1 | /* Handle aliases for locale names. |
2 | Copyright (C) 1995-2017 Free Software Foundation, Inc. |
3 | |
4 | This program is free software: you can redistribute it and/or modify |
5 | it under the terms of the GNU Lesser General Public License as published by |
6 | the Free Software Foundation; either version 2.1 of the License, or |
7 | (at your option) any later version. |
8 | |
9 | This program is distributed in the hope that it will be useful, |
10 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
11 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
12 | GNU Lesser General Public License for more details. |
13 | |
14 | You should have received a copy of the GNU Lesser General Public License |
15 | along with this program. If not, see <http://www.gnu.org/licenses/>. */ |
16 | |
17 | /* Tell glibc's <string.h> to provide a prototype for mempcpy(). |
18 | This must come before <config.h> because <config.h> may include |
19 | <features.h>, and once <features.h> has been included, it's too late. */ |
20 | #ifndef _GNU_SOURCE |
21 | # define _GNU_SOURCE 1 |
22 | #endif |
23 | |
24 | #ifdef HAVE_CONFIG_H |
25 | # include <config.h> |
26 | #endif |
27 | |
28 | #include <ctype.h> |
29 | #include <stdio.h> |
30 | #if defined _LIBC || defined HAVE___FSETLOCKING |
31 | # include <stdio_ext.h> |
32 | #endif |
33 | #include <sys/types.h> |
34 | |
35 | #ifdef __GNUC__ |
36 | # undef alloca |
37 | # define alloca __builtin_alloca |
38 | # define HAVE_ALLOCA 1 |
39 | #else |
40 | # ifdef _MSC_VER |
41 | # include <malloc.h> |
42 | # define alloca _alloca |
43 | # else |
44 | # if defined HAVE_ALLOCA_H || defined _LIBC |
45 | # include <alloca.h> |
46 | # else |
47 | # ifdef _AIX |
48 | #pragma alloca |
49 | # else |
50 | # ifndef alloca |
51 | char *alloca (); |
52 | # endif |
53 | # endif |
54 | # endif |
55 | # endif |
56 | #endif |
57 | |
58 | #include <stdlib.h> |
59 | #include <string.h> |
60 | |
61 | #include "gettextP.h" |
62 | |
63 | #ifdef ENABLE_RELOCATABLE |
64 | # include "relocatable.h" |
65 | #else |
66 | # define relocate(pathname) (pathname) |
67 | #endif |
68 | |
69 | /* @@ end of prolog @@ */ |
70 | |
71 | #ifdef _LIBC |
72 | /* Rename the non ANSI C functions. This is required by the standard |
73 | because some ANSI C functions will require linking with this object |
74 | file and the name space must not be polluted. */ |
75 | # define strcasecmp(s1, s2) __strcasecmp_l (s1, s2, _nl_C_locobj_ptr) |
76 | |
77 | # ifndef mempcpy |
78 | # define mempcpy __mempcpy |
79 | # endif |
80 | # define HAVE_MEMPCPY 1 |
81 | # define HAVE___FSETLOCKING 1 |
82 | #endif |
83 | |
84 | /* Handle multi-threaded applications. */ |
85 | #ifdef _LIBC |
86 | # include <libc-lock.h> |
87 | #else |
88 | # include "lock.h" |
89 | #endif |
90 | |
91 | #ifndef internal_function |
92 | # define internal_function |
93 | #endif |
94 | |
95 | /* Some optimizations for glibc. */ |
96 | #ifdef _LIBC |
97 | # define FEOF(fp) feof_unlocked (fp) |
98 | # define FGETS(buf, n, fp) __fgets_unlocked (buf, n, fp) |
99 | #else |
100 | # define FEOF(fp) feof (fp) |
101 | # define FGETS(buf, n, fp) fgets (buf, n, fp) |
102 | #endif |
103 | |
104 | /* For those losing systems which don't have `alloca' we have to add |
