233; return 1; / GREEK CAPITAL LETTER THETA */
675	case 937: p = 234; return 1; / GREEK CAPITAL LETTER OMEGA */
676	case 948: p = 235; return 1; / GREEK SMALL LETTER DELTA */
677	case 8734: p = 236; return 1; / INFINITY */
678	case 966: p = 237; return 1; / GREEK SMALL LETTER PHI */
679	case 949: p = 238; return 1; / GREEK SMALL LETTER EPSILON */
680	case 8745: p = 239; return 1; / INTERSECTION */
681	case 8801: p = 240; return 1; / IDENTICAL TO */
682	case 8805: p = 242; return 1; / GREATER-THAN OR EQUAL TO */
683	case 8804: p = 243; return 1; / LESS-THAN OR EQUAL TO */
684	case 8992: p = 244; return 1; / TOP HALF INTEGRAL */
685	case 8993: p = 245; return 1; / BOTTOM HALF INTEGRAL */
686	case 8776: p = 247; return 1; / ALMOST EQUAL TO */
687	case 8729: p = 249; return 1; / BULLET OPERATOR */
688	case 8730: p = 251; return 1; / SQUARE ROOT */
689	case 8319: p = 252; return 1; / SUPERSCRIPT LATIN SMALL LETTER N */
690	case 9632: p = 254; return 1; / BLACK SQUARE */
691	}
692	#endif
693	break;
694	}
695	#endif
696	#if (OS==MSDOS \|\| OS==WIN32)
697	case CHARSET_DOSCP850: {
698	unsigned char uni2dostab[] = {
699	255, 173, 189, 156, 207, 190, 221, 245,
700	249, 184, 166, 174, 170, 240, 169, 238,
701	248, 241, 253, 252, 239, 230, 244, 250,
702	247, 251, 167, 175, 172, 171, 243, 168,
703	183, 181, 182, 199, 142, 143, 146, 128,
704	212, 144, 210, 211, 222, 214, 215, 216,
705	209, 165, 227, 224, 226, 229, 153, 158,
706	157, 235, 233, 234, 154, 237, 232, 225,
707	133, 160, 131, 198, 132, 134, 145, 135,
708	138, 130, 136, 137, 141, 161, 140, 139,
709	208, 164, 149, 162, 147, 228, 148, 246,
710	155, 151, 163, 150, 129, 236, 231, 152
711	};
712	if (value >= 160 && value < 256) {
713	*p = (int)uni2dostab[value - 160];
714	return 1;
715	}
716	#if 0
717	if (value == 305) { /* LATIN SMALL LETTER DOTLESS I */
718	*p = 213;
719	return 1;
720	}
721	if (value == 402) { /* LATIN SMALL LETTER F WITH HOOK */
722	*p = 159;
723	return 1;
724	}
725	#endif
726	break;
727	}
728	#endif
729	}
730	/* Transliterate characters we can't represent */
731	#ifdef DEBUG
732	fprintf(stderr, "transliterate `%c' 0x%x\n", value, value);
733	#endif
734	switch (value) {
735	case 160:
736	*p = ' '; return 1;
737	case 161 /* ¡ */:
738	*p = '!'; return 1;
739	case 171 /* « */:
740	p[1] = *p = '<'; return 2;
741	case 187 /* » */:
742	p[1] = *p = '>'; return 2;
743	case 191 /* ¿ */:
744	*p = '?'; return 1;
745	case 192 /* À /: case 193 / Á /: case 194 / Â /: case 195 / Ã */:
746	*p = 'A'; return 1;
747	case 197 /* Å */:
748	p[1] = *p = 'A'; return 2;
749	case 196 /* Ä /: / Ä */
750	*p = 'A';
751	if (!umlaut_to_e()) return 1;
752	p[1] = 'E'; return 2;
753	case 198 /* Æ */:
754	*p = 'A'; p[1] = 'E'; return 2;
755	case 199 /* Ç /: case 268: / &Ccaron; */
756	*p = 'C'; return 1;
757	case 270: /* &Dcaron; */
758	*p = 'D'; return 1;
759	case 200 /* È /: case 201 / É /: case 202 / Ê /: case 203 / Ë */:
760	*p = 'E'; return 1;
761	case 204 /* Ì /: case 205 / Í /: case 206 / Î /: case 207 / Ï */:
762	*p = 'I'; return 1;
763	case 208 /* Ð /: case 222 / Þ */:
764	*p = 'T'; p[1] = 'H'; return 2;
765	case 315: /* &Lacute; */
766	case 317: /* &Lcaron; */
767	*p = 'L'; return 1;
768	case 209 /* Ñ */:
769	*p = 'N'; return 1;
770	case 210 /* Ò /: case 211 / Ó /: case 212 / Ô /: case 213 / Õ */:
771	*p = 'O'; return 1;
772	case 214 /* Ö /: / Ö / case 0x152: / &OElig; */
773	*p = 'O'; p[1] = 'E'; return 2;
774	case 352: /* &Scaron; */
775	case 0x15e: /* &Scedil; */
776	*p = 'S'; return 1;
777	case 0x162: /* &Tcedil; */
778	case 0x164: /* &Tcaron; */
779	*p = 'T'; return 1;
780	case 217 /* Ù /: case 218 / Ú /: case 219 / Û */:
781	*p = 'U'; return 1;
782	case 220 /* Ü /: / Ü */
783	*p = 'U'; p[1] = 'E'; return 2;
784	case 221 /* Ý */:
785	*p = 'Y'; return 1;
786	case 381: /* &Zcaron; */
787	*p = 'Z'; return 1;
788	case 223 /* ß */:
789	p[1] = *p = 's'; return 2;
