1
2 #line 1 "hb-ot-shaper-khmer-machine.rl"
3 /*
4 * Copyright © 2011,2012 Google, Inc.
5 *
6 * This is part of HarfBuzz, a text shaping library.
7 *
8 * Permission is hereby granted, without written agreement and without
9 * license or royalty fees, to use, copy, modify, and distribute this
10 * software and its documentation for any purpose, provided that the
11 * above copyright notice and the following two paragraphs appear in
12 * all copies of this software.
13 *
14 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
15 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
16 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
17 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
18 * DAMAGE.
19 *
20 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
21 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
22 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
23 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
24 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
25 *
26 * Google Author(s): Behdad Esfahbod
27 */
28
29 #ifndef HB_OT_SHAPER_KHMER_MACHINE_HH
30 #define HB_OT_SHAPER_KHMER_MACHINE_HH
31
32 #include "hb.hh"
33
34 #include "hb-ot-layout.hh"
35 #include "hb-ot-shaper-indic.hh"
36
37 /* buffer var allocations */
38 #define khmer_category() ot_shaper_var_u8_category() /* khmer_category_t */
39
40 using khmer_category_t = unsigned;
41
42 #define K_Cat(Cat) khmer_syllable_machine_ex_##Cat
43
44 enum khmer_syllable_type_t {
45 khmer_consonant_syllable,
46 khmer_broken_cluster,
47 khmer_non_khmer_cluster,
48 };
49
50
51 #line 52 "hb-ot-shaper-khmer-machine.hh"
52 #define khmer_syllable_machine_ex_C 1u
53 #define khmer_syllable_machine_ex_DOTTEDCIRCLE 11u
54 #define khmer_syllable_machine_ex_H 4u
55 #define khmer_syllable_machine_ex_PLACEHOLDER 10u
56 #define khmer_syllable_machine_ex_Ra 15u
57 #define khmer_syllable_machine_ex_Robatic 25u
58 #define khmer_syllable_machine_ex_V 2u
59 #define khmer_syllable_machine_ex_VAbv 20u
60 #define khmer_syllable_machine_ex_VBlw 21u
61 #define khmer_syllable_machine_ex_VPre 22u
62 #define khmer_syllable_machine_ex_VPst 23u
63 #define khmer_syllable_machine_ex_Xgroup 26u
64 #define khmer_syllable_machine_ex_Ygroup 27u
65 #define khmer_syllable_machine_ex_ZWJ 6u
66 #define khmer_syllable_machine_ex_ZWNJ 5u
67
68
69 #line 70 "hb-ot-shaper-khmer-machine.hh"
70 static const unsigned char _khmer_syllable_machine_trans_keys[] = {
71 5u, 26u, 5u, 26u, 1u, 15u, 5u, 26u, 5u, 26u, 5u, 26u, 5u, 26u, 5u, 26u,
72 5u, 26u, 5u, 26u, 5u, 26u, 5u, 26u, 5u, 26u, 1u, 15u, 5u, 26u, 5u, 26u,
73 5u, 26u, 5u, 26u, 5u, 26u, 5u, 26u, 5u, 26u, 1u, 27u, 4u, 27u, 1u, 15u,