105 | some additional code emulating it. */ |
106 | #ifdef HAVE_ALLOCA |
107 | # define freea(p) /* nothing */ |
108 | #else |
109 | # define alloca(n) malloc (n) |
110 | # define freea(p) free (p) |
111 | #endif |
112 | |
113 | #if defined _LIBC_REENTRANT || defined HAVE_DECL_FGETS_UNLOCKED |
114 | # undef fgets |
115 | # define fgets(buf, len, s) fgets_unlocked (buf, len, s) |
116 | #endif |
117 | #if defined _LIBC_REENTRANT || defined HAVE_DECL_FEOF_UNLOCKED |
118 | # undef feof |
119 | # define feof(s) feof_unlocked (s) |
120 | #endif |
121 | |
122 | |
123 | __libc_lock_define_initialized (static, lock) |
124 | |
125 | |
126 | struct alias_map |
127 | { |
128 | const char *alias; |
129 | const char *value; |
130 | }; |
131 | |
132 | |
133 | #ifndef _LIBC |
134 | # define libc_freeres_ptr(decl) decl |
135 | #endif |
136 | |
137 | libc_freeres_ptr (static char *string_space); |
138 | static size_t string_space_act; |
139 | static size_t string_space_max; |
140 | libc_freeres_ptr (static struct alias_map *map); |
141 | static size_t nmap; |
142 | static size_t maxmap; |
143 | |
144 | |
145 | /* Prototypes for local functions. */ |
146 | static size_t read_alias_file (const char *fname, int fname_len) |
147 | internal_function; |
148 | static int extend_alias_table (void); |
149 | static int alias_compare (const struct alias_map *map1, |
150 | const struct alias_map *map2); |
151 | |
152 | |
153 | const char * |
154 | _nl_expand_alias (const char *name) |
155 | { |
156 | static const char *locale_alias_path; |
157 | struct alias_map *retval; |
158 | const char *result = NULL; |
159 | size_t added; |
160 | |
161 | __libc_lock_lock (lock); |
162 | |
163 | if (locale_alias_path == NULL) |
164 | locale_alias_path = LOCALE_ALIAS_PATH; |
165 | |
166 | do |
167 | { |
168 | struct alias_map item; |
169 | |
170 | item.alias = name; |
171 | |
172 | if (nmap > 0) |
173 | retval = (struct alias_map *) bsearch (&item, map, nmap, |
174 | sizeof (struct alias_map), |
175 | (int (*) (const void *, |
176 | const void *) |
177 | ) alias_compare); |
178 | else |
179 | retval = NULL; |
180 | |
181 | /* We really found an alias. Return the value. */ |
182 | if (retval != NULL) |
183 | { |
184 | result = retval->value; |
185 | break; |
186 | } |
187 | |
188 | /* Perhaps we can find another alias file. */ |
189 | added = 0; |
190 | while (added == 0 && locale_alias_path[0] != '\0') |
191 | { |
192 | const char *start; |
193 | |
194 | while (locale_alias_path[0] == PATH_SEPARATOR) |
195 | ++locale_alias_path; |
196 | start = locale_alias_path; |
197 | |
198 | while (locale_alias_path[0] != '\0' |
199 | && locale_alias_path[0] != PATH_SEPARATOR) |
200 | ++locale_alias_path; |
201 | |
202 | if (start < locale_alias_path) |
203 | added = read_alias_file (start, locale_alias_path - start); |
204 | } |
205 | } |
206 | while (added != 0); |
207 | |
208 | __libc_lock_unlock (lock); |
209 | |
210 | return result; |
211 | } |
212 | |
213 | |
214 | static size_t |
215 | internal_function |
216 | read_alias_file (const char *fname, int fname_len) |
217 | { |
218 | FILE *fp; |
219 | char *full_fname; |
220 | size_t added; |
221 | static const char aliasfile[] = "/locale.alias" ; |