790	case 224 /* à /: case 225 / á /: case 226 / â /: case 227 / ã */:
791	case 259: /* &abreve; */
792	*p = 'a'; return 1;
793	case 228 /* ä /: / ä / case 230 / æ */:
794	*p = 'a'; p[1] = 'e'; return 2;
795	case 229 /* å */:
796	p[1] = *p = 'a'; return 2;
797	case 231 /* ç /: case 269 / &ccaron; */:
798	*p = 'c'; return 1;
799	case 271: /* &dcaron; */
800	*p = 'd'; return 1;
801	case 232 /* è /: case 233 / é /: case 234 / ê /: case 235 / ë */:
802	case 283 /* &ecaron; */:
803	*p = 'e'; return 1;
804	case 236 /* ì /: case 237 / í /: case 238 / î /: case 239 / ï */:
805	*p = 'i'; return 1;
806	case 316 /* &lacute; */:
807	case 318 /* &lcaron; */:
808	*p = 'l'; return 1;
809	case 241 /* ñ /: case 328 / &ncaron; */:
810	*p = 'n'; return 1;
811	case 345: /* &rcaron; */
812	*p = 'r'; return 1;
813	case 353: /* &scaron; */
814	case 0x15f: /* &scedil; */
815	*p = 's'; return 1;
816	case 357: /* &tcaron; */
817	case 0x163: /* &tcedil; */
818	*p = 't'; return 1;
819	case 240 /* ð /: case 254 / þ */:
820	*p = 't'; p[1] = 'h'; return 2;
821	case 242 /* ò /: case 243 / ó /: case 244 / ô /: case 245 / õ */:
822	*p = 'o'; return 1;
823	case 246 /* ö /: / ö / case 0x153: / &oelig; */
824	*p = 'o'; p[1] = 'e'; return 2;
825	case 249 /* ù /: case 250 / ú /: case 251 / û */:
826	case 367 /* &uring; */:
827	*p = 'u'; return 1;
828	case 252 /* ü /: / ü */
829	*p = 'u'; p[1] = 'e'; return 2;
830	case 253 /* ý /: case 255 / ÿ */:
831	*p = 'y'; return 1;
832	case 382: /* &zcaron; */
833	*p = 'z'; return 1;
834	}
835	#ifdef DEBUG
836	fprintf(stderr, "failed to transliterate\n");
837	#endif
838	return 0;
839	}
840
841	#if (OS==UNIX) && defined(DATADIR) && defined(PACKAGE)
842	/* Under Unix, we compile in the configured path */
843	static const char *pth_cfg_files = DATADIR "/" PACKAGE;
844	#else
845	/* On other platforms, we fall back on looking in the current directory */
846	static const char *pth_cfg_files = "";
847	#endif
848
849	static int num_msgs = 0;
850	static char **msg_array = NULL;
851
852	const char *msg_lang = NULL;
853	const char *msg_lang2 = NULL;
854
855	static char **
856	parse_msgs(int n, unsigned char *p, int charset_code) {
857	int i;
858
859	char *msgs = osmalloc(n sizeof(char *));
860
861	for (i = 0; i < n; i++) {
862	unsigned char *to = p;
863	int ch;
864	msgs[i] = (char *)p;
865
866	/* If we want UTF8 anyway, we just need to find the start of each
867	* message */
868	if (charset_code == CHARSET_UTF8) {
869	p += strlen((char *)p) + 1;
870	continue;
871	}
872
873	while ((ch = *p++) != 0) {
874	/* A byte in the range 0x80-0xbf or 0xf0-0xff isn't valid in
875	* this state, (0xf0-0xfd mean values > 0xffff) so treat as
876	* literal and try to resync so we cope better when fed
877	* non-utf-8 data. Similarly we abandon a multibyte sequence
878	* if we hit an invalid character. */
879	if (ch >= 0xc0 && ch < 0xf0) {
880	int ch1 = *p;
881	if ((ch1 & 0xc0) != 0x80) goto resync;
882
883	if (ch < 0xe0) {
884	/* 2 byte sequence */
885	ch = ((ch & 0x1f) << 6) \| (ch1 & 0x3f);
886	p++;
887	} else {
888	/* 3 byte sequence */
889	int ch2 = p[1];
890	if ((ch2 & 0xc0) != 0x80) goto resync;
891	ch = ((ch & 0x1f) << 12) \| ((ch1 & 0x3f) << 6) \| (ch2 & 0x3f);
892	p += 2;
893	}
894	}
895
896	resync:
897
898	if (ch < 127) {
899	*to++ = (char)ch;
900	} else {
901	/* We assume an N byte UTF-8 code never transliterates to more
902	* than N characters (so we can't transliterate © to (C) or
903	* ® to (R) for example) */
904	to += add_unicode(charset_code, to, ch);
905	}
906	}
907	*to++ = '\0';
908	}
909	return msgs;
910	}
911
912	/* This is the name of the default language, which can be set like so:
913	* ./configure --enable-defaultlang=fr
914	*/
915	#ifdef DEFAULTLANG
916	/* No point extracting these errors as they won't get used if file opens */
917	# include "../lib/defaultlang.h"
918	#else
