1 /* FriBidi
2 * gen-arabic-shaping-tab.c - generate arabic-shaping.tab.i
3 *
4 * Author:
5 * Behdad Esfahbod, 2004, 2005
6 *
7 * Copyright (C) 2004 Sharif FarsiWeb, Inc
8 * Copyright (C) 2004, 2005 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-arabic-shaping-tab"
58 #define outputname "arabic-shaping.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 const char *p,
85 const char *q
86 )
87 {
88 fprintf (stderr, appname ": ");
89 fprintf (stderr, fmt, p, q);
90 fprintf (stderr, "\n");
91 exit (1);
92 }
93
94 static void
95 die3l (
96 const char *fmt,
97 unsigned long l,
98 const char *p
99 )
100 {
101 fprintf (stderr, appname ": ");
102 fprintf (stderr, fmt, l, p);
103 fprintf (stderr, "\n");
104 exit (1);
105 }
106
107 static const char *arabic_shaping_tags[] = {
108 "isolated",
109 "final",
110 "initial",
111 "medial",
112 NULL
113 };
114
115 static int
116 shape_is (
117 const char *s,
118 const char *type_list[]
119 )
120 {
121 const char **p = type_list;
122 for (; *p; p++)
123 if (!strcmp (s, p[0]))
124 return p - type_list;
125 return -1;
126 }
127
128 #define table_name "ArShap"
129 #define macro_name "FRIBIDI_GET_ARABIC_SHAPE_PRES"
130
131 #define START_CHAR 0x600
132 #define END_CHAR 0x700
133
134 static FriBidiChar table[FRIBIDI_UNICODE_CHARS][4];
135 static char buf[4000];
136 static char tag[sizeof (buf)], buf2[sizeof (buf)];
137 static FriBidiChar minshaped, maxshaped;
138
139 static void
140 clear_tab (
141 void
142 )
143 {
144 register FriBidiChar c;
145 register int shape;
146
147 for (c = 0; c < FRIBIDI_UNICODE_CHARS; c++)
148 for (shape = 0; shape < 4; shape++)
149 table[c][shape] = c;
150 }
151
152 static void
153 init (
154 void
155 )
156 {
157 minshaped = FRIBIDI_UNICODE_CHARS;
158 maxshaped = 0;
159
160 clear_tab ();
161 }
162
163 static void
164 read_unicode_data_txt (
165 FILE *f
166 )
167 {
168 unsigned long c, unshaped, l;
169
170 l = 0;
171 while (fgets (buf, sizeof buf, f))
172 {
173 int i, shape;
174 const char *s = buf;
175
176 l++;
177
178 while (*s == ' ')
179 s++;
180
181 if (*s == '#' || *s == '\0' || *s == '\n')
182 continue;
183
184 i = sscanf (s, "%lx;%*[^;];%*[^;];%*[^;];%*[^;];<%[^;> ]> %lx %[^; ]", &c, tag, &unshaped, buf2);
185
186 if (i != 3)
187 continue;
188
189 if (i != 3 || c >= FRIBIDI_UNICODE_CHARS || unshaped >= FRIBIDI_UNICODE_CHARS)
190 die3l ("UnicodeData.txt: invalid input at line %ld: %s", l, s);
191
192 shape = shape_is (tag, arabic_shaping_tags);
193 if (shape >= 0)
194 {
195 table[unshaped][shape] = c;
196 if (unshaped < minshaped)
197 minshaped = unshaped;
198 if (unshaped > maxshaped)
199 maxshaped = unshaped;
200 }
201 }
202 }
203
204 static void
205 read_data (
206 const char *data_file_type[],
207 const char *data_file_name[]
208 )
209 {
210 FILE *f;
211
212 for (; data_file_name[0] && data_file_type[0];
213 data_file_name++, data_file_type++)
214 {
215 if (!(f = fopen (data_file_name[0], "rt")))
216 die2 ("error: cannot open `%s' for reading", data_file_name[0]);
217
218 if (!strcmp (data_file_type[0], "UnicodeData.txt"))
219 read_unicode_data_txt (f);
220 else
221 die2 ("error: unknown data-file type %s", data_file_type[0]);
222
223 fclose (f);
224 }
225
226 }
227
228 static void
229 gen_arabic_shaping_tab (
230 int max_depth /* currently unused */,
231 const char *data_file_type[]
232 )
233 {
234 register FriBidiChar c;
235 register int shape;
236
237 if (maxshaped < minshaped)
238 die ("error: no shaping pair found, something wrong with reading input");
239
240 printf ("/* " outputname "\n * generated by " appname " (" FRIBIDI_NAME " "
241 FRIBIDI_VERSION ")\n" " * from the files %s of Unicode version "
242 FRIBIDI_UNICODE_VERSION ". */\n\n", data_file_type[0]);
243
244 printf ("/*\n"
245 " use %s(key,shape) to access your table\n\n"
246 " required memory: %ld\n"
247 " */\n\n",
248 macro_name, (long)(maxshaped - minshaped + 1) * 4 * sizeof (FriBidiChar));
249
250 printf ("\n" "/* *IND" "ENT-OFF* */\n\n");
251
252 printf ("static const FriBidiChar %s[%d][%d] = {\n", table_name, maxshaped - minshaped + 1, 4);
253 for (c = minshaped; c <= maxshaped; c++)
254 {
255 printf (" {");
256 for (shape = 0; shape < 4; shape++)
257 printf ("0x%04lx,", (unsigned long)table[c][shape]);
258 printf ("},\n");
259 }
260
261
262 printf ("};\n\n");
263
264 printf ("/* *IND" "ENT-ON* */\n\n");
265
266 printf ("#ifndef FRIBIDI_ACCESS_SHAPE_TABLE\n"
267 "# define FRIBIDI_ACCESS_SHAPE_TABLE(table,min,max,x,shape) \\\n"
268 " (((x)<(min)||(x)>(max))?(x):(table)[(x)-(min)][(shape)])\n"
269 "#endif\n\n");
270 printf ("#define %s(x,shape) "
271 "FRIBIDI_ACCESS_SHAPE_TABLE(%s, 0x%04lx, 0x%04lx, (x), (shape))\n\n",
272 macro_name, table_name, (unsigned long)minshaped, (unsigned long)maxshaped);
273
274 printf ("/* End of generated " outputname " */\n");
275 }
276
277 int
278 main (
279 int argc,
280 const char **argv
281 )
282 {
283 const char *data_file_type[] =
284 { "UnicodeData.txt", NULL };
285
286 if (argc < 3)
287 die3 ("usage:\n " appname " max-depth /path/to/%s /path/to/%s [junk...]",
288 data_file_type[0], data_file_type[1]);
289
290 {
291 int max_depth = atoi (argv[1]);
292 const char *data_file_name[] = { NULL, NULL };
293 data_file_name[0] = argv[2];
294
295 if (max_depth < 2)
296 die ("invalid depth");
297
298 init ();
299 read_data (data_file_type, data_file_name);
300 gen_arabic_shaping_tab (max_depth, data_file_type);
301 }
302
303 return 0;
304 }