1 /*
2 *
3 * (C) Copyright IBM Corp. 1998-2011 - All Rights Reserved
4 *
5 */
6
7 #ifndef __INDICREORDERING_H
8 #define __INDICREORDERING_H
9
10 /**
11 * \file
12 * \internal
13 */
14
15 #include "LETypes.h"
16 #include "OpenTypeTables.h"
17
18 U_NAMESPACE_BEGIN
19
20 // Characters that get refered to by name...
21 #define C_SIGN_ZWNJ 0x200C
22 #define C_SIGN_ZWJ 0x200D
23
24 // Character class values
25 #define CC_RESERVED 0U
26 #define CC_VOWEL_MODIFIER 1U
27 #define CC_STRESS_MARK 2U
28 #define CC_INDEPENDENT_VOWEL 3U
29 #define CC_INDEPENDENT_VOWEL_2 4U
30 #define CC_INDEPENDENT_VOWEL_3 5U
31 #define CC_CONSONANT 6U
32 #define CC_CONSONANT_WITH_NUKTA 7U
33 #define CC_NUKTA 8U
34 #define CC_DEPENDENT_VOWEL 9U
35 #define CC_SPLIT_VOWEL_PIECE_1 10U
36 #define CC_SPLIT_VOWEL_PIECE_2 11U
37 #define CC_SPLIT_VOWEL_PIECE_3 12U
38 #define CC_VIRAMA 13U
39 #define CC_ZERO_WIDTH_MARK 14U
40 #define CC_AL_LAKUNA 15U
41 #define CC_COUNT 16U
42
43 // Character class flags
44 #define CF_CLASS_MASK 0x0000FFFFU
45
46 #define CF_CONSONANT 0x80000000U
47
48 #define CF_REPH 0x40000000U
49 #define CF_VATTU 0x20000000U
50 #define CF_BELOW_BASE 0x10000000U
51 #define CF_POST_BASE 0x08000000U
52 #define CF_LENGTH_MARK 0x04000000U
53 #define CF_PRE_BASE 0x02000000U
54
55 #define CF_POS_BEFORE 0x00300000U
56 #define CF_POS_BELOW 0x00200000U
57 #define CF_POS_ABOVE 0x00100000U
58 #define CF_POS_AFTER 0x00000000U
59 #define CF_POS_MASK 0x00300000U
60
61 #define CF_INDEX_MASK 0x000F0000U
62 #define CF_INDEX_SHIFT 16
63
64 // Script flag bits
65 #define SF_MATRAS_AFTER_BASE 0x80000000U
66 #define SF_REPH_AFTER_BELOW 0x40000000U
67 #define SF_EYELASH_RA 0x20000000U
68 #define SF_MPRE_FIXUP 0x10000000U
69 #define SF_FILTER_ZERO_WIDTH 0x08000000U
70
71 #define SF_POST_BASE_LIMIT_MASK 0x0000FFFFU
72 #define SF_NO_POST_BASE_LIMIT 0x00007FFFU
73
74 #define SM_MAX_PIECES 3
75
76 typedef LEUnicode SplitMatra[SM_MAX_PIECES];
77
78 class MPreFixups;
79 class LEGlyphStorage;
80
81 // Dynamic Properties ( v2 fonts only )
82 typedef le_uint32 DynamicProperties;
83
84 #define DP_REPH 0x80000000U
85 #define DP_HALF 0x40000000U
86 #define DP_PREF 0x20000000U
87 #define DP_BLWF 0x10000000U
88 #define DP_PSTF 0x08000000U
89
90 struct IndicClassTable
91 {
92 typedef le_uint32 CharClass;
93 typedef le_uint32 ScriptFlags;
94
95 LEUnicode firstChar;
96 LEUnicode lastChar;
97 le_int32 worstCaseExpansion;
98 ScriptFlags scriptFlags;
99 const CharClass *classTable;
100 const SplitMatra *splitMatraTable;
101
102 inline le_int32 getWorstCaseExpansion() const;
103 inline le_bool getFilterZeroWidth() const;
104
105 CharClass getCharClass(LEUnicode ch) const;
106
107 inline const SplitMatra *getSplitMatra(CharClass charClass) const;
108
109 inline le_bool isVowelModifier(LEUnicode ch) const;
110 inline le_bool isStressMark(LEUnicode ch) const;
111 inline le_bool isConsonant(LEUnicode ch) const;
112 inline le_bool isReph(LEUnicode ch) const;