919	#define N_DONTEXTRACTMSGS 5
920	static unsigned char dontextractmsgs[] =
921	"Can't open message file `%s' using path `%s'\0"/1000/
922	"Problem with message file `%s'\0"/1001/
923	"I don't understand this message file version\0"/1002/
924	"Message file truncated?\0"/1003/
925	"Out of memory (couldn't find %lu bytes).\0"/1004/;
926	#endif
927
928	static char **dontextract = NULL;
929
930	static void
931	parse_msg_file(int charset_code)
932	{
933	FILE *fh;
934	unsigned char header[20];
935	int i;
936	unsigned len;
937	unsigned char *p;
938	char fnm, s;
939	int n;
940
941	#ifdef DEBUG
942	fprintf(stderr, "parse_msg_file(%d)\n", charset_code);
943	#endif
944
945	/* sort out messages we need to print if we can't open the message file */
946	dontextract = parse_msgs(N_DONTEXTRACTMSGS, dontextractmsgs, charset_code);
947
948	fnm = osstrdup(msg_lang);
949	/* trim off charset from stuff like "de_DE.iso8859_1" */
950	s = strchr(fnm, '.');
951	if (s) *s = '\0';
952
953	fh = fopenWithPthAndExt(pth_cfg_files, fnm, EXT_SVX_MSG, "rb", NULL);
954
955	if (!fh) {
956	/* e.g. if 'en_GB' is unknown, see if we know 'en' */
957	if (strlen(fnm) > 3 && fnm[2] == '_') {
958	fnm[2] = '\0';
959	fh = fopenWithPthAndExt(pth_cfg_files, fnm, EXT_SVX_MSG, "rb", NULL);
960	if (!fh) fnm[2] = '_'; /* for error reporting */
961	}
962	}
963
964	if (!fh) {
965	fatalerror(/Can't open message file `%s' using path `%s'/1000,
966	fnm, pth_cfg_files);
967	}
968
969	if (fread(header, 1, 20, fh) < 20 \|\|
970	memcmp(header, "Svx\nMsg\r\n\xfe\xff", 12) != 0) {
971	fatalerror(/Problem with message file `%s'/1001, fnm);
972	}
973
974	if (header[12] != 0)
975	fatalerror(/I don't understand this message file version/1002);
976
977	n = (header[14] << 8) \| header[15];
978
979	len = 0;
980	for (i = 16; i < 20; i++) len = (len << 8) \| header[i];
981
982	p = osmalloc(len);
983	if (fread(p, 1, len, fh) < len)
984	fatalerror(/Message file truncated?/1003);
985
986	fclose(fh);
987
988	#ifdef DEBUG
989	fprintf(stderr, "fnm = `%s', n = %d, len = %d\n", fnm, n, len);
990	#endif
991	osfree(fnm);
992
993	msg_array = parse_msgs(n, p, charset_code);
994	num_msgs = n;
995	}
996
997	const char *
998	msg_cfgpth(void)
999	{
1000	return pth_cfg_files;
1001	}
1002
1003	const char *
1004	msg_appname(void)
1005	{
1006	return appname_copy;
1007	}
1008
1009	void
1010	msg_init(char * const *argv)
1011	{
1012	char *p;
1013	SVX_ASSERT(argv);
1014
1015	#ifdef HAVE_SIGNAL
1016	init_signals();
1017	#endif
1018	/* Point to argv[0] itself so we report a more helpful error if the
1019	* code to work out the clean appname generates a signal */
1020	appname_copy = argv[0];
1021	#if (OS == UNIX)
1022	/* use name as-is on Unix - programs run from path get name as supplied */
1023	appname_copy = osstrdup(argv[0]);
1024	#else
1025	/* use the lower-cased leafname on other platforms */
1026	p = leaf_from_fnm(argv[0]);
1027	appname_copy = p;
1028	while (*p) {
1029	p = tolower(p);
1030	++p;
1031	}
1032	#endif
1033
1034	/* shortcut --version so you can check the version number even when the
1035	* correct message file can't be found... */
1036	if (argv[1] && strcmp(argv[1], "--version") == 0) {
1037	cmdline_version();
1038	exit(0);
1039	}
1040	if (argv[0]) {
1041	#ifdef MACOSX_BUNDLE
1042	/* If we're being built into a bundle, always look relative to
1043	* the path to the binary. */
1044	char * pth = path_from_fnm(argv[0]);
1045	pth_cfg_files = use_path(pth, "share/survex");
1046	osfree(pth);
1047	#elif (OS==UNIX) && defined(DATADIR) && defined(PACKAGE)
1048	bool free_pth = fFalse;
1049	char *pth = getenv("srcdir");
1050	if (!pth \|\| !pth[0]) {
1051	pth = path_from_fnm(argv[0]);
1052	free_pth = fTrue;
1053	}
1054	if (pth[0]) {
1055	struct stat buf;
1056	#if defined(__GNUC__) && defined(__APPLE_CC__)
1057	/* On MacOS X the programs may be installed anywhere, with the
