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
26 /*
27 *
28 * (C) Copyright IBM Corp. 1998-2013 - All Rights Reserved
29 *
30 */
31
32 #ifndef __INDICREORDERING_H
33 #define __INDICREORDERING_H
34
35 /**
36 * \file
37 * \internal
38 */
39
40 #include "LETypes.h"
41 #include "OpenTypeTables.h"
42
43 U_NAMESPACE_BEGIN
44
45 // Characters that get referred to by name...
46 #define C_SIGN_ZWNJ 0x200C
47 #define C_SIGN_ZWJ 0x200D
48
49 // Character class values
50 #define CC_RESERVED 0U
51 #define CC_VOWEL_MODIFIER 1U
52 #define CC_STRESS_MARK 2U
53 #define CC_INDEPENDENT_VOWEL 3U
54 #define CC_INDEPENDENT_VOWEL_2 4U
55 #define CC_INDEPENDENT_VOWEL_3 5U
56 #define CC_CONSONANT 6U
57 #define CC_CONSONANT_WITH_NUKTA 7U
58 #define CC_NUKTA 8U
59 #define CC_DEPENDENT_VOWEL 9U
60 #define CC_SPLIT_VOWEL_PIECE_1 10U
61 #define CC_SPLIT_VOWEL_PIECE_2 11U
62 #define CC_SPLIT_VOWEL_PIECE_3 12U
63 #define CC_VIRAMA 13U
64 #define CC_ZERO_WIDTH_MARK 14U
65 #define CC_AL_LAKUNA 15U
66 #define CC_COUNT 16U
67
68 // Character class flags
69 #define CF_CLASS_MASK 0x0000FFFFU
70
71 #define CF_CONSONANT 0x80000000U
72
73 #define CF_REPH 0x40000000U
74 #define CF_VATTU 0x20000000U
75 #define CF_BELOW_BASE 0x10000000U
76 #define CF_POST_BASE 0x08000000U
77 #define CF_LENGTH_MARK 0x04000000U
78 #define CF_PRE_BASE 0x02000000U
79
80 #define CF_POS_BEFORE 0x00300000U
81 #define CF_POS_BELOW 0x00200000U
82 #define CF_POS_ABOVE 0x00100000U
83 #define CF_POS_AFTER 0x00000000U
84 #define CF_POS_MASK 0x00300000U
85
86 #define CF_INDEX_MASK 0x000F0000U
87 #define CF_INDEX_SHIFT 16
88
89 // Script flag bits
90 #define SF_MATRAS_AFTER_BASE 0x80000000U
91 #define SF_REPH_AFTER_BELOW 0x40000000U
92 #define SF_EYELASH_RA 0x20000000U
93 #define SF_MPRE_FIXUP 0x10000000U
94 #define SF_FILTER_ZERO_WIDTH 0x08000000U
95
96 #define SF_POST_BASE_LIMIT_MASK 0x0000FFFFU
97 #define SF_NO_POST_BASE_LIMIT 0x00007FFFU
98
99 #define SM_MAX_PIECES 3
100
101 typedef LEUnicode SplitMatra[SM_MAX_PIECES];
102
103 class MPreFixups;
104 class LEGlyphStorage;
105
106 // Dynamic Properties ( v2 fonts only )
107 typedef le_uint32 DynamicProperties;
108
109 #define DP_REPH 0x80000000U
110 #define DP_HALF 0x40000000U
111 #define DP_PREF 0x20000000U
112 #define DP_BLWF 0x10000000U
113 #define DP_PSTF 0x08000000U
114
115 struct IndicClassTable
116 {
117 typedef le_uint32 CharClass;
118 typedef le_uint32 ScriptFlags;
119
120 LEUnicode firstChar;
121 LEUnicode lastChar;
122 le_int32 worstCaseExpansion;
123 ScriptFlags scriptFlags;
124 const CharClass *classTable;
125 const SplitMatra *splitMatraTable;
126
127 inline le_int32 getWorstCaseExpansion() const;
128 inline le_bool getFilterZeroWidth() const;
129
130 CharClass getCharClass(LEUnicode ch) const;
131
132 inline const SplitMatra *getSplitMatra(CharClass charClass) const;
133
134 inline le_bool isVowelModifier(LEUnicode ch) const;
135 inline le_bool isStressMark(LEUnicode ch) const;
136 inline le_bool isConsonant(LEUnicode ch) const;
137 inline le_bool isReph(LEUnicode ch) const;
138 inline le_bool isVirama(LEUnicode ch) const;
139 inline le_bool isAlLakuna(LEUnicode ch) const;
140 inline le_bool isNukta(LEUnicode ch) const;
141 inline le_bool isVattu(LEUnicode ch) const;
142 inline le_bool isMatra(LEUnicode ch) const;
143 inline le_bool isSplitMatra(LEUnicode ch) const;
144 inline le_bool isLengthMark(LEUnicode ch) const;
145 inline le_bool hasPostOrBelowBaseForm(LEUnicode ch) const;
146 inline le_bool hasPostBaseForm(LEUnicode ch) const;
147 inline le_bool hasBelowBaseForm(LEUnicode ch) const;
148 inline le_bool hasAboveBaseForm(LEUnicode ch) const;