113 inline le_bool isVirama(LEUnicode ch) const;
114 inline le_bool isAlLakuna(LEUnicode ch) const;
115 inline le_bool isNukta(LEUnicode ch) const;
116 inline le_bool isVattu(LEUnicode ch) const;
117 inline le_bool isMatra(LEUnicode ch) const;
118 inline le_bool isSplitMatra(LEUnicode ch) const;
119 inline le_bool isLengthMark(LEUnicode ch) const;
120 inline le_bool hasPostOrBelowBaseForm(LEUnicode ch) const;
121 inline le_bool hasPostBaseForm(LEUnicode ch) const;
122 inline le_bool hasBelowBaseForm(LEUnicode ch) const;
123 inline le_bool hasAboveBaseForm(LEUnicode ch) const;
124 inline le_bool hasPreBaseForm(LEUnicode ch) const;
125
126 inline static le_bool isVowelModifier(CharClass charClass);
127 inline static le_bool isStressMark(CharClass charClass);
128 inline static le_bool isConsonant(CharClass charClass);
129 inline static le_bool isReph(CharClass charClass);
130 inline static le_bool isVirama(CharClass charClass);
131 inline static le_bool isAlLakuna(CharClass charClass);
132 inline static le_bool isNukta(CharClass charClass);
133 inline static le_bool isVattu(CharClass charClass);
134 inline static le_bool isMatra(CharClass charClass);
135 inline static le_bool isSplitMatra(CharClass charClass);
136 inline static le_bool isLengthMark(CharClass charClass);
137 inline static le_bool hasPostOrBelowBaseForm(CharClass charClass);
138 inline static le_bool hasPostBaseForm(CharClass charClass);
139 inline static le_bool hasBelowBaseForm(CharClass charClass);
140 inline static le_bool hasAboveBaseForm(CharClass charClass);
141 inline static le_bool hasPreBaseForm(CharClass charClass);
142
143 static const IndicClassTable *getScriptClassTable(le_int32 scriptCode);
144 };
145
146 class IndicReordering /* not : public UObject because all methods are static */ {
147 public:
148 static le_int32 getWorstCaseExpansion(le_int32 scriptCode);
149
150 static le_bool getFilterZeroWidth(le_int32 scriptCode);
151
152 static le_int32 reorder(const LEUnicode *theChars, le_int32 charCount, le_int32 scriptCode,
153 LEUnicode *outChars, LEGlyphStorage &glyphStorage,
154 MPreFixups **outMPreFixups, LEErrorCode& success);
155
156 static void adjustMPres(MPreFixups *mpreFixups, LEGlyphStorage &glyphStorage, LEErrorCode& success);
157
158 static le_int32 v2process(const LEUnicode *theChars, le_int32 charCount, le_int32 scriptCode,
159 LEUnicode *outChars, LEGlyphStorage &glyphStorage);
160
161 static const FeatureMap *getFeatureMap(le_int32 &count);
162
163 static const FeatureMap *getv2FeatureMap(le_int32 &count);
164
165 static void applyPresentationForms(LEGlyphStorage &glyphStorage, le_int32 count);
166
167 static void finalReordering(LEGlyphStorage &glyphStorage, le_int32 count);
168
169 static void getDynamicProperties(DynamicProperties *dProps, const IndicClassTable *classTable);
170
171 private:
172 // do not instantiate
173 IndicReordering();
174