1058	* share directory and the binaries in the same directory. */
1059	p = use_path(pth, "share/survex/en.msg");
1060	if (lstat(p, &buf) == 0 && S_ISREG(buf.st_mode)) {
1061	pth_cfg_files = use_path(pth, "share/survex");
1062	goto macosx_got_msg;
1063	}
1064	osfree(p);
1065	#endif
1066	/* If we're run with an explicit path, check if "../lib/en.msg"
1067	* from the program's path exists, and if so look there for
1068	* support files - this allows us to test binaries in the build
1069	* tree easily. */
1070	p = use_path(pth, "../lib/en.msg");
1071	if (lstat(p, &buf) == 0) {
1072	#ifdef S_ISDIR
1073	/* POSIX way */
1074	if (S_ISREG(buf.st_mode)) {
1075	pth_cfg_files = use_path(pth, "../lib");
1076	}
1077	#else
1078	/* BSD way */
1079	if ((buf.st_mode & S_IFMT) == S_IFREG) {
1080	pth_cfg_files = use_path(pth, "../lib");
1081	}
1082	#endif
1083	}
1084	#if defined(__GNUC__) && defined(__APPLE_CC__)
1085	macosx_got_msg:
1086	#endif
1087	osfree(p);
1088	}
1089
1090	if (free_pth) osfree(pth);
1091	#elif (OS==WIN32)
1092	DWORD len = 256;
1093	char buf = NULL, modname;
1094	while (1) {
1095	DWORD got;
1096	buf = osrealloc(buf, len);
1097	got = GetModuleFileName(NULL, buf, len);
1098	if (got < len) break;
1099	len += len;
1100	}
1101	modname = buf;
1102	/* Strange Win32 nastiness - strip prefix "\\?\" if present */
1103	if (strncmp(modname, "\\\\?\\", 4) == 0) modname += 4;
1104	pth_cfg_files = path_from_fnm(modname);
1105	osfree(buf);
1106	#else
1107	/* Get the path to the support files from argv[0] */
1108	pth_cfg_files = path_from_fnm(argv[0]);
1109	#endif
1110	}
1111
1112	msg_lang = getenv("SURVEXLANG");
1113	#ifdef DEBUG
1114	fprintf(stderr, "msg_lang = %p (= \"%s\")\n", msg_lang, msg_lang?msg_lang:"(null)");
1115	#endif
1116
1117	if (!msg_lang \|\| !*msg_lang) {
1118	msg_lang = getenv("LANG");
1119	if (!msg_lang \|\| !*msg_lang) {
1120	#if (OS==WIN32)
1121	LCID locid;
1122	#elif (OS==RISCOS)
1123	territory_t t;
1124	#endif
1125	#ifdef DEFAULTLANG
1126	msg_lang = STRING(DEFAULTLANG);
1127	#else
1128	msg_lang = "en";
1129	#endif
1130	#if (OS==WIN32)
1131	locid = GetUserDefaultLCID();
1132	if (locid) {
1133	WORD langid = LANGIDFROMLCID(locid);
1134	switch (PRIMARYLANGID(langid)) {
1135	/* older mingw compilers don't seem to supply this value */
1136	#ifndef LANG_CATALAN
1137	# define LANG_CATALAN 0x03
1138	#endif
1139	case LANG_CATALAN:
1140	msg_lang = "ca";
1141	break;
1142	case LANG_CHINESE:
1143	msg_lang = "zh";
1144	break;
1145	case LANG_ENGLISH:
1146	if (SUBLANGID(langid) == SUBLANG_ENGLISH_US)
1147	msg_lang = "en_US";
1148	else
1149	msg_lang = "en";
1150	break;
1151	case LANG_FRENCH:
1152	msg_lang = "fr";
1153	break;
1154	case LANG_GERMAN:
1155	switch (SUBLANGID(langid)) {
1156	case SUBLANG_GERMAN_SWISS:
1157	msg_lang = "de_CH";
1158	break;
1159	case SUBLANG_GERMAN:
1160	msg_lang = "de_DE";
1161	break;
1162	default:
1163	msg_lang = "de";
1164	}
1165	break;
1166	case LANG_ITALIAN:
1167	msg_lang = "it";
1168	break;
1169	case LANG_PORTUGUESE:
1170	if (SUBLANGID(langid) == SUBLANG_PORTUGUESE_BRAZILIAN)
1171	msg_lang = "pt_BR";
1172	else
1173	msg_lang = "pt";
1174	break;
1175	case LANG_ROMANIAN:
1176	msg_lang = "ro";
1177	break;
1178	case LANG_SLOVAK:
1179	msg_lang = "sk";
1180	break;
1181	case LANG_SPANISH:
1182	msg_lang = "es";
1183	break;
1184	}
1185	}
1186	#elif (OS==RISCOS)
1187	if (!xterritory_number(&t)) switch (t) {
1188	case 1: /* UK */
1189	case 2: /* Master */
1190	case 3: /* Compact */
1191	case 17: /* Canada1 */
1192	case 19: /* Canada */
1193	case 22: /* Ireland */
1194	msg_lang = "en";
1195	break;
1196	case 4: /* Italy */
1197	msg_lang = "it";
1198	break;
1199	case 5: /* Spain (or ca) */
1200	case 27: /* Mexico */
1201	case 28: /* LatinAm (or pt_BR) */
1202	msg_lang = "es";
1203	break;
1204	case 6: /* France */
1205	case 18: /* Canada2 */
1206	msg_lang = "fr";
1207	break;
