1 /* FriBidi
2 * fribidi-char-sets-cap-rtl.c - CapRTL character set conversion routines
3 *
4 * Authors:
5 * Behdad Esfahbod, 2001, 2002, 2004
6 * Dov Grobgeld, 1999, 2000
7 *
8 * Copyright (C) 2004 Sharif FarsiWeb, Inc
9 * Copyright (C) 2001,2002 Behdad Esfahbod
10 * Copyright (C) 1999,2000 Dov Grobgeld
11 *
12 * This library is free software; you can redistribute it and/or
13 * modify it under the terms of the GNU Lesser General Public
14 * License as published by the Free Software Foundation; either
15 * version 2.1 of the License, or (at your option) any later version.
16 *
17 * This library is distributed in the hope that it will be useful,
18 * but WITHOUT ANY WARRANTY; without even the implied warranty of
19 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20 * Lesser General Public License for more details.
21 *
22 * You should have received a copy of the GNU Lesser General Public License
23 * along with this library, in a file named COPYING; if not, write to the
24 * Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
25 * Boston, MA 02110-1301, USA
26 *
27 * For licensing issues, contact <fribidi.license@gmail.com>.
28 */
29
30 #include <common.h>
31
32 #include <fribidi-char-sets-cap-rtl.h>
33
34 #include <fribidi-unicode.h>
35 #include <fribidi-mirroring.h>
36 #include <fribidi-bidi-types.h>
37
38 #include <bidi-types.h>
39
40 #include <stdio.h>
41
42 enum
43 {
44 # define _FRIBIDI_ADD_TYPE(TYPE,SYMBOL) TYPE = FRIBIDI_TYPE_##TYPE,
45 # include "fribidi-bidi-types-list.h"
46 # undef _FRIBIDI_ADD_TYPE
47 _FRIBIDI_MAX_TYPES_VALUE
48 };
49
50 enum
51 {
52 # define _FRIBIDI_ADD_TYPE(TYPE,SYMBOL) DUMMY_##TYPE,
53 # include "fribidi-bidi-types-list.h"
54 # undef _FRIBIDI_ADD_TYPE
55 _FRIBIDI_NUM_TYPES
56 };
57
58 static FriBidiCharType CapRTLCharTypes[] = {
59 /* *INDENT-OFF* */
60 ON, ON, ON, ON, LTR,RTL,ON, ON, ON, ON, ON, ON, ON, BS, RLO,RLE, /* 00-0f */
61 LRO,LRE,PDF,WS, LRI, RLI, FSI, PDI, ON, ON, ON, ON, ON, ON, ON, ON, /* 10-1f */
62 WS, ON, ON, ON, ET, ON, ON, ON, ON, ON, ON, ET, CS, ON, ES, ES, /* 20-2f */
63 EN, EN, EN, EN, EN, EN, AN, AN, AN, AN, CS, ON, ON, ON, ON, ON, /* 30-3f */
64 RTL,AL, AL, AL, AL, AL, AL, RTL,RTL,RTL,RTL,RTL,RTL,RTL,RTL,RTL, /* 40-4f */
65 RTL,RTL,RTL,RTL,RTL,RTL,RTL,RTL,RTL,RTL,RTL,ON, BS, ON, BN, ON, /* 50-5f */
66 NSM,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR, /* 60-6f */
67 LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,LTR,ON, SS, ON, WS, ON, /* 70-7f */
68 /* *INDENT-ON* */
69 };
70
71 #define CAPRTL_CHARS (int)(sizeof CapRTLCharTypes / sizeof CapRTLCharTypes[0])
72
73 static FriBidiChar *caprtl_to_unicode = NULL;
74
75 static void
76 init_cap_rtl (
77 void
78 )
79 {
80 int request[_FRIBIDI_NUM_TYPES];
81 FriBidiCharType to_type[_FRIBIDI_NUM_TYPES];
82 int num_types = 0, count = 0;
83 FriBidiCharType i;
84 char mark[CAPRTL_CHARS];
85
86 caprtl_to_unicode =
87 (FriBidiChar *) fribidi_malloc (CAPRTL_CHARS *