74 4u, 27u, 4u, 27u, 27u, 27u, 4u, 27u, 4u, 27u, 4u, 27u, 4u, 27u, 4u, 27u,
75 4u, 27u, 1u, 15u, 4u, 27u, 4u, 27u, 27u, 27u, 4u, 27u, 4u, 27u, 4u, 27u,
76 4u, 27u, 4u, 27u, 5u, 26u, 0
77 };
78
79 static const char _khmer_syllable_machine_key_spans[] = {
80 22, 22, 15, 22, 22, 22, 22, 22,
81 22, 22, 22, 22, 22, 15, 22, 22,
82 22, 22, 22, 22, 22, 27, 24, 15,
83 24, 24, 1, 24, 24, 24, 24, 24,
84 24, 15, 24, 24, 1, 24, 24, 24,
85 24, 24, 22
86 };
87
88 static const short _khmer_syllable_machine_index_offsets[] = {
89 0, 23, 46, 62, 85, 108, 131, 154,
90 177, 200, 223, 246, 269, 292, 308, 331,
91 354, 377, 400, 423, 446, 469, 497, 522,
92 538, 563, 588, 590, 615, 640, 665, 690,
93 715, 740, 756, 781, 806, 808, 833, 858,
94 883, 908, 933
95 };
96
97 static const char _khmer_syllable_machine_indicies[] = {
98 1, 1, 0, 0, 0, 0, 0, 0,
99 0, 0, 0, 0, 0, 0, 0, 2,
100 0, 0, 0, 0, 3, 4, 0, 1,
101 1, 0, 0, 0, 0, 0, 0, 0,
102 0, 0, 0, 0, 0, 0, 0, 0,
103 0, 0, 0, 0, 4, 0, 5, 5,
104 0, 0, 0, 0, 0, 0, 0, 0,
105 0, 0, 0, 0, 5, 0, 1, 1,
106 0, 0, 0, 0, 0, 0, 0, 0,
107 0, 0, 0, 0, 0, 2, 0, 0,
108 0, 0, 0, 4, 0, 6, 6, 0,
109 0, 0, 0, 0, 0, 0, 0, 0,
110 0, 0, 0, 0, 0, 0, 0, 0,
111 0, 0, 2, 0, 7, 7, 0, 0,
112 0, 0, 0, 0, 0, 0, 0, 0,
113 0, 0, 0, 0, 0, 0, 0, 0,
114 0, 8, 0, 9, 9, 0, 0, 0,
115 0, 0, 0, 0, 0, 0, 0, 0,
116 0, 0, 2, 0, 0, 0, 0, 0,
117 10, 0, 9, 9, 0, 0, 0, 0,
118 0, 0, 0, 0, 0, 0, 0, 0,
119 0, 0, 0, 0, 0, 0, 0, 10,
120 0, 11, 11, 0, 0, 0, 0, 0,
121 0, 0, 0, 0, 0, 0, 0, 0,
122 2, 0, 0, 0, 0, 0, 12, 0,
123 11, 11, 0, 0, 0, 0, 0, 0,
124 0, 0, 0, 0, 0, 0, 0, 0,
125 0, 0, 0, 0, 0, 12, 0, 1,
126 1, 0, 0, 0, 0, 0, 0, 0,
127 0, 0, 0, 0, 0, 0, 2, 0,
128 0, 0, 0, 13, 4, 0, 15, 15,
129 14, 14, 14, 14, 14, 14, 14, 14,
130 14, 14, 14, 14, 14, 16, 14, 14,
131 14, 14, 17, 18, 14, 15, 15, 19,
132 19, 19, 19, 19, 19, 19, 19, 19,
133 19, 19, 19, 19, 19, 19, 19, 19,
134 19, 19, 18, 19, 20, 20, 14, 14,
135 14, 14, 14, 14, 14, 14, 14, 14,
136 14, 14, 20, 14, 15, 15, 14, 14,
137 14, 14, 14, 14, 14, 14, 14, 14,
138 14, 14, 14, 16, 14, 14, 14, 14,
139 14, 18, 14, 21, 21, 14, 14, 14,
140 14, 14, 14, 14, 14, 14, 14, 14,
141 14, 14, 14, 14, 14, 14, 14, 14,
142 16, 14, 22, 22, 14, 14, 14, 14,
143 14, 14, 14, 14, 14, 14, 14, 14,
144 14, 14, 14, 14, 14, 14, 14, 23,
145 14, 24, 24, 14, 14, 14, 14, 14,
146 14, 14, 14, 14, 14, 14, 14, 14,
147 16, 14, 14, 14, 14, 14, 25, 14,
148 24, 24, 14, 14, 14, 14, 14, 14,
149 14, 14, 14, 14, 14, 14, 14, 14,
150 14, 14, 14, 14, 14, 25, 14, 26,
151 26, 14, 14, 14, 14, 14, 14, 14,
152 14, 14, 14, 14, 14, 14, 16, 14,
153 14, 14, 14, 14, 27, 14, 26, 26,
154 14, 14, 14, 14, 14, 14, 14, 14,
155 14, 14, 14, 14, 14, 14, 14, 14,
156 14, 14, 14, 27, 14, 29, 29, 28,