1208	case 7: /* Germany */
1209	msg_lang = "de_DE";
1210	break;
1211	case 8: /* Portugal */
1212	msg_lang = "pt";
1213	break;
1214	case 23: /* Hong Kong */
1215	msg_lang = "zh";
1216	break;
1217	case 48: /* USA */
1218	msg_lang = "en_US";
1219	break;
1220	#if 0
1221	case 9: /* Esperanto */
1222	case 10: /* Greece */
1223	case 11: /* Sweden */
1224	case 12: /* Finland */
1225	case 13: /* Unused */
1226	case 14: /* Denmark */
1227	case 15: /* Norway */
1228	case 16: /* Iceland */
1229	case 20: /* Turkey */
1230	case 21: /* Arabic */
1231	case 24: /* Russia */
1232	case 25: /* Russia2 */
1233	case 26: /* Israel */
1234	#endif
1235	}
1236	#elif (OS==MSDOS)
1237	{
1238	int country_code;
1239	# ifdef __DJGPP__
1240	__dpmi_regs r;
1241	r.x.ax = 0x6501;
1242	r.x.bx = 0xffff;
1243	r.x.dx = 0xffff;
1244	/* Use DJGPP's transfer buffer (which is at least 2K) */
1245	r.x.es = __tb >> 4;
1246	r.x.di = __tb & 0x0f;
1247	r.x.cx = 2048;
1248	/* bit 1 is the carry flag */
1249	if (__dpmi_int(0x21, &r) != -1 && !(r.x.flags & 1)) {
1250	unsigned short val;
1251	dosmemget(__tb + 3, 2, &val);
1252	country_code = val;
1253	# else
1254	union REGS r;
1255	r.x.ax = 0x3800; /* get current country info */
1256	r.x.dx = 0;
1257	intdos(&r, &r);
1258	if (!r.x.cflag) {
1259	country_code = r.x.bx;
1260	# endif
1261	/* List of country codes taken from:
1262	* http://www.delorie.com/djgpp/doc/rbinter/it/00/14.html */
1263	/* The mappings here are guesses at best in most cases.
1264	* In a lot of cases we pick a language because we have
1265	* a translation in it, rather than because it's the most
1266	* widely used or understood in that country. */
1267	/* Improvements welcome */
1268	switch (country_code) {
1269	case 1: /* United States */
1270	case 670: /* Saipan / N. Mariana Island */
1271	case 671: /* Guam */
1272	case 680: /* Palau */
1273	case 684: /* American Samoa */
1274	case 691: /* Micronesia */
1275	case 692: /* Marshall Islands */
1276	msg_lang = "en_US";
1277	break;
1278	case 4: /* Canada (English) */
1279	case 27: /* South Africa */
1280	case 44: /* United Kingdom */
1281	case 61: /* International English / Australia */
1282	case 64: /* New Zealand */
1283	case 99: /* Asia (English) */
1284	case 220: /* Gambia */
1285	case 231: /* Liberia */
1286	case 232: /* Sierra Leone */
1287	case 233: /* Ghana */
1288	case 254: /* Kenya */
1289	case 256: /* Uganda */
1290	case 260: /* Zambia */
1291	case 263: /* Zimbabwe */
1292	case 264: /* Namibia */
1293	case 267: /* Botswana */
1294	case 268: /* Swaziland */
1295	case 290: /* St. Helena */
1296	case 297: /* Aruba */
1297	case 350: /* Gibraltar */
1298	case 353: /* Ireland */
1299	case 356: /* Malta */
1300	case 500: /* Falkland Islands */
1301	case 501: /* Belize */
1302	case 592: /* Guyana */
1303	case 672: /* Norfolk Island (Australia) / Christmas Island/Cocos Islands / Antartica */
1304	case 673: /* Brunei Darussalam */
1305	case 674: /* Nauru */
1306	case 675: /* Papua New Guinea */
1307	case 676: /* Tonga Islands */
1308	case 677: /* Solomon Islands */
1309	case 679: /* Fiji */
1310	case 682: /* Cook Islands */
1311	case 683: /* Niue */
1312	case 685: /* Western Samoa */
1313	case 686: /* Kiribati */
1314	/* I believe only some of these are English speaking... */
1315	case 809: /* Antigua and Barbuda / Anguilla / Bahamas / Barbados / Bermuda
1316	British Virgin Islands / Cayman Islands / Dominica
1317	Dominican Republic / Grenada / Jamaica / Montserra
1318	St. Kitts and Nevis / St. Lucia / St. Vincent and Grenadines
1319	Trinidad and Tobago / Turks and Caicos */
1320	msg_lang = "en";
1321	break;
1322	case 2: /* Canadian-French */
1323	case 32: /* Belgium / / maybe */
1324	case 33: /* France */
1325	case 213: /* Algeria */
1326	case 216: /* Tunisia */
1327	case 221: /* Senegal */
1328	case 223: /* Mali */
1329	case 225: /* Ivory Coast */
1330	case 226: /* Burkina Faso */
