1 /*
2 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
3 *
4 * This code is free software; you can redistribute it and/or modify it
5 * under the terms of the GNU General Public License version 2 only, as
6 * published by the Free Software Foundation. Oracle designates this
7 * particular file as subject to the "Classpath" exception as provided
8 * by Oracle in the LICENSE file that accompanied this code.
9 *
10 * This code is distributed in the hope that it will be useful, but WITHOUT
11 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
12 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
13 * version 2 for more details (a copy is included in the LICENSE file that
14 * accompanied this code).
15 *
16 * You should have received a copy of the GNU General Public License version
17 * 2 along with this work; if not, write to the Free Software Foundation,
18 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
19 *
20 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
21 * or visit www.oracle.com if you need additional information or have any
22 * questions.
23 */
24
25 // This file is available under and governed by the GNU General Public
26 // License version 2 only, as published by the Free Software Foundation.
27 // However, the following notice accompanied the original version of this
28 // file:
29 //
30 /*
31 * Copyright © 2010,2012 Google, Inc.
32 *
33 * This is part of HarfBuzz, a text shaping library.
34 *
35 * Permission is hereby granted, without written agreement and without
36 * license or royalty fees, to use, copy, modify, and distribute this
37 * software and its documentation for any purpose, provided that the
38 * above copyright notice and the following two paragraphs appear in
39 * all copies of this software.
40 *
41 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
42 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
43 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
44 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
45 * DAMAGE.
46 *
47 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
48 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
49 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
50 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
51 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
52 *
53 * Google Author(s): Behdad Esfahbod
54 */
55
56 #include "hb-ot-shape-complex-private.hh"
57
58
59 /* Thai / Lao shaper */
60
61
62 /* PUA shaping */
63
64
65 enum thai_consonant_type_t
66 {
67 NC,
68 AC,
69 RC,
70 DC,
71 NOT_CONSONANT,
72 NUM_CONSONANT_TYPES = NOT_CONSONANT
73 };
74
75 static thai_consonant_type_t
76 get_consonant_type (hb_codepoint_t u)
77 {
78 if (u == 0x0E1Bu || u == 0x0E1Du || u == 0x0E1Fu/* || u == 0x0E2Cu*/)
79 return AC;
80 if (u == 0x0E0Du || u == 0x0E10u)
81 return RC;
82 if (u == 0x0E0Eu || u == 0x0E0Fu)
83 return DC;
84 if (hb_in_range (u, 0x0E01u, 0x0E2Eu))
85 return NC;
86 return NOT_CONSONANT;
87 }
88
89
90 enum thai_mark_type_t
91 {
92 AV,
93 BV,
94 T,
95 NOT_MARK,
96 NUM_MARK_TYPES = NOT_MARK
97 };
98
99 static thai_mark_type_t
100 get_mark_type (hb_codepoint_t u)
101 {
102 if (u == 0x0E31u || hb_in_range (u, 0x0E34u, 0x0E37u) ||
103 u == 0x0E47u || hb_in_range (u, 0x0E4Du, 0x0E4Eu))
104 return AV;
105 if (hb_in_range (u, 0x0E38u, 0x0E3Au))
106 return BV;
107 if (hb_in_range (u, 0x0E48u, 0x0E4Cu))
108 return T;
109 return NOT_MARK;