88 sizeof caprtl_to_unicode[0]);
89 for (i = 0; i < CAPRTL_CHARS; i++)
90 if (CapRTLCharTypes[i] == fribidi_get_bidi_type (i))
91 {
92 caprtl_to_unicode[i] = i;
93 mark[i] = 1;
94 }
95 else
96 {
97 int j;
98
99 caprtl_to_unicode[i] = FRIBIDI_UNICODE_CHARS;
100 mark[i] = 0;
101 if (fribidi_get_mirror_char (i, NULL))
102 {
103 DBG ("warning: I could not map mirroring character map to itself in CapRTL");
104 }
105
106 for (j = 0; j < num_types; j++)
107 if (to_type[j] == CapRTLCharTypes[i])
108 break;
109 if (j == num_types)
110 {
111 num_types++;
112 to_type[j] = CapRTLCharTypes[i];
113 request[j] = 0;
114 }
115 request[j]++;
116 count++;
117 }
118 for (i = 0; i < 0x10000 && count; i++) /* Assign BMP chars to CapRTL entries */
119 if (!fribidi_get_mirror_char (i, NULL) && !(i < CAPRTL_CHARS && mark[i]))
120 {
121 int j, k;
122 FriBidiCharType t = fribidi_get_bidi_type (i);
123 for (j = 0; j < num_types; j++)
124 if (to_type[j] == t)
125 break;
126 if (j >= num_types || !request[j]) /* Do not need this type */
127 continue;
128 for (k = 0; k < CAPRTL_CHARS; k++)
129 if (caprtl_to_unicode[k] == FRIBIDI_UNICODE_CHARS
130 && to_type[j] == CapRTLCharTypes[k])
131 {
132 request[j]--;
133 count--;
134 caprtl_to_unicode[k] = i;
135 break;
136 }
137 }
138 if (count)
139 {
140 int j;
141
142 DBG ("warning: could not find a mapping for CapRTL to Unicode:");
143 for (j = 0; j < num_types; j++)
144 if (request[j])
145 {
146 DBG2 (" need this type: %s", fribidi_get_bidi_type_name (to_type[j]));
147 }
148 }
149 }
150
151 static char
152 fribidi_unicode_to_cap_rtl_c (
153 /* input */
154 FriBidiChar uch
155 )
156 {
157 int i;
158
159 if (!caprtl_to_unicode)
160 init_cap_rtl ();
161
162 for (i = 0; i < CAPRTL_CHARS; i++)
163 if (uch == caprtl_to_unicode[i])
164 return (unsigned char) i;
165 return '?';
166 }
167
168 FriBidiStrIndex
169 fribidi_cap_rtl_to_unicode (
170 /* input */
171 const char *s,
172 FriBidiStrIndex len,
173 /* output */
174 FriBidiChar *us
175 )
176 {
177 FriBidiStrIndex i, j;
178
179 if (!caprtl_to_unicode)
180 init_cap_rtl ();
181
182 j = 0;
183 for (i = 0; i < len; i++)
184 {
185 char ch;
186
187 ch = s[i];
188 if (ch == '_')
189 {
190 switch (ch = s[++i])
191 {
192 case '>':
193 us[j++] = FRIBIDI_CHAR_LRM;
194 break;
195 case '<':
196 us[j++] = FRIBIDI_CHAR_RLM;
197 break;
198 case 'l':
199 us[j++] = FRIBIDI_CHAR_LRE;
200 break;
201 case 'r':
202 us[j++] = FRIBIDI_CHAR_RLE;
203 break;
204 case 'o':
205 us[j++] = FRIBIDI_CHAR_PDF;
206 break;
207 case 'L':
208 us[j++] = FRIBIDI_CHAR_LRO;
209 break;
210 case 'R':
211 us[j++] = FRIBIDI_CHAR_RLO;
212 break;
213 case 'i':
214 us[j++] = FRIBIDI_CHAR_LRI;
215 break;
216 case 'y':
217 us[j++] = FRIBIDI_CHAR_RLI;
218 break;
219 case 'f':
220 us[j++] = FRIBIDI_CHAR_FSI;
221 break;
222 case 'I':
223 us[j++] = FRIBIDI_CHAR_PDI;
224 break;
225 case '_':
226 us[j++] = '_';
227 break;
228 default:
229 us[j++] = '_';
230 i--;
