]>
git.ipfire.org Git - thirdparty/glibc.git/blob - locale/programs/repertoire.c
1 /* Copyright (C) 1998, 1999 Free Software Foundation, Inc.
2 This file is part of the GNU C Library.
3 Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
5 The GNU C Library is free software; you can redistribute it and/or
6 modify it under the terms of the GNU Library General Public License as
7 published by the Free Software Foundation; either version 2 of the
8 License, or (at your option) any later version.
10 The GNU C Library 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 GNU
13 Library General Public License for more details.
15 You should have received a copy of the GNU Library General Public
16 License along with the GNU C Library; see the file COPYING.LIB. If not,
17 write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
18 Boston, MA 02111-1307, USA. */
33 #include "linereader.h"
35 #include "repertoire.h"
36 #include "simple-hash.h"
37 #include "localedef.h"
40 /* Simple keyword hashing for the repertoiremap. */
41 static const struct keyword_t
*repertoiremap_hash (const char *str
,
43 static void repertoire_new_char (struct linereader
*lr
, hash_table
*ht
,
44 hash_table
*rt
, struct obstack
*ob
,
45 uint32_t value
, const char *from
,
46 const char *to
, int decimal_ellipsis
);
47 static int repertoire_compare (const void *p1
, const void *p2
);
49 /* Already known repertoire maps. */
54 repertoire_read (const char *filename
)
56 struct linereader
*repfile
;
57 struct repertoire_t
*result
;
58 struct repertoire_t
**resultp
;
59 struct repertoire_t search
;
61 char *from_name
= NULL
;
63 enum token_t ellipsis
= tok_none
;
65 search
.name
= filename
;
66 resultp
= tfind (&search
, &known
, &repertoire_compare
);
71 repfile
= lr_open (filename
, repertoiremap_hash
);
74 if (strchr (filename
, '/') == NULL
)
76 char *i18npath
= getenv ("I18NPATH");
77 if (i18npath
!= NULL
&& *i18npath
!= '\0')
79 char path
[strlen (filename
) + 1 + strlen (i18npath
)
80 + sizeof ("/repertoiremaps/") - 1];
82 i18npath
= strdupa (i18npath
);
85 while (repfile
== NULL
86 && (next
= strsep (&i18npath
, ":")) != NULL
)
88 stpcpy (stpcpy (stpcpy (path
, next
), "/repertoiremaps/"),
91 repfile
= lr_open (path
, repertoiremap_hash
);
95 stpcpy (stpcpy (path
, next
), filename
);
97 repfile
= lr_open (path
, repertoiremap_hash
);
104 /* Look in the systems charmap directory. */
105 char *buf
= xmalloc (strlen (filename
) + 1
106 + sizeof (REPERTOIREMAP_PATH
));
108 stpcpy (stpcpy (stpcpy (buf
, REPERTOIREMAP_PATH
), "/"),
110 repfile
= lr_open (buf
, repertoiremap_hash
);
119 error (0, errno
, _("repertoire map file `%s' not found"), filename
);
124 /* We don't want symbolic names in string to be translated. */
125 repfile
->translate_strings
= 0;
127 /* Allocate room for result. */
128 result
= (struct repertoire_t
*) xmalloc (sizeof (struct repertoire_t
));
129 memset (result
, '\0', sizeof (struct repertoire_t
));
131 result
->name
= xstrdup (filename
);
133 #define obstack_chunk_alloc malloc
134 #define obstack_chunk_free free
135 obstack_init (&result
->mem_pool
);
137 if (init_hash (&result
->char_table
, 256)
138 || init_hash (&result
->reverse_table
, 256)
139 || init_hash (&result
->seq_table
, 256))
145 /* We use a state machine to describe the charmap description file
151 struct token
*now
= lr_token (repfile