110 }
111
112
113 enum thai_action_t
114 {
115 NOP,
116 SD, /* Shift combining-mark down */
117 SL, /* Shift combining-mark left */
118 SDL, /* Shift combining-mark down-left */
119 RD /* Remove descender from base */
120 };
121
122 static hb_codepoint_t
123 thai_pua_shape (hb_codepoint_t u, thai_action_t action, hb_font_t *font)
124 {
125 struct thai_pua_mapping_t {
126 hb_codepoint_t u;
127 hb_codepoint_t win_pua;
128 hb_codepoint_t mac_pua;
129 } const *pua_mappings = NULL;
130 static const thai_pua_mapping_t SD_mappings[] = {
131 {0x0E48u, 0xF70Au, 0xF88Bu}, /* MAI EK */
132 {0x0E49u, 0xF70Bu, 0xF88Eu}, /* MAI THO */
133 {0x0E4Au, 0xF70Cu, 0xF891u}, /* MAI TRI */
134 {0x0E4Bu, 0xF70Du, 0xF894u}, /* MAI CHATTAWA */
135 {0x0E4Cu, 0xF70Eu, 0xF897u}, /* THANTHAKHAT */
136 {0x0E38u, 0xF718u, 0xF89Bu}, /* SARA U */
137 {0x0E39u, 0xF719u, 0xF89Cu}, /* SARA UU */
138 {0x0E3Au, 0xF71Au, 0xF89Du}, /* PHINTHU */
139 {0x0000u, 0x0000u, 0x0000u}
140 };
141 static const thai_pua_mapping_t SDL_mappings[] = {
142 {0x0E48u, 0xF705u, 0xF88Cu}, /* MAI EK */
143 {0x0E49u, 0xF706u, 0xF88Fu}, /* MAI THO */
144 {0x0E4Au, 0xF707u, 0xF892u}, /* MAI TRI */
145 {0x0E4Bu, 0xF708u, 0xF895u}, /* MAI CHATTAWA */
146 {0x0E4Cu, 0xF709u, 0xF898u}, /* THANTHAKHAT */
147 {0x0000u, 0x0000u, 0x0000u}
148 };
149 static const thai_pua_mapping_t SL_mappings[] = {
150 {0x0E48u, 0xF713u, 0xF88Au}, /* MAI EK */
151 {0x0E49u, 0xF714u, 0xF88Du}, /* MAI THO */
152 {0x0E4Au, 0xF715u, 0xF890u}, /* MAI TRI */
153 {0x0E4Bu, 0xF716u, 0xF893u}, /* MAI CHATTAWA */
154 {0x0E4Cu, 0xF717u, 0xF896u}, /* THANTHAKHAT */
155 {0x0E31u, 0xF710u, 0xF884u}, /* MAI HAN-AKAT */
156 {0x0E34u, 0xF701u, 0xF885u}, /* SARA I */
157 {0x0E35u, 0xF702u, 0xF886u}, /* SARA II */
158 {0x0E36u, 0xF703u, 0xF887u}, /* SARA UE */
159 {0x0E37u, 0xF704u, 0xF888u}, /* SARA UEE */
160 {0x0E47u, 0xF712u, 0xF889u}, /* MAITAIKHU */
161 {0x0E4Du, 0xF711u, 0xF899u}, /* NIKHAHIT */
162 {0x0000u, 0x0000u, 0x0000u}
163 };
164 static const thai_pua_mapping_t RD_mappings[] = {
165 {0x0E0Du, 0xF70Fu, 0xF89Au}, /* YO YING */
166 {0x0E10u, 0xF700u, 0xF89Eu}, /* THO THAN */
167 {0x0000u, 0x0000u, 0x0000u}
168 };
169
170 switch (action) {
171 default: assert (false); /* Fallthrough */
172 case NOP: return u;
173 case SD: pua_mappings = SD_mappings; break;
174 case SDL: pua_mappings = SDL_mappings; break;
175 case SL: pua_mappings = SL_mappings; break;
176 case RD: pua_mappings = RD_mappings; break;
177 }
178 for (; pua_mappings->u; pua_mappings++)
179 if (pua_mappings->u == u)
180 {
181 hb_codepoint_t glyph;
182 if (hb_font_get_glyph (font, pua_mappings->win_pua, 0, &glyph))
183 return pua_mappings->win_pua;
184 if (hb_font_get_glyph (font, pua_mappings->mac_pua, 0, &glyph))
185 return pua_mappings->mac_pua;
186 break;
187 }
188 return u;
189 }
190
191
192 static enum thai_above_state_t
193 { /* Cluster above looks like: */
194 T0, /* ⣤ */
195 T1, /* ⣼ */
196 T2, /* ⣾ */
197 T3, /* ⣿ */
198 NUM_ABOVE_STATES
199 } thai_above_start_state[NUM_CONSONANT_TYPES + 1/* For NOT_CONSONANT */] =
200 {
201 T0, /* NC */
202 T1, /* AC */
203 T0, /* RC */
204 T0, /* DC */
205 T3, /* NOT_CONSONANT */
206 };
207
208 static const struct thai_above_state_machine_edge_t {
209 thai_action_t action;
210 thai_above_state_t next_state;
211 } thai_above_state_machine[NUM_ABOVE_STATES][NUM_MARK_TYPES] =
212 { /*AV*/ /*BV*/ /*T*/