149 inline le_bool hasPreBaseForm(LEUnicode ch) const;
150
151 inline static le_bool isVowelModifier(CharClass charClass);
152 inline static le_bool isStressMark(CharClass charClass);
153 inline static le_bool isConsonant(CharClass charClass);
154 inline static le_bool isReph(CharClass charClass);
155 inline static le_bool isVirama(CharClass charClass);
156 inline static le_bool isAlLakuna(CharClass charClass);
157 inline static le_bool isNukta(CharClass charClass);
158 inline static le_bool isVattu(CharClass charClass);
159 inline static le_bool isMatra(CharClass charClass);
160 inline static le_bool isSplitMatra(CharClass charClass);
161 inline static le_bool isLengthMark(CharClass charClass);
162 inline static le_bool hasPostOrBelowBaseForm(CharClass charClass);
163 inline static le_bool hasPostBaseForm(CharClass charClass);
164 inline static le_bool hasBelowBaseForm(CharClass charClass);
165 inline static le_bool hasAboveBaseForm(CharClass charClass);
166 inline static le_bool hasPreBaseForm(CharClass charClass);
167
168 static const IndicClassTable *getScriptClassTable(le_int32 scriptCode);
169 };
170
171 class IndicReordering /* not : public UObject because all methods are static */ {
172 public:
173 static le_int32 getWorstCaseExpansion(le_int32 scriptCode);
174
175 static le_bool getFilterZeroWidth(le_int32 scriptCode);
176
177 static le_int32 reorder(const LEUnicode *theChars, le_int32 charCount, le_int32 scriptCode,
178 LEUnicode *outChars, LEGlyphStorage &glyphStorage,
179 MPreFixups **outMPreFixups, LEErrorCode& success);
180
181 static void adjustMPres(MPreFixups *mpreFixups, LEGlyphStorage &glyphStorage, LEErrorCode& success);
182
183 static le_int32 v2process(const LEUnicode *theChars, le_int32 charCount, le_int32 scriptCode,
184 LEUnicode *outChars, LEGlyphStorage &glyphStorage, LEErrorCode& success);
185
186 static const FeatureMap *getFeatureMap(le_int32 &count);
187
188 static const FeatureMap *getv2FeatureMap(le_int32 &count);
189
190 static void applyPresentationForms(LEGlyphStorage &glyphStorage, le_int32 count);
191
192 static void finalReordering(LEGlyphStorage &glyphStorage, le_int32 count);
193
194 static void getDynamicProperties(DynamicProperties *dProps, const IndicClassTable *classTable);
195
196 private:
197 // do not instantiate
198 IndicReordering();
199
200 static le_int32 findSyllable(const IndicClassTable *classTable, const LEUnicode *chars, le_int32 prev, le_int32 charCount);
201
202 };
203
204 inline le_int32 IndicClassTable::getWorstCaseExpansion() const
205 {
206 return worstCaseExpansion;
207 }
208
209 inline le_bool IndicClassTable::getFilterZeroWidth() const
210 {
211 return (scriptFlags & SF_FILTER_ZERO_WIDTH) != 0;
212 }
213
214 inline const SplitMatra *IndicClassTable::getSplitMatra(CharClass charClass) const
215 {
216 le_int32 index = (charClass & CF_INDEX_MASK) >> CF_INDEX_SHIFT;
217
218 return &splitMatraTable[index - 1];
219 }
220
221 inline le_bool IndicClassTable::isVowelModifier(CharClass charClass)
222 {
223 return (charClass & CF_CLASS_MASK) == CC_VOWEL_MODIFIER;
224 }
225
226 inline le_bool IndicClassTable::isStressMark(CharClass charClass)
227 {
228 return (charClass & CF_CLASS_MASK) == CC_STRESS_MARK;
229 }
230
231 inline le_bool IndicClassTable::isConsonant(CharClass charClass)
232 {
233 return (charClass & CF_CONSONANT) != 0;
234 }
235
236 inline le_bool IndicClassTable::isReph(CharClass charClass)
237 {
238 return (charClass & CF_REPH) != 0;
239 }
240
241 inline le_bool IndicClassTable::isNukta(CharClass charClass)
242 {
243 return (charClass & CF_CLASS_MASK) == CC_NUKTA;
244 }
245
