1 /* FriBidi
2 * gen-bidi-type-tab.c - generate bidi-type.tab.i
3 *
4 * Author:
5 * Behdad Esfahbod, 2001, 2002, 2004
6 *
7 * Copyright (C) 2004 Sharif FarsiWeb, Inc
8 * Copyright (C) 2001,2002,2004 Behdad Esfahbod
9 *
10 * This library is free software; you can redistribute it and/or
11 * modify it under the terms of the GNU Lesser General Public
12 * License as published by the Free Software Foundation; either
13 * version 2.1 of the License, or (at your option) any later version.
14 *
15 * This library is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
18 * Lesser General Public License for more details.
19 *
20 * You should have received a copy of the GNU Lesser General Public License
21 * along with this library, in a file named COPYING; if not, write to the
22 * Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
23 * Boston, MA 02110-1301, USA
24 *
25 * For licensing issues, contact <fribidi.license@gmail.com>.
26 */
27
28 #include <common.h>
29
30 #include <fribidi-unicode.h>
31
32 #include <stdio.h>
33 #ifdef HAVE_CONFIG_H
34 # include <config.h>
35 #endif
36
37 #ifdef STDC_HEADERS
38 # include <stdlib.h>
39 # include <stddef.h>
40 #else
41 # if HAVE_STDLIB_H
42 # include <stdlib.h>
43 # endif
44 #endif
45 #ifdef HAVE_STRING_H
46 # if !STDC_HEADERS && HAVE_MEMORY_H
47 # include <memory.h>
48 # endif
49 # include <string.h>
50 #endif
51 #ifdef HAVE_STRINGS_H
52 # include <strings.h>
53 #endif
54
55 #include "packtab.h"
56
57 #define appname "gen-bidi-type-tab"
58 #define outputname "bidi-type.tab.i"
59
60 static void
61 die (
62 const char *msg
63 )
64 {
65 fprintf (stderr, appname ": %s\n", msg);
66 exit (1);
67 }
68
69 static void
70 die2 (
71 const char *fmt,
72 const char *p
73 )
74 {
75 fprintf (stderr, appname ": ");
76 fprintf (stderr, fmt, p);
77 fprintf (stderr, "\n");
78 exit (1);
79 }
80
81 static void
82 die3 (
83 const char *fmt,
84 unsigned long l,
85 const char *p
86 )
87 {
88 fprintf (stderr, appname ": ");
89 fprintf (stderr, fmt, l, p);
90 fprintf (stderr, "\n");
91 exit (1);
92 }
93
94 enum FriBidiCharTypeLinearEnumOffsetOne
95 {
96 # define _FRIBIDI_ADD_TYPE(TYPE,SYMBOL) TYPE,
97 # define _FRIBIDI_ADD_ALIAS(TYPE1,TYPE2) TYPE1 = TYPE2,
98 # include <fribidi-bidi-types-list.h>
99 # undef _FRIBIDI_ADD_ALIAS
100 # undef _FRIBIDI_ADD_TYPE
101 NUM_TYPES
102 };
103
104 struct
105 {
106 const char *name;
107 int key;
108 }
109 type_names[] =
110 {
111 # define _FRIBIDI_ADD_TYPE(TYPE,SYMBOL) {STRINGIZE(TYPE), TYPE},
112 # define _FRIBIDI_ADD_ALIAS(TYPE1,TYPE2) _FRIBIDI_ADD_TYPE(TYPE1, SYMBOL)
113 # include <fribidi-bidi-types-list.h>
114 # undef _FRIBIDI_ADD_ALIAS
115 # undef _FRIBIDI_ADD_TYPE
116 };
117
118 #define type_names_count (sizeof (type_names) / sizeof (type_names[0]))
119
120 static const char *names[type_names_count];
121
122 static char
123 get_type (
124 const char *s
125 )
126 {
127 unsigned int i;
128
129 for (i = 0; i < type_names_count; i++)
130 if (!strcmp (s, type_names[i].name))
131 return type_names[i].key;
132 die2 ("bidi type name `%s' not found", s);
133 return 0;
134 }
135
136 #define table_name "Bid"
137 #define macro_name "FRIBIDI_GET_BIDI_TYPE"
138
139 static signed int table[FRIBIDI_UNICODE_CHARS];
140 static char buf[4000];
141 static char tp[sizeof (buf)];
142
143 static void
144 init (
145 void
146 )
147 {
148 register int i;
149
150 for (i = 0; i < type_names_count; i++)
151 names[i] = 0;
152 for (i = type_names_count - 1; i >= 0; i--)
153 names[type_names[i].key] = type_names[i].name;
154 }
155
156 static void
157 clear_tab (
158 void
159 )
160 {
161 register FriBidiChar c;
162
163 for (c = 0; c < FRIBIDI_UNICODE_CHARS; c++)
164 table[c] = LTR;
165 }
166
167 static void
168 init_tab_unicode_data_txt (
169 void
170 )
171 {
172 register FriBidiChar c;
173
174 clear_tab ();
175
176 for (c = 0x0590; c < 0x0600; c++)
177 table[c] = RTL;
178 for (c = 0x07C0; c < 0x0900; c++)
179 table[c] = RTL;
180 for (c = 0xFB1D; c < 0xFB50; c++)
181 table[c] = RTL;
182
183 for (c = 0x0600; c < 0x07C0; c++)