213 /*T0*/ {{NOP,T3}, {NOP,T0}, {SD, T3}},
214 /*T1*/ {{SL, T2}, {NOP,T1}, {SDL,T2}},
215 /*T2*/ {{NOP,T3}, {NOP,T2}, {SL, T3}},
216 /*T3*/ {{NOP,T3}, {NOP,T3}, {NOP,T3}},
217 };
218
219
220 static enum thai_below_state_t
221 {
222 B0, /* No descender */
223 B1, /* Removable descender */
224 B2, /* Strict descender */
225 NUM_BELOW_STATES
226 } thai_below_start_state[NUM_CONSONANT_TYPES + 1/* For NOT_CONSONANT */] =
227 {
228 B0, /* NC */
229 B0, /* AC */
230 B1, /* RC */
231 B2, /* DC */
232 B2, /* NOT_CONSONANT */
233 };
234
235 static const struct thai_below_state_machine_edge_t {
236 thai_action_t action;
237 thai_below_state_t next_state;
238 } thai_below_state_machine[NUM_BELOW_STATES][NUM_MARK_TYPES] =
239 { /*AV*/ /*BV*/ /*T*/
240 /*B0*/ {{NOP,B0}, {NOP,B2}, {NOP, B0}},
241 /*B1*/ {{NOP,B1}, {RD, B2}, {NOP, B1}},
242 /*B2*/ {{NOP,B2}, {SD, B2}, {NOP, B2}},
243 };
244
245
246 static void
247 do_thai_pua_shaping (const hb_ot_shape_plan_t *plan HB_UNUSED,
248 hb_buffer_t *buffer,
249 hb_font_t *font)
250 {
251 thai_above_state_t above_state = thai_above_start_state[NOT_CONSONANT];
252 thai_below_state_t below_state = thai_below_start_state[NOT_CONSONANT];
253 unsigned int base = 0;
254
255 hb_glyph_info_t *info = buffer->info;
256 unsigned int count = buffer->len;
257 for (unsigned int i = 0; i < count; i++)
258 {
259 thai_mark_type_t mt = get_mark_type (info[i].codepoint);
260
261 if (mt == NOT_MARK) {
262 thai_consonant_type_t ct = get_consonant_type (info[i].codepoint);
263 above_state = thai_above_start_state[ct];
264 below_state = thai_below_start_state[ct];
265 base = i;
266 continue;
267 }
268
269 const thai_above_state_machine_edge_t &above_edge = thai_above_state_machine[above_state][mt];
270 const thai_below_state_machine_edge_t &below_edge = thai_below_state_machine[below_state][mt];
271 above_state = above_edge.next_state;
272 below_state = below_edge.next_state;
273
274 /* At least one of the above/below actions is NOP. */
275 thai_action_t action = above_edge.action != NOP ? above_edge.action : below_edge.action;
276
277 if (action == RD)
278 info[base].codepoint = thai_pua_shape (info[base].codepoint, action, font);
279 else
280 info[i].codepoint = thai_pua_shape (info[i].codepoint, action, font);
281 }
282 }
283
284
285 static void
286 preprocess_text_thai (const hb_ot_shape_plan_t *plan,
287 hb_buffer_t *buffer,
288 hb_font_t *font)
289 {
290 /* This function implements the shaping logic documented here:
291 *
292 * http://linux.thai.net/~thep/th-otf/shaping.html
293 *
294 * The first shaping rule listed there is needed even if the font has Thai
295 * OpenType tables. The rest do fallback positioning based on PUA codepoints.
296 * We implement that only if there exist no Thai GSUB in the font.
297 */
298
299 /* The following is NOT specified in the MS OT Thai spec, however, it seems
300 * to be what Uniscribe and other engines implement. According to Eric Muller:
301 *
302 * When you have a SARA AM, decompose it in NIKHAHIT + SARA AA, *and* move the
303 * NIKHAHIT backwards over any tone mark (0E48-0E4B).
304 *
305 * <0E14, 0E4B, 0E33> -> <0E14, 0E4D, 0E4B, 0E32>
306 *
307 * This reordering is legit only when the NIKHAHIT comes from a SARA AM, not
308 * when it's there to start with. The string <0E14, 0E4B, 0E4D> is probably
309 * not what a user wanted, but the rendering is nevertheless nikhahit above
310 * chattawa.