157 30, 31, 31, 28, 28, 28, 13, 13,
158 28, 28, 28, 29, 28, 28, 28, 28,
159 16, 25, 27, 23, 28, 17, 18, 20,
160 28, 33, 34, 34, 32, 32, 32, 32,
161 32, 32, 32, 32, 32, 32, 32, 32,
162 32, 2, 10, 12, 8, 32, 13, 4,
163 5, 32, 35, 35, 32, 32, 32, 32,
164 32, 32, 32, 32, 32, 32, 32, 32,
165 35, 32, 33, 36, 36, 32, 32, 32,
166 32, 32, 32, 32, 32, 32, 32, 32,
167 32, 32, 2, 10, 12, 8, 32, 3,
168 4, 5, 32, 37, 38, 38, 32, 32,
169 32, 32, 32, 32, 32, 32, 32, 32,
170 32, 32, 32, 2, 10, 12, 8, 32,
171 32, 4, 5, 32, 5, 32, 37, 6,
172 6, 32, 32, 32, 32, 32, 32, 32,
173 32, 32, 32, 32, 32, 32, 32, 32,
174 32, 8, 32, 32, 2, 5, 32, 37,
175 7, 7, 32, 32, 32, 32, 32, 32,
176 32, 32, 32, 32, 32, 32, 32, 32,
177 32, 32, 32, 32, 32, 8, 5, 32,
178 37, 39, 39, 32, 32, 32, 32, 32,
179 32, 32, 32, 32, 32, 32, 32, 32,
180 2, 32, 32, 8, 32, 32, 10, 5,
181 32, 37, 40, 40, 32, 32, 32, 32,
182 32, 32, 32, 32, 32, 32, 32, 32,
183 32, 2, 10, 32, 8, 32, 32, 12,
184 5, 32, 33, 38, 38, 32, 32, 32,
185 32, 32, 32, 32, 32, 32, 32, 32,
186 32, 32, 2, 10, 12, 8, 32, 32,
187 4, 5, 32, 33, 38, 38, 32, 32,
188 32, 32, 32, 32, 32, 32, 32, 32,
189 32, 32, 32, 2, 10, 12, 8, 32,
190 3, 4, 5, 32, 42, 42, 41, 41,
191 41, 41, 41, 41, 41, 41, 41, 41,
192 41, 41, 42, 41, 30, 43, 43, 41,
193 41, 41, 41, 41, 41, 41, 41, 41,
194 41, 41, 41, 41, 16, 25, 27, 23,
195 41, 17, 18, 20, 41, 44, 45, 45,
196 41, 41, 41, 41, 41, 41, 41, 41,
197 41, 41, 41, 41, 41, 16, 25, 27,
198 23, 41, 41, 18, 20, 41, 20, 41,
199 44, 21, 21, 41, 41, 41, 41, 41,
200 41, 41, 41, 41, 41, 41, 41, 41,
201 41, 41, 41, 23, 41, 41, 16, 20,
202 41, 44, 22, 22, 41, 41, 41, 41,
203 41, 41, 41, 41, 41, 41, 41, 41,
204 41, 41, 41, 41, 41, 41, 41, 23,
205 20, 41, 44, 46, 46, 41, 41, 41,
206 41, 41, 41, 41, 41, 41, 41, 41,
207 41, 41, 16, 41, 41, 23, 41, 41,
208 25, 20, 41, 44, 47, 47, 41, 41,
209 41, 41, 41, 41, 41, 41, 41, 41,
210 41, 41, 41, 16, 25, 41, 23, 41,
211 41, 27, 20, 41, 30, 45, 45, 41,
212 41, 41, 41, 41, 41, 41, 41, 41,
213 41, 41, 41, 41, 16, 25, 27, 23,
214 41, 41, 18, 20, 41, 15, 15, 48,
215 48, 48, 48, 48, 48, 48, 48, 48,
216 48, 48, 48, 48, 16, 48, 48, 48,
217 48, 48, 18, 48, 0
218 };
219
220 static const char _khmer_syllable_machine_trans_targs[] = {
221 21, 1, 27, 31, 25, 26, 4, 5,
222 28, 7, 29, 9, 30, 32, 21, 12,
223 37, 41, 35, 21, 36, 15, 16, 38,
224 18, 39, 20, 40, 21, 22, 33, 42,
225 21, 23, 10, 24, 0, 2, 3, 6,
226 8, 21, 34, 11, 13, 14, 17, 19,
227 21
228 };
229
230 static const char _khmer_syllable_machine_trans_actions[] = {
231 1, 0, 2, 2, 2, 0, 0, 0,
232 2, 0, 2, 0, 2, 2, 3, 0,
233 2, 4, 4, 5, 0, 0, 0, 2,
234 0, 2, 0, 2, 8, 2, 0, 9,
235 10, 0, 0, 2, 0, 0, 0, 0,