184 table[c] = AL;
185 for (c = 0xFB50; c < 0xFDD0; c++)
186 table[c] = AL;
187 for (c = 0xFDF0; c < 0xFE00; c++)
188 table[c] = AL;
189 for (c = 0xFE70; c < 0xFF00; c++)
190 table[c] = AL;
191
192 for (c = 0x2060; c < 0x2070; c++)
193 table[c] = BN;
194 for (c = 0xFDD0; c < 0xFDF0; c++)
195 table[c] = BN;
196 for (c = 0xFFF0; c < 0xFFF9; c++)
197 table[c] = BN;
198 for (c = 0xFFFF; c < FRIBIDI_UNICODE_CHARS; c += 0x10000)
199 table[c - 1] = table[c] = BN;
200
201 if (FRIBIDI_UNICODE_CHARS > 0x10000)
202 {
203 for (c = 0x10800; c < 0x11000; c++)
204 table[c] = RTL;
205
206 for (c = 0xE0000; c < 0xE1000; c++)
207 table[c] = BN;
208 }
209 }
210
211 static void
212 init_tab_derived_bidi_class_txt (
213 void
214 )
215 {
216 clear_tab ();
217 }
218
219 static void
220 read_unicode_data_txt (
221 FILE *f
222 )
223 {
224 unsigned long c, l;
225
226 init_tab_unicode_data_txt ();
227
228 l = 0;
229 while (fgets (buf, sizeof buf, f))
230 {
231 int i;
232 const char *s = buf;
233
234 l++;
235
236 while (*s == ' ')
237 s++;
238
239 if (s[0] == '#' || s[0] == '\0' || s[0] == '\n')
240 continue;
241
242 i = sscanf (s, "%lx;%*[^;];%*[^;];%*[^;];%[^; ]", &c, tp);
243 if (i != 2 || c >= FRIBIDI_UNICODE_CHARS)
244 die3 ("invalid input at line %ld: %s", l, s);
245
246 /* FIXME: We don't handle First..Last parts of UnicodeData.txt,
247 * but it works, since all those are LTR. */
248 table[c] = get_type (tp);
249 }
250 }
251
252 static void
253 read_derived_bidi_class_txt (
254 FILE *f
255 )
256 {
257 unsigned long c, c2, l;
258
259 init_tab_derived_bidi_class_txt ();
260
261 l = 0;
262 while (fgets (buf, sizeof buf, f))
263 {
264 int i;
265 register char typ;
266 const char *s = buf;
267
268 l++;
269
270 while (*s == ' ')
271 s++;
272
273 if (s[0] == '#' || s[0] == '\0' || s[0] == '\n')
274 continue;
275
276 i = sscanf (s, "%lx ; %[^; ]", &c, tp);
277 if (i == 2)
278 c2 = c;
279 else
280 i = sscanf (s, "%lx..%lx ; %[^; ]", &c, &c2, tp) - 1;
281
282 if (i != 2 || c > c2 || c2 >= FRIBIDI_UNICODE_CHARS)
283 die3 ("invalid input at line %ld: %s", l, s);
284
285 typ = get_type (tp);
286 for (; c <= c2; c++)
287 table[c] = typ;
288 }
289 }
290
291 static void
292 read_data (
293 const char *data_file_type,
294 const char *data_file_name
295 )
296 {
297 FILE *f;
298
299 if (!(f = fopen (data_file_name, "rt")))
300 die2 ("error: cannot open `%s' for reading", data_file_name);
301
302 if (!strcmp (data_file_type, "UnicodeData.txt"))
303 read_unicode_data_txt (f);
304 else if (!strcmp (data_file_type, "DerivedBidiClass.txt"))
305 read_derived_bidi_class_txt (f);
306 else
307 die2 ("error: unknown data-file type %s", data_file_type);
308
309 fclose (f);
310 }
311
312 static void
313 gen_bidi_type_tab (
314 int max_depth,
315 const char *data_file_type
316 )
317 {
318 printf ("/* " outputname "\n * generated by " appname " (" FRIBIDI_NAME " "
319 FRIBIDI_VERSION ")\n" " * from the file %s of Unicode version "
320 FRIBIDI_UNICODE_VERSION ". */\n\n", data_file_type);
321
322 printf ("#define PACKTAB_UINT8 uint8_t\n"
323 "#define PACKTAB_UINT16 uint16_t\n"
324 "#define PACKTAB_UINT32 uint32_t\n\n");
325
326 if (!pack_table
327 (table, FRIBIDI_UNICODE_CHARS, 1, LTR, max_depth, 3, names,
328 "unsigned char", table_name, macro_name, stdout))
329 die ("error: insufficient memory, decrease max_depth");
330
331 printf ("#undef PACKTAB_UINT8\n"
332 "#undef PACKTAB_UINT16\n" "#undef PACKTAB_UINT32\n\n");
333
334 printf ("/* End of generated " outputname " */\n");
335 }
336
337 int
338 main (
339 int argc,
340 const char **argv
341 )
342 {
343 const char *data_file_type = "UnicodeData.txt";
344
345 if (argc < 3)
346 die2 ("usage:\n " appname " max-depth /path/to/%s [junk...]",
347 data_file_type);
348
349 {
350 int max_depth = atoi (argv[1]);
351 const char *data_file_name = argv[2];
352
353 if (max_depth < 2)
354 die ("invalid depth");
355
356 init ();
357 read_data (data_file_type, data_file_name);
358 gen_bidi_type_tab (max_depth, data_file_type);
359 }
360
361 return 0;
362 }