1331	case 227: /* Niger */
1332	case 228: /* Togo */
1333	case 229: /* Benin */
1334	case 230: /* Mauritius */
1335	case 235: /* Chad */
1336	case 236: /* Central African Republic */
1337	case 237: /* Cameroon */
1338	case 241: /* Gabon */
1339	case 242: /* Congo */
1340	case 250: /* Rwhanda */
1341	case 253: /* Djibouti */
1342	case 257: /* Burundi */
1343	case 261: /* Madagascar */
1344	case 262: /* Reunion Island */
1345	case 269: /* Comoros */
1346	case 270: /* Mayotte */
1347	case 352: /* Luxembourg (or de or ...) */
1348	case 508: /* St. Pierre and Miquelon */
1349	case 509: /* Haiti */
1350	case 590: /* Guadeloupe */
1351	case 594: /* French Guiana */
1352	case 596: /* Martinique / French Antilles */
1353	case 678: /* Vanuatu */
1354	case 681: /* Wallis & Futuna */
1355	case 687: /* New Caledonia */
1356	case 689: /* French Polynesia */
1357	case 961: /* Lebanon */
1358	msg_lang = "fr";
1359	break;
1360	case 3: /* Latin America */
1361	case 34: /* Spain */
1362	case 51: /* Peru */
1363	case 52: /* Mexico */
1364	case 53: /* Cuba */
1365	case 54: /* Argentina */
1366	case 56: /* Chile */
1367	case 57: /* Columbia */
1368	case 58: /* Venezuela */
1369	case 63: /* Philippines */
1370	case 240: /* Equatorial Guinea */
1371	case 502: /* Guatemala */
1372	case 503: /* El Salvador */
1373	case 504: /* Honduras */
1374	case 505: /* Nicraragua */
1375	case 506: /* Costa Rica */
1376	case 507: /* Panama */
1377	case 591: /* Bolivia */
1378	case 593: /* Ecuador */
1379	case 595: /* Paraguay */
1380	case 598: /* Uruguay */
1381	msg_lang = "es";
1382	break;
1383	case 39: /* Italy / San Marino / Vatican City */
1384	msg_lang = "it";
1385	break;
1386	case 41: /* Switzerland / Liechtenstein / / or fr or ... */
1387	msg_lang = "de_CH";
1388	break;
1389	case 43: /* Austria (DR DOS 5.0) */
1390	msg_lang = "de";
1391	break;
1392	case 49: /* Germany */
1393	msg_lang = "de_DE";
1394	break;
1395	case 55: /* Brazil (not supported by DR DOS 5.0) */
1396	msg_lang = "pt_BR";
1397	break;
1398	case 238: /* Cape Verde Islands */
1399	case 244: /* Angola */
1400	case 245: /* Guinea-Bissau */
1401	case 259: /* Mozambique */
1402	case 351: /* Portugal */
1403	msg_lang = "pt";
1404	break;
1405	case 42: /* Czechoslovakia / Tjekia / Slovakia (not supported by DR DOS 5.0) */
1406	case 421: /* Czech Republic / Tjekia (PC DOS 7+) */
1407	case 422: /* Slovakia (reported as 421 due to a bug in COUNTRY.SYS) */
1408	msg_lang = "sk";
1409	break;
1410	case 65: /* Singapore */
1411	case 86: /* China (MS-DOS 5.0+) */
1412	case 88: /* Taiwan (MS-DOS 5.0+) */
1413	case 852: /* Hong Kong */
1414	case 853: /* Macao */
1415	case 886: /* Taiwan (MS-DOS 6.22+) */
1416	msg_lang = "zh";
1417	break;
1418	#if 0
1419	case 7: /* Russia */
1420	case 20: /* Egypt */
1421	case 30: /* Greece */
1422	case 31: /* Netherlands */
1423	case 35: /* Bulgaria??? */
1424	case 36: /* Hungary (not supported by DR DOS 5.0) */
1425	case 38: /* Yugoslavia (not supported by DR DOS 5.0) -- obsolete */
1426	#endif
1427	case 40: /* Romania */
1428	msg_lang = "ro";
1429	break;
1430	#if 0
1431	case 45: /* Denmark */
1432	case 46: /* Sweden */
1433	case 47: /* Norway */
1434	case 48: /* Poland (not supported by DR DOS 5.0) */
1435	case 60: /* Malaysia */
1436	case 62: /* Indonesia / East Timor */
1437	case 66: /* Thailand (or Taiwan??? ) */
1438	case 81: /* Japan (DR DOS 5.0, MS-DOS 5.0+) */
1439	case 82: /* South Korea (DR DOS 5.0) */
1440	case 84: /* Vietnam */
1441	case 90: /* Turkey (MS-DOS 5.0+) */
1442	case 91: /* India */
1443	case 92: /* Pakistan */
1444	case 93: /* Afghanistan */
1445	case 94: /* Sri Lanka */
1446	case 98: /* Iran */
1447	case 102: /* ??? (Hebrew MS-DOS 5.0) */
1448	case 112: /* Belarus */
1449	case 200: /* Thailand (PC DOS 6.1+) (reported as 01due to a bug in PC DOS COUNTRY.SYS) */
1450	case 212: /* Morocco */
1451	case 218: /* Libya */