175 static le_int32 findSyllable(const IndicClassTable *classTable, const LEUnicode *chars, le_int32 prev, le_int32 charCount);
176
177 };
178
getWorstCaseExpansion()179 inline le_int32 IndicClassTable::getWorstCaseExpansion() const
180 {
181 return worstCaseExpansion;
182 }
183
getFilterZeroWidth()184 inline le_bool IndicClassTable::getFilterZeroWidth() const
185 {
186 return (scriptFlags & SF_FILTER_ZERO_WIDTH) != 0;
187 }
188
getSplitMatra(CharClass charClass)189 inline const SplitMatra *IndicClassTable::getSplitMatra(CharClass charClass) const
190 {
191 le_int32 index = (charClass & CF_INDEX_MASK) >> CF_INDEX_SHIFT;
192
193 return &splitMatraTable[index - 1];
194 }
195
isVowelModifier(CharClass charClass)196 inline le_bool IndicClassTable::isVowelModifier(CharClass charClass)
197 {
198 return (charClass & CF_CLASS_MASK) == CC_VOWEL_MODIFIER;
199 }
200
isStressMark(CharClass charClass)201 inline le_bool IndicClassTable::isStressMark(CharClass charClass)
202 {
203 return (charClass & CF_CLASS_MASK) == CC_STRESS_MARK;
204 }
205
isConsonant(CharClass charClass)206 inline le_bool IndicClassTable::isConsonant(CharClass charClass)
207 {
208 return (charClass & CF_CONSONANT) != 0;
209 }
210
isReph(CharClass charClass)211 inline le_bool IndicClassTable::isReph(CharClass charClass)
212 {
213 return (charClass & CF_REPH) != 0;
214 }
215
isNukta(CharClass charClass)216 inline le_bool IndicClassTable::isNukta(CharClass charClass)
217 {
218 return (charClass & CF_CLASS_MASK) == CC_NUKTA;
219 }
220
isVirama(CharClass charClass)221 inline le_bool IndicClassTable::isVirama(CharClass charClass)
222 {
223 return (charClass & CF_CLASS_MASK) == CC_VIRAMA;
224 }
225
isAlLakuna(CharClass charClass)226 inline le_bool IndicClassTable::isAlLakuna(CharClass charClass)
227 {
228 return (charClass & CF_CLASS_MASK) == CC_AL_LAKUNA;
229 }
230
isVattu(CharClass charClass)231 inline le_bool IndicClassTable::isVattu(CharClass charClass)
232 {
233 return (charClass & CF_VATTU) != 0;
234 }
235
isMatra(CharClass charClass)236 inline le_bool IndicClassTable::isMatra(CharClass charClass)
237 {
238 charClass &= CF_CLASS_MASK;
239
240 return charClass >= CC_DEPENDENT_VOWEL && charClass <= CC_SPLIT_VOWEL_PIECE_3;
241 }
242
isSplitMatra(CharClass charClass)243 inline le_bool IndicClassTable::isSplitMatra(CharClass charClass)
244 {
245 return (charClass & CF_INDEX_MASK) != 0;
246 }
247
isLengthMark(CharClass charClass)248 inline le_bool IndicClassTable::isLengthMark(CharClass charClass)
249 {
250 return (charClass & CF_LENGTH_MARK) != 0;
251 }
252
hasPostOrBelowBaseForm(CharClass charClass)253 inline le_bool IndicClassTable::hasPostOrBelowBaseForm(CharClass charClass)
254 {
255 return (charClass & (CF_POST_BASE | CF_BELOW_BASE)) != 0;
256 }
257
hasPostBaseForm(CharClass charClass)258 inline le_bool IndicClassTable::hasPostBaseForm(CharClass charClass)
259 {
260 return (charClass & CF_POST_BASE) != 0;
261 }
262
hasPreBaseForm(CharClass charClass)263 inline le_bool IndicClassTable::hasPreBaseForm(CharClass charClass)