, NULL
, NULL
);
152 enum token_t nowtok
= now
->tok
;
155 if (nowtok
== tok_eof
)
161 /* We haven't yet read any character definition. This is where
162 we accept escape_char and comment_char definitions. */
163 if (nowtok
== tok_eol
)
164 /* Ignore empty lines. */
167 if (nowtok
== tok_escape_char
|| nowtok
== tok_comment_char
)
169 /* We know that we need an argument. */
170 arg
= lr_token (repfile
, NULL
, NULL
);
172 if (arg
->tok
!= tok_ident
)
174 lr_error (repfile
, _("syntax error in prolog: %s"),
177 lr_ignore_rest (repfile
, 0);
181 if (arg
->val
.str
.lenmb
!= 1)
183 lr_error (repfile
, _("\
184 argument to <%s> must be a single character"),
185 nowtok
== tok_escape_char
? "escape_char"
188 lr_ignore_rest (repfile
, 0);
192 if (nowtok
== tok_escape_char
)
193 repfile
->escape_char
= *arg
->val
.str
.startmb
;
195 repfile
->comment_char
= *arg
->val
.str
.startmb
;
197 lr_ignore_rest (repfile
, 1);
201 if (nowtok
== tok_charids
)
203 lr_ignore_rest (repfile
, 1);
209 /* Otherwise we start reading the character definitions. */
214 /* We are now are in the body. Each line
215 must have the format "%s %s %s\n" or "%s...%s %s %s\n". */
216 if (nowtok
== tok_eol
)
217 /* Ignore empty lines. */
220 if (nowtok
== tok_end
)
226 if (nowtok
!= tok_bsymbol
)
229 _("syntax error in repertoire map definition: %s"),
230 _("no symbolic name given"));
232 lr_ignore_rest (repfile
, 0);
236 /* If the previous line was not completely correct free the
238 if (from_name
!= NULL
)
239 obstack_free (&result
->mem_pool
, from_name
);
241 from_name
= (char *) obstack_copy0 (&result
->mem_pool
,
242 now
->val
.str
.startmb
,
250 /* We have two possibilities: We can see an ellipsis or an
252 if (nowtok
== tok_ellipsis3
|| nowtok
== tok_ellipsis4
253 || nowtok
== tok_ellipsis2
)
262 /* We expect a value of the form <Uxxxx> or <Uxxxxxxxx> where
263 the xxx mean a hexadecimal value. */
267 if (nowtok
!= tok_ucs4
)
270 _("syntax error in repertoire map definition: %s"),
271 _("no <Uxxxx> or <Uxxxxxxxx> value given"));
273 lr_ignore_rest (repfile
, 0);
277 /* We've found a new valid definition. */
278 repertoire_new_char (repfile
, &result
->char_table
,
279 &result
->reverse_table
, &result
->mem_pool
,
280 now
->val
.ucs4
, from_name
, to_name
,
281 ellipsis
!= tok_ellipsis2
);
283 /* Ignore the rest of the line. */
284 lr_ignore_rest (repfile
, 0);
292 if (nowtok
!= tok_bsymbol
)
295 _("syntax error in repertoire map definition: %s"),
296 _("no symbolic name given for end of range"));
298 lr_ignore_rest (repfile
, 0);
303 /* Copy the to-name in a safe place. */
304 to_name
= (char *) obstack_copy0 (&result
->mem_pool
,
305 repfile
->token
.val
.str
.startmb
,
306 repfile
->token
.val
.str
.lenmb
);
312 if (nowtok
!= tok_charids
)
313 lr_error (repfile
, _("\
314 `%1$s' definition does not end with `END %1$s'"), "CHARIDS");
316 lr_ignore_rest (repfile
, nowtok
== tok_charids
);
323 if (state
!= 2 && state
!= 90 && !be_quiet
)
324 error (0, 0, _("%s: premature end of file"), repfile
->fname
);
328 if (tsearch (result
, &known
, &repertoire_compare
) == NULL
)
329 /* Something went wrong. */
330 error (0, errno
, _("cannot safe new repertoire map"));
337 repertoire_compare (const void *p1
, const void *p2
)
339 struct repertoire_t
*r1
= (struct repertoire_t
*) p1
;
340 struct repertoire_t
*r2
= (struct repertoire_t