222 | |
223 | full_fname = (char *) alloca (fname_len + sizeof aliasfile); |
224 | #ifdef HAVE_MEMPCPY |
225 | mempcpy (mempcpy (full_fname, fname, fname_len), |
226 | aliasfile, sizeof aliasfile); |
227 | #else |
228 | memcpy (full_fname, fname, fname_len); |
229 | memcpy (&full_fname[fname_len], aliasfile, sizeof aliasfile); |
230 | #endif |
231 | |
232 | #ifdef _LIBC |
233 | /* Note the file is opened with cancellation in the I/O functions |
234 | disabled. */ |
235 | fp = fopen (relocate (full_fname), "rce" ); |
236 | #else |
237 | fp = fopen (relocate (full_fname), "r" ); |
238 | #endif |
239 | freea (full_fname); |
240 | if (fp == NULL) |
241 | return 0; |
242 | |
243 | #ifdef HAVE___FSETLOCKING |
244 | /* No threads present. */ |
245 | __fsetlocking (fp, FSETLOCKING_BYCALLER); |
246 | #endif |
247 | |
248 | added = 0; |
249 | while (!FEOF (fp)) |
250 | { |
251 | /* It is a reasonable approach to use a fix buffer here because |
252 | a) we are only interested in the first two fields |
253 | b) these fields must be usable as file names and so must not |
254 | be that long |
255 | We avoid a multi-kilobyte buffer here since this would use up |
256 | stack space which we might not have if the program ran out of |
257 | memory. */ |
258 | char buf[400]; |
259 | char *alias; |
260 | char *value; |
261 | char *cp; |
262 | int complete_line; |
263 | |
264 | if (FGETS (buf, sizeof buf, fp) == NULL) |
265 | /* EOF reached. */ |
266 | break; |
267 | |
268 | /* Determine whether the line is complete. */ |
269 | complete_line = strchr (buf, '\n') != NULL; |
270 | |
271 | cp = buf; |
272 | /* Ignore leading white space. */ |
273 | while (isspace ((unsigned char) cp[0])) |
274 | ++cp; |
275 | |
276 | /* A leading '#' signals a comment line. */ |
277 | if (cp[0] != '\0' && cp[0] != '#') |
278 | { |
279 | alias = cp++; |
280 | while (cp[0] != '\0' && !isspace ((unsigned char) cp[0])) |
281 | ++cp; |
282 | /* Terminate alias name. */ |
283 | if (cp[0] != '\0') |
284 | *cp++ = '\0'; |
285 | |
286 | /* Now look for the beginning of the value. */ |
287 | while (isspace ((unsigned char) cp[0])) |
288 | ++cp; |
289 | |
290 | if (cp[0] != '\0') |
291 | { |
292 | value = cp++; |
293 | while (cp[0] != '\0' && !isspace ((unsigned char) cp[0])) |
294 | ++cp; |
295 | /* Terminate value. */ |
296 | if (cp[0] == '\n') |
297 | { |
298 | /* This has to be done to make the following test |
299 | for the end of line possible. We are looking for |
300 | the terminating '\n' which do not overwrite here. */ |
301 | *cp++ = '\0'; |
302 | *cp = '\n'; |
303 | } |
304 | else if (cp[0] != '\0') |
305 | *cp++ = '\0'; |
306 | |
307 | #ifdef IN_LIBGLOCALE |
308 | /* glibc's locale.alias contains entries for ja_JP and ko_KR |
309 | that make it impossible to use a Japanese or Korean UTF-8 |
310 | locale under the name "ja_JP" or "ko_KR". Ignore these |
311 | entries. */ |
312 | if (strchr (alias, '_') == NULL) |
313 | #endif |
314 | { |
315 | size_t alias_len; |
316 | size_t value_len; |
317 | |
318 | if (nmap >= maxmap) |
319 | if (__builtin_expect (extend_alias_table (), 0)) |
320 | goto out; |
321 | |
322 | alias_len = strlen (alias) + 1; |
323 | value_len = strlen (value) + 1; |
324 | |