246 inline le_bool IndicClassTable::isVirama(CharClass charClass)
247 {
248 return (charClass & CF_CLASS_MASK) == CC_VIRAMA;
249 }
250
251 inline le_bool IndicClassTable::isAlLakuna(CharClass charClass)
252 {
253 return (charClass & CF_CLASS_MASK) == CC_AL_LAKUNA;
254 }
255
256 inline le_bool IndicClassTable::isVattu(CharClass charClass)
257 {
258 return (charClass & CF_VATTU) != 0;
259 }
260
261 inline le_bool IndicClassTable::isMatra(CharClass charClass)
262 {
263 charClass &= CF_CLASS_MASK;
264
265 return charClass >= CC_DEPENDENT_VOWEL && charClass <= CC_SPLIT_VOWEL_PIECE_3;
266 }
267
268 inline le_bool IndicClassTable::isSplitMatra(CharClass charClass)
269 {
270 return (charClass & CF_INDEX_MASK) != 0;
271 }
272
273 inline le_bool IndicClassTable::isLengthMark(CharClass charClass)
274 {
275 return (charClass & CF_LENGTH_MARK) != 0;
276 }
277
278 inline le_bool IndicClassTable::hasPostOrBelowBaseForm(CharClass charClass)
279 {
280 return (charClass & (CF_POST_BASE | CF_BELOW_BASE)) != 0;
281 }
282
283 inline le_bool IndicClassTable::hasPostBaseForm(CharClass charClass)
284 {
285 return (charClass & CF_POST_BASE) != 0;
286 }
287
288 inline le_bool IndicClassTable::hasPreBaseForm(CharClass charClass)
289 {
290 return (charClass & CF_PRE_BASE) != 0;
291 }
292
293 inline le_bool IndicClassTable::hasBelowBaseForm(CharClass charClass)
294 {
295 return (charClass & CF_BELOW_BASE) != 0;
296 }
297
298 inline le_bool IndicClassTable::hasAboveBaseForm(CharClass charClass)
299 {
300 return ((charClass & CF_POS_MASK) == CF_POS_ABOVE);
301 }
302
303 inline le_bool IndicClassTable::isVowelModifier(LEUnicode ch) const
304 {
305 return isVowelModifier(getCharClass(ch));
306 }
307
308 inline le_bool IndicClassTable::isStressMark(LEUnicode ch) const
309 {
310 return isStressMark(getCharClass(ch));
311 }
312
313 inline le_bool IndicClassTable::isConsonant(LEUnicode ch) const
314 {
315 return isConsonant(getCharClass(ch));
316 }
317
318 inline le_bool IndicClassTable::isReph(LEUnicode ch) const
319 {
320 return isReph(getCharClass(ch));
321 }
322
323 inline le_bool IndicClassTable::isVirama(LEUnicode ch) const
324 {
325 return isVirama(getCharClass(ch));
326 }
327
328 inline le_bool IndicClassTable::isAlLakuna(LEUnicode ch) const
329 {
330 return isAlLakuna(getCharClass(ch));
331 }
332
333 inline le_bool IndicClassTable::isNukta(LEUnicode ch) const
334 {
335 return isNukta(getCharClass(ch));
336 }
337
338 inline le_bool IndicClassTable::isVattu(LEUnicode ch) const
339 {
340 return isVattu(getCharClass(ch));
341 }
342
343 inline le_bool IndicClassTable::isMatra(LEUnicode ch) const
344 {
345 return isMatra(getCharClass(ch));
346 }
347
348 inline le_bool IndicClassTable::isSplitMatra(LEUnicode ch) const
349 {
350 return isSplitMatra(getCharClass(ch));
351 }
352
353 inline le_bool IndicClassTable::isLengthMark(LEUnicode ch) const
354 {
355 return isLengthMark(getCharClass(ch));
356 }
357
358 inline le_bool IndicClassTable::hasPostOrBelowBaseForm(LEUnicode ch) const
359 {
360 return hasPostOrBelowBaseForm(getCharClass(ch));
361 }
362
363 inline le_bool IndicClassTable::hasPostBaseForm(LEUnicode ch) const
364 {
365 return hasPostBaseForm(getCharClass(ch));
366 }
367
368 inline le_bool IndicClassTable::hasBelowBaseForm(LEUnicode ch) const
369 {
370 return hasBelowBaseForm(getCharClass(ch));
371 }
372
373 inline le_bool IndicClassTable::hasPreBaseForm(LEUnicode ch) const
374 {
375 return hasPreBaseForm(getCharClass(ch));
376 }
377
378 inline le_bool IndicClassTable::hasAboveBaseForm(LEUnicode ch) const
379 {
380 return hasAboveBaseForm(getCharClass(ch));
381 }
382 U_NAMESPACE_END
383 #endif