231 break;
232 }
233 }
234 else
235 {
236 if ((int)s[i] < 0)
237 us[j++] = '?';
238 else
239 us[j++] = caprtl_to_unicode[(int) s[i]];
240 }
241 }
242
243 return j;
244 }
245
246 FriBidiStrIndex
247 fribidi_unicode_to_cap_rtl (
248 /* input */
249 const FriBidiChar *us,
250 FriBidiStrIndex len,
251 /* output */
252 char *s
253 )
254 {
255 FriBidiStrIndex i;
256 int j;
257
258 j = 0;
259 for (i = 0; i < len; i++)
260 {
261 FriBidiChar ch = us[i];
262 if (!FRIBIDI_IS_EXPLICIT (fribidi_get_bidi_type (ch))
263 && !FRIBIDI_IS_ISOLATE (fribidi_get_bidi_type (ch))
264 && ch != '_' && ch != FRIBIDI_CHAR_LRM && ch != FRIBIDI_CHAR_RLM)
265 s[j++] = fribidi_unicode_to_cap_rtl_c (ch);
266 else
267 {
268 s[j++] = '_';
269 switch (ch)
270 {
271 case FRIBIDI_CHAR_LRM:
272 s[j++] = '>';
273 break;
274 case FRIBIDI_CHAR_RLM:
275 s[j++] = '<';
276 break;
277 case FRIBIDI_CHAR_LRE:
278 s[j++] = 'l';
279 break;
280 case FRIBIDI_CHAR_RLE:
281 s[j++] = 'r';
282 break;
283 case FRIBIDI_CHAR_PDF:
284 s[j++] = 'o';
285 break;
286 case FRIBIDI_CHAR_LRO:
287 s[j++] = 'L';
288 break;
289 case FRIBIDI_CHAR_RLO:
290 s[j++] = 'R';
291 break;
292 case FRIBIDI_CHAR_LRI:
293 s[j++] = 'i';
294 break;
295 case FRIBIDI_CHAR_RLI:
296 s[j++] = 'y';
297 break;
298 case FRIBIDI_CHAR_FSI:
299 s[j++] = 'f';
300 break;
301 case FRIBIDI_CHAR_PDI:
302 s[j++] = 'I';
303 break;
304 case '_':
305 s[j++] = '_';
306 break;
307 default:
308 j--;
309 if (ch < 256)
310 s[j++] = fribidi_unicode_to_cap_rtl_c (ch);
311 else
312 s[j++] = '?';
313 break;
314 }
315 }
316 }
317 s[j] = 0;
318
319 return j;
320 }
321
322 const char *
323 fribidi_char_set_desc_cap_rtl (
324 void
325 )
326 {
327 static char *s = 0;
328 int l, i, j;
329
330 if (s)
331 return s;
332
333 l = 10000;
334 s = (char *) fribidi_malloc (l);
335 i = 0;
336 i += sprintf (s + i, /*l - i, */
337 "CapRTL is a character set for testing with the reference\n"
338 "implementation, with explicit marks escape strings, and\n"
339 "the property that contains all unicode character types in\n"
340 "ASCII range 1-127.\n"
341 "\n"
342 "Warning: CapRTL character types are subject to change.\n"
343 "\n" "CapRTL's character types:\n");
344 for (j = 0; j < CAPRTL_CHARS; j++)
345 {
346 if (j % 4 == 0)
347 s[i++] = '\n';
348 i += sprintf (s + i, /*l - i, */ " * 0x%02x %c%c %-3s ", j,
349 j < 0x20 ? '^' : ' ',
350 j < 0x20 ? j + '@' : j < 0x7f ? j : ' ',
351 fribidi_get_bidi_type_name (CapRTLCharTypes[j]));
352 }
353 i += sprintf (s + i, /*l - i, */
354 "\n\n"
355 "Escape sequences:\n"
356 " Character `_' is used to escape explicit marks. The list is:\n"
357 " * _> LRM\n" " * _< RLM\n"
358 " * _l LRE\n" " * _r RLE\n"
359 " * _L LRO\n" " * _R RLO\n"
360 " * _o PDF\n" " * _i LRI\n"
361 " * _y RLI\n" " * _f FSI\n"
362 " * _I PDI\n" " * __ `_' itself\n"
363 "\n");
364 return s;
365 }
366
367 /* Editor directions:
368 * vim:textwidth=78:tabstop=8:shiftwidth=2:autoindent:cindent
369 */