1452	case 222: /* Maruitania */
1453	case 224: /* African Guinea */
1454	case 234: /* Nigeria */
1455	case 239: /* Sao Tome and Principe */
1456	case 243: /* Zaire */
1457	case 246: /* Diego Garcia */
1458	case 247: /* Ascension Isle */
1459	case 248: /* Seychelles */
1460	case 249: /* Sudan */
1461	case 251: /* Ethiopia */
1462	case 252: /* Somalia */
1463	case 255: /* Tanzania */
1464	case 265: /* Malawi */
1465	case 266: /* Lesotho */
1466	case 298: /* Faroe Islands */
1467	case 299: /* Greenland */
1468	case 354: /* Iceland */
1469	case 355: /* Albania */
1470	case 357: /* Cyprus */
1471	case 358: /* Finland */
1472	case 359: /* Bulgaria */
1473	case 370: /* Lithuania (reported as 372 due to a bug in MS-DOS COUNTRY.SYS) */
1474	case 371: /* Latvia (reported as 372 due to a bug in MS-DOS COUNTRY.SYS) */
1475	case 372: /* Estonia */
1476	case 373: /* Moldova / / FIXME: similar to Romanian? */
1477	case 375: /* ??? (MS-DOS 7.10 / Windows98) */
1478	case 380: /* Ukraine */
1479	case 381: /* Serbia / Montenegro */
1480	case 384: /* Croatia */
1481	case 385: /* Croatia (PC DOS 7+) */
1482	case 386: /* Slovenia */
1483	case 387: /* Bosnia-Herzegovina (Latin) */
1484	case 388: /* Bosnia-Herzegovina (Cyrillic) (PC DOS 7+) (reported as 381 due to a bug in PC DOS COUNTRY.SYS) */
1485	case 389: /* FYR Macedonia */
1486	case 597: /* Suriname (nl) */
1487	case 599: /* Netherland Antilles (nl) */
1488	case 666: /* Russia??? (PTS-DOS 6.51 KEYB) */
1489	case 667: /* Poland??? (PTS-DOS 6.51 KEYB) */
1490	case 668: /* Poland??? (Slavic??? ) (PTS-DOS 6.51 KEYB) */
1491	case 688: /* Tuvalu */
1492	case 690: /* Tokealu */
1493	case 711: /* ??? (currency = EA$, code pages 437,737,850,852,855,857) */
1494	case 785: /* Arabic (Middle East/Saudi Arabia/etc.) */
1495	case 804: /* Ukraine */
1496	case 850: /* North Korea */
1497	case 855: /* Cambodia */
1498	case 856: /* Laos */
1499	case 880: /* Bangladesh */
1500	case 960: /* Maldives */
1501	case 962: /* Jordan */
1502	case 963: /* Syria / Syrian Arab Republic */
1503	case 964: /* Iraq */
1504	case 965: /* Kuwait */
1505	case 966: /* Saudi Arabia */
1506	case 967: /* Yemen */
1507	case 968: /* Oman */
1508	case 969: /* Yemen??? (Arabic MS-DOS 5.0) */
1509	case 971: /* United Arab Emirates */
1510	case 972: /* Israel (Hebrew) (DR DOS 5.0,MS-DOS 5.0+) */
1511	case 973: /* Bahrain */
1512	case 974: /* Qatar */
1513	case 975: /* Bhutan */
1514	case 976: /* Mongolia */
1515	case 977: /* Nepal */
1516	case 995: /* Myanmar (Burma) */
1517	#endif
1518	}
1519	}
1520	}
1521	#endif
1522	}
1523	}
1524	#ifdef DEBUG
1525	fprintf(stderr, "msg_lang = %p (= \"%s\")\n", msg_lang, msg_lang?msg_lang:"(null)");
1526	#endif
1527
1528	/* On Mandrake LANG defaults to C */
1529	if (strcmp(msg_lang, "C") == 0) msg_lang = "en";
1530
1531	msg_lang = osstrdup(msg_lang);
1532
1533	/* Convert en-us to en_US, etc */
1534	p = strchr(msg_lang, '-');
1535	if (p) {
1536	*p++ = '_';
1537	while (*p) {
1538	p = toupper(p);
1539	p++;
1540	}
1541	}
1542
1543	p = strchr(msg_lang, '_');
1544	if (p) {
1545	*p = '\0';
1546	msg_lang2 = osstrdup(msg_lang);
1547	*p = '_';
1548	}
1549
1550	#ifdef LC_MESSAGES
1551	/* try to setlocale() appropriately too */
1552	if (!setlocale(LC_MESSAGES, msg_lang)) {
1553	if (msg_lang2) setlocale(LC_MESSAGES, msg_lang2);
1554	}
1555	#endif
1556
1557	select_charset(default_charset());
1558	}
1559
1560	/* Message may be overwritten by next call
1561	* (but not in current implementation) */
1562	const char *
1563	msg(int en)
1564	{
1565	/* NB can't use SVX_ASSERT here! */
1566	static char badbuf[256];
1567	if (en >= 1000 && en < 1000 + N_DONTEXTRACTMSGS)
1568	return dontextract[en - 1000];
1569	if (!msg_array) {
1570	if (en != 1) {
1571	sprintf(badbuf, "Message %d requested before msg_array initialised\n",
1572	en);
1573	return badbuf;
1574	}