325 | if (string_space_act + alias_len + value_len > string_space_max) |
326 | { |
327 | /* Increase size of memory pool. */ |
328 | size_t new_size = (string_space_max |
329 | + (alias_len + value_len > 1024 |
330 | ? alias_len + value_len : 1024)); |
331 | char *new_pool = (char *) realloc (string_space, new_size); |
332 | if (new_pool == NULL) |
333 | goto out; |
334 | |
335 | if (__builtin_expect (string_space != new_pool, 0)) |
336 | { |
337 | size_t i; |
338 | |
339 | for (i = 0; i < nmap; i++) |
340 | { |
341 | map[i].alias += new_pool - string_space; |
342 | map[i].value += new_pool - string_space; |
343 | } |
344 | } |
345 | |
346 | string_space = new_pool; |
347 | string_space_max = new_size; |
348 | } |
349 | |
350 | map[nmap].alias = |
351 | (const char *) memcpy (&string_space[string_space_act], |
352 | alias, alias_len); |
353 | string_space_act += alias_len; |
354 | |
355 | map[nmap].value = |
356 | (const char *) memcpy (&string_space[string_space_act], |
357 | value, value_len); |
358 | string_space_act += value_len; |
359 | |
360 | ++nmap; |
361 | ++added; |
362 | } |
363 | } |
364 | } |
365 | |
366 | /* Possibly not the whole line fits into the buffer. Ignore |
367 | the rest of the line. */ |
368 | if (! complete_line) |
369 | do |
370 | if (FGETS (buf, sizeof buf, fp) == NULL) |
371 | /* Make sure the inner loop will be left. The outer loop |
372 | will exit at the `feof' test. */ |
373 | break; |
374 | while (strchr (buf, '\n') == NULL); |
375 | } |
376 | |
377 | out: |
378 | /* Should we test for ferror()? I think we have to silently ignore |
379 | errors. --drepper */ |
380 | fclose (fp); |
381 | |
382 | if (added > 0) |
383 | qsort (map, nmap, sizeof (struct alias_map), |
384 | (int (*) (const void *, const void *)) alias_compare); |
385 | |
386 | return added; |
387 | } |
388 | |
389 | |
390 | static int |
391 | extend_alias_table (void) |
392 | { |
393 | size_t new_size; |
394 | struct alias_map *new_map; |
395 | |
396 | new_size = maxmap == 0 ? 100 : 2 * maxmap; |
397 | new_map = (struct alias_map *) realloc (map, (new_size |
398 | * sizeof (struct alias_map))); |
399 | if (new_map == NULL) |
400 | /* Simply don't extend: we don't have any more core. */ |
401 | return -1; |
402 | |
403 | map = new_map; |
404 | maxmap = new_size; |
405 | return 0; |
406 | } |
407 | |
408 | |
409 | static int |
410 | alias_compare (const struct alias_map *map1, const struct alias_map *map2) |
411 | { |
412 | #if defined _LIBC || defined HAVE_STRCASECMP |
413 | return strcasecmp (map1->alias, map2->alias); |
414 | #else |
415 | const unsigned char *p1 = (const unsigned char *) map1->alias; |
416 | const unsigned char *p2 = (const unsigned char *) map2->alias; |
417 | unsigned char c1, c2; |
418 | |
419 | if (p1 == p2) |
420 | return 0; |
421 | |
422 | do |
423 | { |
424 | /* I know this seems to be odd but the tolower() function in |
425 | some systems libc cannot handle nonalpha characters. */ |
426 | c1 = isupper (*p1) ? tolower (*p1) : *p1; |
427 | c2 = isupper (*p2) ? tolower (*p2) : *p2; |
428 | if (c1 == '\0') |
429 | break; |
430 | ++p1; |
431 | ++p2; |
432 | } |
433 | while (c1 == c2); |
434 | |
435 | return c1 - c2; |
436 | #endif |
437 | } |
438 | |