264 {
265 return (charClass & CF_PRE_BASE) != 0;
266 }
267
hasBelowBaseForm(CharClass charClass)268 inline le_bool IndicClassTable::hasBelowBaseForm(CharClass charClass)
269 {
270 return (charClass & CF_BELOW_BASE) != 0;
271 }
272
hasAboveBaseForm(CharClass charClass)273 inline le_bool IndicClassTable::hasAboveBaseForm(CharClass charClass)
274 {
275 return ((charClass & CF_POS_MASK) == CF_POS_ABOVE);
276 }
277
isVowelModifier(LEUnicode ch)278 inline le_bool IndicClassTable::isVowelModifier(LEUnicode ch) const
279 {
280 return isVowelModifier(getCharClass(ch));
281 }
282
isStressMark(LEUnicode ch)283 inline le_bool IndicClassTable::isStressMark(LEUnicode ch) const
284 {
285 return isStressMark(getCharClass(ch));
286 }
287
isConsonant(LEUnicode ch)288 inline le_bool IndicClassTable::isConsonant(LEUnicode ch) const
289 {
290 return isConsonant(getCharClass(ch));
291 }
292
isReph(LEUnicode ch)293 inline le_bool IndicClassTable::isReph(LEUnicode ch) const
294 {
295 return isReph(getCharClass(ch));
296 }
297
isVirama(LEUnicode ch)298 inline le_bool IndicClassTable::isVirama(LEUnicode ch) const
299 {
300 return isVirama(getCharClass(ch));
301 }
302
isAlLakuna(LEUnicode ch)303 inline le_bool IndicClassTable::isAlLakuna(LEUnicode ch) const
304 {
305 return isAlLakuna(getCharClass(ch));
306 }
307
isNukta(LEUnicode ch)308 inline le_bool IndicClassTable::isNukta(LEUnicode ch) const
309 {
310 return isNukta(getCharClass(ch));
311 }
312
isVattu(LEUnicode ch)313 inline le_bool IndicClassTable::isVattu(LEUnicode ch) const
314 {
315 return isVattu(getCharClass(ch));
316 }
317
isMatra(LEUnicode ch)318 inline le_bool IndicClassTable::isMatra(LEUnicode ch) const
319 {
320 return isMatra(getCharClass(ch));
321 }
322
isSplitMatra(LEUnicode ch)323 inline le_bool IndicClassTable::isSplitMatra(LEUnicode ch) const
324 {
325 return isSplitMatra(getCharClass(ch));
326 }
327
isLengthMark(LEUnicode ch)328 inline le_bool IndicClassTable::isLengthMark(LEUnicode ch) const
329 {
330 return isLengthMark(getCharClass(ch));
331 }
332
hasPostOrBelowBaseForm(LEUnicode ch)333 inline le_bool IndicClassTable::hasPostOrBelowBaseForm(LEUnicode ch) const
334 {
335 return hasPostOrBelowBaseForm(getCharClass(ch));
336 }
337
hasPostBaseForm(LEUnicode ch)338 inline le_bool IndicClassTable::hasPostBaseForm(LEUnicode ch) const
339 {
340 return hasPostBaseForm(getCharClass(ch));
341 }
342
hasBelowBaseForm(LEUnicode ch)343 inline le_bool IndicClassTable::hasBelowBaseForm(LEUnicode ch) const
344 {
345 return hasBelowBaseForm(getCharClass(ch));
346 }
347
hasPreBaseForm(LEUnicode ch)348 inline le_bool IndicClassTable::hasPreBaseForm(LEUnicode ch) const
349 {
350 return hasPreBaseForm(getCharClass(ch));
351 }
352
hasAboveBaseForm(LEUnicode ch)353 inline le_bool IndicClassTable::hasAboveBaseForm(LEUnicode ch) const
354 {
355 return hasAboveBaseForm(getCharClass(ch));
356 }
357 U_NAMESPACE_END
358 #endif
359