236 0, 11, 4, 0, 0, 0, 0, 0,
237 12
238 };
239
240 static const char _khmer_syllable_machine_to_state_actions[] = {
241 0, 0, 0, 0, 0, 0, 0, 0,
242 0, 0, 0, 0, 0, 0, 0, 0,
243 0, 0, 0, 0, 0, 6, 0, 0,
244 0, 0, 0, 0, 0, 0, 0, 0,
245 0, 0, 0, 0, 0, 0, 0, 0,
246 0, 0, 0
247 };
248
249 static const char _khmer_syllable_machine_from_state_actions[] = {
250 0, 0, 0, 0, 0, 0, 0, 0,
251 0, 0, 0, 0, 0, 0, 0, 0,
252 0, 0, 0, 0, 0, 7, 0, 0,
253 0, 0, 0, 0, 0, 0, 0, 0,
254 0, 0, 0, 0, 0, 0, 0, 0,
255 0, 0, 0
256 };
257
258 static const short _khmer_syllable_machine_eof_trans[] = {
259 1, 1, 1, 1, 1, 1, 1, 1,
260 1, 1, 1, 15, 20, 15, 15, 15,
261 15, 15, 15, 15, 15, 0, 33, 33,
262 33, 33, 33, 33, 33, 33, 33, 33,
263 33, 42, 42, 42, 42, 42, 42, 42,
264 42, 42, 49
265 };
266
267 static const int khmer_syllable_machine_start = 21;
268 static const int khmer_syllable_machine_first_final = 21;
269 static const int khmer_syllable_machine_error = -1;
270
271 static const int khmer_syllable_machine_en_main = 21;
272
273
274 #line 53 "hb-ot-shaper-khmer-machine.rl"
275
276
277
278 #line 102 "hb-ot-shaper-khmer-machine.rl"
279
280
281 #define found_syllable(syllable_type) \
282 HB_STMT_START { \
283 if (0) fprintf (stderr, "syllable %u..%u %s\n", ts, te, #syllable_type); \
284 for (unsigned int i = ts; i < te; i++) \
285 info[i].syllable() = (syllable_serial << 4) | syllable_type; \
286 syllable_serial++; \
287 if (syllable_serial == 16) syllable_serial = 1; \
288 } HB_STMT_END
289
290 inline void
find_syllables_khmer(hb_buffer_t * buffer)291 find_syllables_khmer (hb_buffer_t *buffer)
292 {
293 unsigned int p, pe, eof, ts, te, act HB_UNUSED;
294 int cs;
295 hb_glyph_info_t *info = buffer->info;
296
297 #line 298 "hb-ot-shaper-khmer-machine.hh"
298 {
299 cs = khmer_syllable_machine_start;
300 ts = 0;
301 te = 0;
302 act = 0;
303 }
304
305 #line 122 "hb-ot-shaper-khmer-machine.rl"
306
307
308 p = 0;
309 pe = eof = buffer->len;
310
311 unsigned int syllable_serial = 1;
312
313 #line 314 "hb-ot-shaper-khmer-machine.hh"
314 {
315 int _slen;
316 int _trans;
317 const unsigned char *_keys;
318 const char *_inds;
319 if ( p == pe )
320 goto _test_eof;
321 _resume:
322 switch ( _khmer_syllable_machine_from_state_actions[cs] ) {
323 case 7:
324 #line 1 "NONE"
325 {ts = p;}
326 break;
327 #line 328 "hb-ot-shaper-khmer-machine.hh"
328 }
329
330 _keys = _khmer_syllable_machine_trans_keys + (cs<<1);
331 _inds = _khmer_syllable_machine_indicies + _khmer_syllable_machine_index_offsets[cs];
332
333 _slen = _khmer_syllable_machine_key_spans[cs];
334 _trans = _inds[ _slen > 0 && _keys[0] <=( info[p].khmer_category()) &&
335 ( info[p].khmer_category()) <= _keys[1] ?