*) p2
;
342 return strcmp (r1
->name
, r2
->name
);
346 static const struct keyword_t
*
347 repertoiremap_hash (const char *str
, unsigned int len
)
349 static const struct keyword_t wordlist
[0] =
351 {"escape_char", tok_escape_char
, 0},
352 {"comment_char", tok_comment_char
, 0},
353 {"CHARIDS", tok_charids
, 0},
357 if (len
== 11 && memcmp (wordlist
[0].name
, str
, 11) == 0)
359 if (len
== 12 && memcmp (wordlist
[1].name
, str
, 12) == 0)
361 if (len
== 7 && memcmp (wordlist
[2].name
, str
, 7) == 0)
363 if (len
== 3 && memcmp (wordlist
[3].name
, str
, 3) == 0)
371 repertoire_new_char (struct linereader
*lr
, hash_table
*ht
, hash_table
*rt
,
372 struct obstack
*ob
, uint32_t value
, const char *from
,
373 const char *to
, int decimal_ellipsis
)
379 int prefix_len
, len1
, len2
;
380 unsigned int from_nr
, to_nr
, cnt
;
384 insert_entry (ht
, from
, strlen (from
),
385 (void *) (unsigned long int) value
);
386 /* Please note that it isn't a bug if a symbol is defined more
387 than once. All later definitions are simply discarded. */
389 insert_entry (rt
, obstack_copy (ob
, &value
, sizeof (value
)),
390 sizeof (value
), (void *) from
);
395 /* We have a range: the names must have names with equal prefixes
396 and an equal number of digits, where the second number is greater
397 or equal than the first. */
398 len1
= strlen (from
);
404 lr_error (lr
, _("invalid names for character range"));
408 cp
= &from
[len1
- 1];
409 if (decimal_ellipsis
)
410 while (isdigit (*cp
) && cp
>= from
)
413 while (isxdigit (*cp
) && cp
>= from
)
415 if (!isdigit (*cp
) && !isupper (*cp
))
417 hexadecimal range format should use only capital characters"));
421 prefix_len
= (cp
- from
) + 1;
423 if (cp
== &from
[len1
- 1] || strncmp (from
, to
, prefix_len
) != 0)
427 from_nr
= strtoul (&from
[prefix_len
], &from_end
, decimal_ellipsis
? 10 : 16);
428 if (*from_end
!= '\0' || (from_nr
== ULONG_MAX
&& errno
== ERANGE
)
429 || ((to_nr
= strtoul (&to
[prefix_len
], &to_end
,
430 decimal_ellipsis
? 10 : 16)) == ULONG_MAX
434 lr_error (lr
, _("<%s> and <%s> are invalid names for range"));
440 lr_error (lr
, _("upper limit in range is not smaller then lower limit"));
444 for (cnt
= from_nr
; cnt
<= to_nr
; ++cnt
)
446 uint32_t this_value
= value
+ (cnt
- from_nr
);
448 obstack_printf (ob
, decimal_ellipsis
? "%.*s%0*d" : "%.*s%0*X",
449 prefix_len
, from
, len1
- prefix_len
, cnt
);
451 insert_entry (ht
, buf
, len1
,
452 (void *) (unsigned long int) this_value
);
453 /* Please note we don't examine the return value since it is no error
454 if we have two definitions for a symbol. */
456 insert_entry (rt
, obstack_copy (ob
, &this_value
, sizeof (this_value
)),
457 sizeof (this_value
), (void *) from
);
463 repertoire_find_value (const struct repertoire_t
*rep
, const char *name
,
468 if (find_entry ((hash_table
*) &rep
->char_table
, name
, len
, &result
) < 0)
469 return ILLEGAL_CHAR_VALUE
;
471 return (uint32_t) ((unsigned long int) result
);
476 repertoire_find_symbol (const struct repertoire_t
*rep
, uint32_t ucs
)
480 if (find_entry ((hash_table
*) &rep
->reverse_table
, &ucs
, sizeof (ucs
),
484 return (const char *) result
;
489 repertoire_find_seq (const struct repertoire_t
*rep
, uint32_t ucs
)
493 if (find_entry ((hash_table
*) &rep
->seq_table
, &ucs
, sizeof (ucs
),
497 return (struct charseq
*) result
;