311 *
312 * Same for Lao.
313 *
314 * Note:
315 *
316 * Uniscribe also does some below-marks reordering. Namely, it positions U+0E3A
317 * after U+0E38 and U+0E39. We do that by modifying the ccc for U+0E3A.
318 * See unicode->modified_combining_class (). Lao does NOT have a U+0E3A
319 * equivalent.
320 */
321
322
323 /*
324 * Here are the characters of significance:
325 *
326 * Thai Lao
327 * SARA AM: U+0E33 U+0EB3
328 * SARA AA: U+0E32 U+0EB2
329 * Nikhahit: U+0E4D U+0ECD
330 *
331 * Testing shows that Uniscribe reorder the following marks:
332 * Thai: <0E31,0E34..0E37,0E47..0E4E>
333 * Lao: <0EB1,0EB4..0EB7,0EC7..0ECE>
334 *
335 * Note how the Lao versions are the same as Thai + 0x80.
336 */
337
338 /* We only get one script at a time, so a script-agnostic implementation
339 * is adequate here. */
340 #define IS_SARA_AM(x) (((x) & ~0x0080u) == 0x0E33u)
341 #define NIKHAHIT_FROM_SARA_AM(x) ((x) - 0x0E33u + 0x0E4Du)
342 #define SARA_AA_FROM_SARA_AM(x) ((x) - 1)
343 #define IS_TONE_MARK(x) (hb_in_ranges ((x) & ~0x0080u, 0x0E34u, 0x0E37u, 0x0E47u, 0x0E4Eu, 0x0E31u, 0x0E31u))
344
345 buffer->clear_output ();
346 unsigned int count = buffer->len;
347 for (buffer->idx = 0; buffer->idx < count;)
348 {
349 hb_codepoint_t u = buffer->cur().codepoint;
350 if (likely (!IS_SARA_AM (u))) {
351 buffer->next_glyph ();
352 continue;
353 }
354
355 /* Is SARA AM. Decompose and reorder. */
356 hb_codepoint_t decomposed[2] = {hb_codepoint_t (NIKHAHIT_FROM_SARA_AM (u)),
357 hb_codepoint_t (SARA_AA_FROM_SARA_AM (u))};
358 buffer->replace_glyphs (1, 2, decomposed);
359 if (unlikely (buffer->in_error))
360 return;
361
362 /* Make Nikhahit be recognized as a mark when zeroing widths. */
363 unsigned int end = buffer->out_len;
364 _hb_glyph_info_set_general_category (&buffer->out_info[end - 2], HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK);
365
366 /* Ok, let's see... */
367 unsigned int start = end - 2;
368 while (start > 0 && IS_TONE_MARK (buffer->out_info[start - 1].codepoint))
369 start--;
370
371 if (start + 2 < end)
372 {
373 /* Move Nikhahit (end-2) to the beginning */
374 buffer->merge_out_clusters (start, end);
375 hb_glyph_info_t t = buffer->out_info[end - 2];
376 memmove (buffer->out_info + start + 1,
377 buffer->out_info + start,
378 sizeof (buffer->out_info[0]) * (end - start - 2));
379 buffer->out_info[start] = t;
380 }
381 else
382 {
383 /* Since we decomposed, and NIKHAHIT is combining, merge clusters with the
384 * previous cluster. */
385 if (start && buffer->cluster_level == HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES)
386 buffer->merge_out_clusters (start - 1, end);
387 }
388 }
389 buffer->swap_buffers ();
390
391 /* If font has Thai GSUB, we are done. */
392 if (plan->props.script == HB_SCRIPT_THAI && !plan->map.found_script[0])
393 do_thai_pua_shaping (plan, buffer, font);
394 }
395
396 const hb_ot_complex_shaper_t _hb_ot_complex_shaper_thai =
397 {
398 "thai",
399 NULL, /* collect_features */
400 NULL, /* override_features */
401 NULL, /* data_create */
402 NULL, /* data_destroy */
403 preprocess_text_thai,
404 HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
405 NULL, /* decompose */
406 NULL, /* compose */
407 NULL, /* setup_masks */
408 HB_OT_SHAPE_ZERO_WIDTH_MARKS_DEFAULT,
409 false,/* fallback_position */
410 };