336 ( info[p].khmer_category()) - _keys[0] : _slen ];
337
338 _eof_trans:
339 cs = _khmer_syllable_machine_trans_targs[_trans];
340
341 if ( _khmer_syllable_machine_trans_actions[_trans] == 0 )
342 goto _again;
343
344 switch ( _khmer_syllable_machine_trans_actions[_trans] ) {
345 case 2:
346 #line 1 "NONE"
347 {te = p+1;}
348 break;
349 case 8:
350 #line 98 "hb-ot-shaper-khmer-machine.rl"
351 {te = p+1;{ found_syllable (khmer_non_khmer_cluster); }}
352 break;
353 case 10:
354 #line 96 "hb-ot-shaper-khmer-machine.rl"
355 {te = p;p--;{ found_syllable (khmer_consonant_syllable); }}
356 break;
357 case 11:
358 #line 97 "hb-ot-shaper-khmer-machine.rl"
359 {te = p;p--;{ found_syllable (khmer_broken_cluster); buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_BROKEN_SYLLABLE; }}
360 break;
361 case 12:
362 #line 98 "hb-ot-shaper-khmer-machine.rl"
363 {te = p;p--;{ found_syllable (khmer_non_khmer_cluster); }}
364 break;
365 case 1:
366 #line 96 "hb-ot-shaper-khmer-machine.rl"
367 {{p = ((te))-1;}{ found_syllable (khmer_consonant_syllable); }}
368 break;
369 case 3:
370 #line 97 "hb-ot-shaper-khmer-machine.rl"
371 {{p = ((te))-1;}{ found_syllable (khmer_broken_cluster); buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_BROKEN_SYLLABLE; }}
372 break;
373 case 5:
374 #line 1 "NONE"
375 { switch( act ) {
376 case 2:
377 {{p = ((te))-1;} found_syllable (khmer_broken_cluster); buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_BROKEN_SYLLABLE; }
378 break;
379 case 3:
380 {{p = ((te))-1;} found_syllable (khmer_non_khmer_cluster); }
381 break;
382 }
383 }
384 break;
385 case 4:
386 #line 1 "NONE"
387 {te = p+1;}
388 #line 97 "hb-ot-shaper-khmer-machine.rl"
389 {act = 2;}
390 break;
391 case 9:
392 #line 1 "NONE"
393 {te = p+1;}
394 #line 98 "hb-ot-shaper-khmer-machine.rl"
395 {act = 3;}
396 break;
397 #line 398 "hb-ot-shaper-khmer-machine.hh"
398 }
399
400 _again:
401 switch ( _khmer_syllable_machine_to_state_actions[cs] ) {
402 case 6:
403 #line 1 "NONE"
404 {ts = 0;}
405 break;
406 #line 407 "hb-ot-shaper-khmer-machine.hh"
407 }
408
409 if ( ++p != pe )
410 goto _resume;
411 _test_eof: {}
412 if ( p == eof )
413 {
414 if ( _khmer_syllable_machine_eof_trans[cs] > 0 ) {
415 _trans = _khmer_syllable_machine_eof_trans[cs] - 1;
416 goto _eof_trans;
417 }
418 }
419
420 }
421
422 #line 130 "hb-ot-shaper-khmer-machine.rl"
423
424 }
425
426 #undef found_syllable
427
428 #endif /* HB_OT_SHAPER_KHMER_MACHINE_HH */
429