1575	/* this should be the only other message which can be requested before
1576	* the message file is opened and read... */
1577	if (!dontextract) return "Out of memory (couldn't find %lu bytes).";
1578	return dontextract[(/Out of memory (couldn't find %lu bytes)./1004)
1579	- 1000];
1580	}
1581
1582	if (en < 0 \|\| en >= num_msgs) {
1583	sprintf(badbuf, "Message %d out of range\n", en);
1584	return badbuf;
1585	}
1586
1587	if (en == 0) {
1588	const char *p = msg_array[0];
1589	if (!*p) p = "(C)";
1590	return p;
1591	}
1592
1593	return msg_array[en];
1594	}
1595
1596	/* returns persistent copy of message */
1597	const char *
1598	msgPerm(int en)
1599	{
1600	return msg(en);
1601	}
1602
1603	void
1604	v_report(int severity, const char *fnm, int line, int en, va_list ap)
1605	{
1606	#ifdef AVEN
1607	aven_v_report(severity, fnm, line, en, ap);
1608	#else
1609	if (fnm) {
1610	fputs(fnm, STDERR);
1611	if (line) fprintf(STDERR, ":%d", line);
1612	} else {
1613	fputs(appname_copy, STDERR);
1614	}
1615	fputs(": ", STDERR);
1616
1617	if (severity == 0) {
1618	fputs(msg(/warning/4), STDERR);
1619	fputs(": ", STDERR);
1620	}
1621
1622	vfprintf(STDERR, msg(en), ap);
1623	fputnl(STDERR);
1624	#endif
1625
1626	switch (severity) {
1627	case 0:
1628	msg_warnings++;
1629	break;
1630	case 1:
1631	msg_errors++;
1632	if (msg_errors == 50)
1633	fatalerror_in_file(fnm, 0, /Too many errors - giving up/19);
1634	break;
1635	case 2:
1636	exit(EXIT_FAILURE);
1637	}
1638	}
1639
1640	void
1641	warning(int en, ...)
1642	{
1643	va_list ap;
1644	va_start(ap, en);
1645	v_report(0, NULL, 0, en, ap);
1646	va_end(ap);
1647	}
1648
1649	void
1650	error(int en, ...)
1651	{
1652	va_list ap;
1653	va_start(ap, en);
1654	v_report(1, NULL, 0, en, ap);
1655	va_end(ap);
1656	}
1657
1658	void
1659	fatalerror(int en, ...)
1660	{
1661	va_list ap;
1662	va_start(ap, en);
1663	v_report(2, NULL, 0, en, ap);
1664	va_end(ap);
1665	}
1666
1667	void
1668	warning_in_file(const char *fnm, int line, int en, ...)
1669	{
1670	va_list ap;
1671	va_start(ap, en);
1672	v_report(0, fnm, line, en, ap);
1673	va_end(ap);
1674	}
1675
1676	void
1677	error_in_file(const char *fnm, int line, int en, ...)
1678	{
1679	va_list ap;
1680	va_start(ap, en);
1681	v_report(1, fnm, line, en, ap);
1682	va_end(ap);
1683	}
1684
1685	void
1686	fatalerror_in_file(const char *fnm, int line, int en, ...)
1687	{
1688	va_list ap;
1689	va_start(ap, en);
1690	v_report(2, fnm, line, en, ap);
1691	va_end(ap);
1692	}
1693
1694	/* Code to support switching character set at runtime (e.g. for a printer
1695	* driver to support different character sets on screen and on the printer)
1696	*/
1697	typedef struct charset_li {
1698	struct charset_li *next;
1699	int code;
1700	char **msg_array;
1701	} charset_li;
1702
1703	static charset_li *charset_head = NULL;
1704
1705	static int charset = CHARSET_BAD;
1706
1707	int
1708	select_charset(int charset_code)
1709	{
1710	int old_charset = charset;
1711	charset_li *p;
1712
1713	#ifdef DEBUG
1714	fprintf(stderr, "select_charset(%d), old charset = %d\n", charset_code,
1715	charset);
1716	#endif
1717
1718	charset = charset_code;
1719
1720	/* check if we've already parsed messages for new charset */
1721	for (p = charset_head; p; p = p->next) {
1722	#ifdef DEBUG
1723	printf("%p: code %d msg_array %p\n", p, p->code, p->msg_array);
1724	#endif
1725	if (p->code == charset) {
1726	msg_array = p->msg_array;
1727	return old_charset;
1728	}
1729	}
1730
1731	/* nope, got to reparse message file */
1732	parse_msg_file(charset_code);
1733
1734	/* add to list */
1735	p = osnew(charset_li);
1736	p->code = charset;
1737	p->msg_array = msg_array;
1738	p->next = charset_head;
1739	charset_head = p;
1740
1741	return old_charset;
1742	}

Note: See TracBrowser for help on using the repository browser.

Context Navigation

source: git/src/message.c @ 863401f

Download in other formats: