1 // Copyright (C) 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4 *******************************************************************************
5 *
6 * Copyright (C) 2002-2016, International Business Machines
7 * Corporation and others. All Rights Reserved.
8 *
9 *******************************************************************************
10 * file name: uprops.cpp
11 * encoding: US-ASCII
12 * tab size: 8 (not used)
13 * indentation:4
14 *
15 * created on: 2002feb24
16 * created by: Markus W. Scherer
17 *
18 * Implementations for mostly non-core Unicode character properties
19 * stored in uprops.icu.
20 *
21 * With the APIs implemented here, almost all properties files and
22 * their associated implementation files are used from this file,
23 * including those for normalization and case mappings.
24 */
25
26 #include "unicode/utypes.h"
27 #include "unicode/uchar.h"
28 #include "unicode/unorm2.h"
29 #include "unicode/uscript.h"
30 #include "unicode/ustring.h"
31 #include "cstring.h"
32 #include "normalizer2impl.h"
33 #include "umutex.h"
34 #include "ubidi_props.h"
35 #include "uprops.h"
36 #include "ucase.h"
37 #include "ustr_imp.h"
38
39 U_NAMESPACE_USE
40
41 #define GET_BIDI_PROPS() ubidi_getSingleton()
42
43 /* general properties API functions ----------------------------------------- */
44
45 struct BinaryProperty;
46
47 typedef UBool BinaryPropertyContains(const BinaryProperty &prop, UChar32 c, UProperty which);
48
49 struct BinaryProperty {
50 int32_t column; // SRC_PROPSVEC column, or "source" if mask==0
51 uint32_t mask;
52 BinaryPropertyContains *contains;
53 };
54
defaultContains(const BinaryProperty & prop,UChar32 c,UProperty)55 static UBool defaultContains(const BinaryProperty &prop, UChar32 c, UProperty /*which*/) {
56 /* systematic, directly stored properties */
57 return (u_getUnicodeProperties(c, prop.column)&prop.mask)!=0;
58 }
59
caseBinaryPropertyContains(const BinaryProperty &,UChar32 c,UProperty which)60 static UBool caseBinaryPropertyContains(const BinaryProperty &/*prop*/, UChar32 c, UProperty which) {
61 return ucase_hasBinaryProperty(c, which);
62 }
63
isBidiControl(const BinaryProperty &,UChar32 c,UProperty)64 static UBool isBidiControl(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
65 return ubidi_isBidiControl(GET_BIDI_PROPS(), c);
66 }
67
isMirrored(const BinaryProperty &,UChar32 c,UProperty)68 static UBool isMirrored(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
69 return ubidi_isMirrored(GET_BIDI_PROPS(), c);
70 }
71
isJoinControl(const BinaryProperty &,UChar32 c,UProperty)72 static UBool isJoinControl(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
73 return ubidi_isJoinControl(GET_BIDI_PROPS(), c);
74 }
75
76 #if UCONFIG_NO_NORMALIZATION
hasFullCompositionExclusion(const BinaryProperty &,UChar32,UProperty)77 static UBool hasFullCompositionExclusion(const BinaryProperty &, UChar32, UProperty) {
78 return FALSE;
79 }
80 #else
hasFullCompositionExclusion(const BinaryProperty &,UChar32 c,UProperty)81 static UBool hasFullCompositionExclusion(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
82 // By definition, Full_Composition_Exclusion is the same as NFC_QC=No.
83 UErrorCode errorCode=U_ZERO_ERROR;
84 const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
85 return U_SUCCESS(errorCode) && impl->isCompNo(impl->getNorm16(c));
86 }
87 #endif
88
89 // UCHAR_NF*_INERT properties
90 #if UCONFIG_NO_NORMALIZATION
isNormInert(const BinaryProperty &,UChar32,UProperty)91 static UBool isNormInert(const BinaryProperty &, UChar32, UProperty) {
92 return FALSE;
93 }
94 #else
isNormInert(const BinaryProperty &,UChar32 c,UProperty which)95 static UBool isNormInert(const BinaryProperty &/*prop*/, UChar32 c, UProperty which) {
96 UErrorCode errorCode=U_ZERO_ERROR;
97 const Normalizer2 *norm2=Normalizer2Factory::getInstance(
98 (UNormalizationMode)(which-UCHAR_NFD_INERT+UNORM_NFD), errorCode);
99 return U_SUCCESS(errorCode) && norm2->isInert(c);
100 }
101 #endif
102
103 #if UCONFIG_NO_NORMALIZATION
changesWhenCasefolded(const BinaryProperty &,UChar32,UProperty)104 static UBool changesWhenCasefolded(const BinaryProperty &, UChar32, UProperty) {
105 return FALSE;
106 }
107 #else
changesWhenCasefolded(const BinaryProperty &,UChar32 c,UProperty)108 static UBool changesWhenCasefolded(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
109 UnicodeString nfd;
110 UErrorCode errorCode=U_ZERO_ERROR;
111 const Normalizer2 *nfcNorm2=Normalizer2::getNFCInstance(errorCode);
112 if(U_FAILURE(errorCode)) {
113 return FALSE;
114 }
115 if(nfcNorm2->getDecomposition(c, nfd)) {
116 /* c has a decomposition */
117 if(nfd.length()==1) {
118 c=nfd[0]; /* single BMP code point */
119 } else if(nfd.length()<=U16_MAX_LENGTH &&
120 nfd.length()==U16_LENGTH(c=nfd.char32At(0))
121 ) {
122 /* single supplementary code point */
123 } else {
124 c=U_SENTINEL;
125 }
126 } else if(c<0) {
127 return FALSE; /* protect against bad input */
128 }
129 if(c>=0) {
130 /* single code point */
131 const UCaseProps *csp=ucase_getSingleton();
132 const UChar *resultString;
133 return (UBool)(ucase_toFullFolding(csp, c, &resultString, U_FOLD_CASE_DEFAULT)>=0);
134 } else {
135 /* guess some large but stack-friendly capacity */
136 UChar dest[2*UCASE_MAX_STRING_LENGTH];
137 int32_t destLength;
138 destLength=u_strFoldCase(dest, UPRV_LENGTHOF(dest),
139 nfd.getBuffer(), nfd.length(),
140 U_FOLD_CASE_DEFAULT, &errorCode);
141 return (UBool)(U_SUCCESS(errorCode) &&
142 0!=u_strCompare(nfd.getBuffer(), nfd.length(),
143 dest, destLength, FALSE));
144 }
145 }
146 #endif
147
148 #if UCONFIG_NO_NORMALIZATION
changesWhenNFKC_Casefolded(const BinaryProperty &,UChar32,UProperty)149 static UBool changesWhenNFKC_Casefolded(const BinaryProperty &, UChar32, UProperty) {
150 return FALSE;
151 }
152 #else
changesWhenNFKC_Casefolded(const BinaryProperty &,UChar32 c,UProperty)153 static UBool changesWhenNFKC_Casefolded(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
154 UErrorCode errorCode=U_ZERO_ERROR;
155 const Normalizer2Impl *kcf=Normalizer2Factory::getNFKC_CFImpl(errorCode);
156 if(U_FAILURE(errorCode)) {
157 return FALSE;
158 }
159 UnicodeString src(c);
160 UnicodeString dest;
161 {
162 // The ReorderingBuffer must be in a block because its destructor
163 // needs to release dest's buffer before we look at its contents.
164 ReorderingBuffer buffer(*kcf, dest);
165 // Small destCapacity for NFKC_CF(c).
166 if(buffer.init(5, errorCode)) {
167 const UChar *srcArray=src.getBuffer();
168 kcf->compose(srcArray, srcArray+src.length(), FALSE,
169 TRUE, buffer, errorCode);
170 }
171 }
172 return U_SUCCESS(errorCode) && dest!=src;
173 }
174 #endif
175
176 #if UCONFIG_NO_NORMALIZATION
isCanonSegmentStarter(const BinaryProperty &,UChar32,UProperty)177 static UBool isCanonSegmentStarter(const BinaryProperty &, UChar32, UProperty) {
178 return FALSE;
179 }
180 #else
isCanonSegmentStarter(const BinaryProperty &,UChar32 c,UProperty)181 static UBool isCanonSegmentStarter(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
182 UErrorCode errorCode=U_ZERO_ERROR;
183 const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
184 return
185 U_SUCCESS(errorCode) && impl->ensureCanonIterData(errorCode) &&
186 impl->isCanonSegmentStarter(c);
187 }
188 #endif
189
isPOSIX_alnum(const BinaryProperty &,UChar32 c,UProperty)190 static UBool isPOSIX_alnum(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
191 return u_isalnumPOSIX(c);
192 }
193
isPOSIX_blank(const BinaryProperty &,UChar32 c,UProperty)194 static UBool isPOSIX_blank(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
195 return u_isblank(c);
196 }
197
isPOSIX_graph(const BinaryProperty &,UChar32 c,UProperty)198 static UBool isPOSIX_graph(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
199 return u_isgraphPOSIX(c);
200 }
201
isPOSIX_print(const BinaryProperty &,UChar32 c,UProperty)202 static UBool isPOSIX_print(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
203 return u_isprintPOSIX(c);
204 }
205
isPOSIX_xdigit(const BinaryProperty &,UChar32 c,UProperty)206 static UBool isPOSIX_xdigit(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
207 return u_isxdigit(c);
208 }
209
210 static const BinaryProperty binProps[UCHAR_BINARY_LIMIT]={
211 /*
212 * column and mask values for binary properties from u_getUnicodeProperties().
213 * Must be in order of corresponding UProperty,
214 * and there must be exactly one entry per binary UProperty.
215 *
216 * Properties with mask==0 are handled in code.
217 * For them, column is the UPropertySource value.
218 */
219 { 1, U_MASK(UPROPS_ALPHABETIC), defaultContains },
220 { 1, U_MASK(UPROPS_ASCII_HEX_DIGIT), defaultContains },
221 { UPROPS_SRC_BIDI, 0, isBidiControl },
222 { UPROPS_SRC_BIDI, 0, isMirrored },
223 { 1, U_MASK(UPROPS_DASH), defaultContains },
224 { 1, U_MASK(UPROPS_DEFAULT_IGNORABLE_CODE_POINT), defaultContains },
225 { 1, U_MASK(UPROPS_DEPRECATED), defaultContains },
226 { 1, U_MASK(UPROPS_DIACRITIC), defaultContains },
227 { 1, U_MASK(UPROPS_EXTENDER), defaultContains },
228 { UPROPS_SRC_NFC, 0, hasFullCompositionExclusion },
229 { 1, U_MASK(UPROPS_GRAPHEME_BASE), defaultContains },
230 { 1, U_MASK(UPROPS_GRAPHEME_EXTEND), defaultContains },
231 { 1, U_MASK(UPROPS_GRAPHEME_LINK), defaultContains },
232 { 1, U_MASK(UPROPS_HEX_DIGIT), defaultContains },
233 { 1, U_MASK(UPROPS_HYPHEN), defaultContains },
234 { 1, U_MASK(UPROPS_ID_CONTINUE), defaultContains },
235 { 1, U_MASK(UPROPS_ID_START), defaultContains },
236 { 1, U_MASK(UPROPS_IDEOGRAPHIC), defaultContains },
237 { 1, U_MASK(UPROPS_IDS_BINARY_OPERATOR), defaultContains },
238 { 1, U_MASK(UPROPS_IDS_TRINARY_OPERATOR), defaultContains },
239 { UPROPS_SRC_BIDI, 0, isJoinControl },
240 { 1, U_MASK(UPROPS_LOGICAL_ORDER_EXCEPTION), defaultContains },
241 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_LOWERCASE
242 { 1, U_MASK(UPROPS_MATH), defaultContains },
243 { 1, U_MASK(UPROPS_NONCHARACTER_CODE_POINT), defaultContains },
244 { 1, U_MASK(UPROPS_QUOTATION_MARK), defaultContains },
245 { 1, U_MASK(UPROPS_RADICAL), defaultContains },
246 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_SOFT_DOTTED
247 { 1, U_MASK(UPROPS_TERMINAL_PUNCTUATION), defaultContains },
248 { 1, U_MASK(UPROPS_UNIFIED_IDEOGRAPH), defaultContains },
249 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_UPPERCASE
250 { 1, U_MASK(UPROPS_WHITE_SPACE), defaultContains },
251 { 1, U_MASK(UPROPS_XID_CONTINUE), defaultContains },
252 { 1, U_MASK(UPROPS_XID_START), defaultContains },
253 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_CASE_SENSITIVE
254 { 1, U_MASK(UPROPS_S_TERM), defaultContains },
255 { 1, U_MASK(UPROPS_VARIATION_SELECTOR), defaultContains },
256 { UPROPS_SRC_NFC, 0, isNormInert }, // UCHAR_NFD_INERT
257 { UPROPS_SRC_NFKC, 0, isNormInert }, // UCHAR_NFKD_INERT
258 { UPROPS_SRC_NFC, 0, isNormInert }, // UCHAR_NFC_INERT
259 { UPROPS_SRC_NFKC, 0, isNormInert }, // UCHAR_NFKC_INERT
260 { UPROPS_SRC_NFC_CANON_ITER, 0, isCanonSegmentStarter },
261 { 1, U_MASK(UPROPS_PATTERN_SYNTAX), defaultContains },
262 { 1, U_MASK(UPROPS_PATTERN_WHITE_SPACE), defaultContains },
263 { UPROPS_SRC_CHAR_AND_PROPSVEC, 0, isPOSIX_alnum },
264 { UPROPS_SRC_CHAR, 0, isPOSIX_blank },
265 { UPROPS_SRC_CHAR, 0, isPOSIX_graph },
266 { UPROPS_SRC_CHAR, 0, isPOSIX_print },
267 { UPROPS_SRC_CHAR, 0, isPOSIX_xdigit },
268 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_CASED
269 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_CASE_IGNORABLE
270 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_CHANGES_WHEN_LOWERCASED
271 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_CHANGES_WHEN_UPPERCASED
272 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_CHANGES_WHEN_TITLECASED
273 { UPROPS_SRC_CASE_AND_NORM, 0, changesWhenCasefolded },
274 { UPROPS_SRC_CASE, 0, caseBinaryPropertyContains }, // UCHAR_CHANGES_WHEN_CASEMAPPED
275 { UPROPS_SRC_NFKC_CF, 0, changesWhenNFKC_Casefolded },
276 { 2, U_MASK(UPROPS_2_EMOJI), defaultContains },
277 { 2, U_MASK(UPROPS_2_EMOJI_PRESENTATION), defaultContains },
278 { 2, U_MASK(UPROPS_2_EMOJI_MODIFIER), defaultContains },
279 { 2, U_MASK(UPROPS_2_EMOJI_MODIFIER_BASE), defaultContains },
280 };
281
282 U_CAPI UBool U_EXPORT2
u_hasBinaryProperty(UChar32 c,UProperty which)283 u_hasBinaryProperty(UChar32 c, UProperty which) {
284 /* c is range-checked in the functions that are called from here */
285 if(which<UCHAR_BINARY_START || UCHAR_BINARY_LIMIT<=which) {
286 /* not a known binary property */
287 return FALSE;
288 } else {
289 const BinaryProperty &prop=binProps[which];
290 return prop.contains(prop, c, which);
291 }
292 }
293
294 struct IntProperty;
295
296 typedef int32_t IntPropertyGetValue(const IntProperty &prop, UChar32 c, UProperty which);
297 typedef int32_t IntPropertyGetMaxValue(const IntProperty &prop, UProperty which);
298
299 struct IntProperty {
300 int32_t column; // SRC_PROPSVEC column, or "source" if mask==0
301 uint32_t mask;
302 int32_t shift; // =maxValue if getMaxValueFromShift() is used
303 IntPropertyGetValue *getValue;
304 IntPropertyGetMaxValue *getMaxValue;
305 };
306
defaultGetValue(const IntProperty & prop,UChar32 c,UProperty)307 static int32_t defaultGetValue(const IntProperty &prop, UChar32 c, UProperty /*which*/) {
308 /* systematic, directly stored properties */
309 return (int32_t)(u_getUnicodeProperties(c, prop.column)&prop.mask)>>prop.shift;
310 }
311
defaultGetMaxValue(const IntProperty & prop,UProperty)312 static int32_t defaultGetMaxValue(const IntProperty &prop, UProperty /*which*/) {
313 return (uprv_getMaxValues(prop.column)&prop.mask)>>prop.shift;
314 }
315
getMaxValueFromShift(const IntProperty & prop,UProperty)316 static int32_t getMaxValueFromShift(const IntProperty &prop, UProperty /*which*/) {
317 return prop.shift;
318 }
319
getBiDiClass(const IntProperty &,UChar32 c,UProperty)320 static int32_t getBiDiClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
321 return (int32_t)u_charDirection(c);
322 }
323
getBiDiPairedBracketType(const IntProperty &,UChar32 c,UProperty)324 static int32_t getBiDiPairedBracketType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
325 return (int32_t)ubidi_getPairedBracketType(GET_BIDI_PROPS(), c);
326 }
327
biDiGetMaxValue(const IntProperty &,UProperty which)328 static int32_t biDiGetMaxValue(const IntProperty &/*prop*/, UProperty which) {
329 return ubidi_getMaxValue(GET_BIDI_PROPS(), which);
330 }
331
332 #if UCONFIG_NO_NORMALIZATION
getCombiningClass(const IntProperty &,UChar32,UProperty)333 static int32_t getCombiningClass(const IntProperty &, UChar32, UProperty) {
334 return 0;
335 }
336 #else
getCombiningClass(const IntProperty &,UChar32 c,UProperty)337 static int32_t getCombiningClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
338 return u_getCombiningClass(c);
339 }
340 #endif
341
getGeneralCategory(const IntProperty &,UChar32 c,UProperty)342 static int32_t getGeneralCategory(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
343 return (int32_t)u_charType(c);
344 }
345
getJoiningGroup(const IntProperty &,UChar32 c,UProperty)346 static int32_t getJoiningGroup(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
347 return ubidi_getJoiningGroup(GET_BIDI_PROPS(), c);
348 }
349
getJoiningType(const IntProperty &,UChar32 c,UProperty)350 static int32_t getJoiningType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
351 return ubidi_getJoiningType(GET_BIDI_PROPS(), c);
352 }
353
getNumericType(const IntProperty &,UChar32 c,UProperty)354 static int32_t getNumericType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
355 int32_t ntv=(int32_t)GET_NUMERIC_TYPE_VALUE(u_getMainProperties(c));
356 return UPROPS_NTV_GET_TYPE(ntv);
357 }
358
getScript(const IntProperty &,UChar32 c,UProperty)359 static int32_t getScript(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
360 UErrorCode errorCode=U_ZERO_ERROR;
361 return (int32_t)uscript_getScript(c, &errorCode);
362 }
363
364 /*
365 * Map some of the Grapheme Cluster Break values to Hangul Syllable Types.
366 * Hangul_Syllable_Type is fully redundant with a subset of Grapheme_Cluster_Break.
367 */
368 static const UHangulSyllableType gcbToHst[]={
369 U_HST_NOT_APPLICABLE, /* U_GCB_OTHER */
370 U_HST_NOT_APPLICABLE, /* U_GCB_CONTROL */
371 U_HST_NOT_APPLICABLE, /* U_GCB_CR */
372 U_HST_NOT_APPLICABLE, /* U_GCB_EXTEND */
373 U_HST_LEADING_JAMO, /* U_GCB_L */
374 U_HST_NOT_APPLICABLE, /* U_GCB_LF */
375 U_HST_LV_SYLLABLE, /* U_GCB_LV */
376 U_HST_LVT_SYLLABLE, /* U_GCB_LVT */
377 U_HST_TRAILING_JAMO, /* U_GCB_T */
378 U_HST_VOWEL_JAMO /* U_GCB_V */
379 /*
380 * Omit GCB values beyond what we need for hst.
381 * The code below checks for the array length.
382 */
383 };
384
getHangulSyllableType(const IntProperty &,UChar32 c,UProperty)385 static int32_t getHangulSyllableType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
386 /* see comments on gcbToHst[] above */
387 int32_t gcb=(int32_t)(u_getUnicodeProperties(c, 2)&UPROPS_GCB_MASK)>>UPROPS_GCB_SHIFT;
388 if(gcb<UPRV_LENGTHOF(gcbToHst)) {
389 return gcbToHst[gcb];
390 } else {
391 return U_HST_NOT_APPLICABLE;
392 }
393 }
394
395 #if UCONFIG_NO_NORMALIZATION
getNormQuickCheck(const IntProperty &,UChar32,UProperty)396 static int32_t getNormQuickCheck(const IntProperty &, UChar32, UProperty) {
397 return 0;
398 }
399 #else
getNormQuickCheck(const IntProperty &,UChar32 c,UProperty which)400 static int32_t getNormQuickCheck(const IntProperty &/*prop*/, UChar32 c, UProperty which) {
401 return (int32_t)unorm_getQuickCheck(c, (UNormalizationMode)(which-UCHAR_NFD_QUICK_CHECK+UNORM_NFD));
402 }
403 #endif
404
405 #if UCONFIG_NO_NORMALIZATION
getLeadCombiningClass(const IntProperty &,UChar32,UProperty)406 static int32_t getLeadCombiningClass(const IntProperty &, UChar32, UProperty) {
407 return 0;
408 }
409 #else
getLeadCombiningClass(const IntProperty &,UChar32 c,UProperty)410 static int32_t getLeadCombiningClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
411 return unorm_getFCD16(c)>>8;
412 }
413 #endif
414
415 #if UCONFIG_NO_NORMALIZATION
getTrailCombiningClass(const IntProperty &,UChar32,UProperty)416 static int32_t getTrailCombiningClass(const IntProperty &, UChar32, UProperty) {
417 return 0;
418 }
419 #else
getTrailCombiningClass(const IntProperty &,UChar32 c,UProperty)420 static int32_t getTrailCombiningClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
421 return unorm_getFCD16(c)&0xff;
422 }
423 #endif
424
425 static const IntProperty intProps[UCHAR_INT_LIMIT-UCHAR_INT_START]={
426 /*
427 * column, mask and shift values for int-value properties from u_getUnicodeProperties().
428 * Must be in order of corresponding UProperty,
429 * and there must be exactly one entry per int UProperty.
430 *
431 * Properties with mask==0 are handled in code.
432 * For them, column is the UPropertySource value.
433 */
434 { UPROPS_SRC_BIDI, 0, 0, getBiDiClass, biDiGetMaxValue },
435 { 0, UPROPS_BLOCK_MASK, UPROPS_BLOCK_SHIFT, defaultGetValue, defaultGetMaxValue },
436 { UPROPS_SRC_NFC, 0, 0xff, getCombiningClass, getMaxValueFromShift },
437 { 2, UPROPS_DT_MASK, 0, defaultGetValue, defaultGetMaxValue },
438 { 0, UPROPS_EA_MASK, UPROPS_EA_SHIFT, defaultGetValue, defaultGetMaxValue },
439 { UPROPS_SRC_CHAR, 0, (int32_t)U_CHAR_CATEGORY_COUNT-1,getGeneralCategory, getMaxValueFromShift },
440 { UPROPS_SRC_BIDI, 0, 0, getJoiningGroup, biDiGetMaxValue },
441 { UPROPS_SRC_BIDI, 0, 0, getJoiningType, biDiGetMaxValue },
442 { 2, UPROPS_LB_MASK, UPROPS_LB_SHIFT, defaultGetValue, defaultGetMaxValue },
443 { UPROPS_SRC_CHAR, 0, (int32_t)U_NT_COUNT-1, getNumericType, getMaxValueFromShift },
444 { 0, UPROPS_SCRIPT_MASK, 0, getScript, defaultGetMaxValue },
445 { UPROPS_SRC_PROPSVEC, 0, (int32_t)U_HST_COUNT-1, getHangulSyllableType, getMaxValueFromShift },
446 // UCHAR_NFD_QUICK_CHECK: max=1=YES -- never "maybe", only "no" or "yes"
447 { UPROPS_SRC_NFC, 0, (int32_t)UNORM_YES, getNormQuickCheck, getMaxValueFromShift },
448 // UCHAR_NFKD_QUICK_CHECK: max=1=YES -- never "maybe", only "no" or "yes"
449 { UPROPS_SRC_NFKC, 0, (int32_t)UNORM_YES, getNormQuickCheck, getMaxValueFromShift },
450 // UCHAR_NFC_QUICK_CHECK: max=2=MAYBE
451 { UPROPS_SRC_NFC, 0, (int32_t)UNORM_MAYBE, getNormQuickCheck, getMaxValueFromShift },
452 // UCHAR_NFKC_QUICK_CHECK: max=2=MAYBE
453 { UPROPS_SRC_NFKC, 0, (int32_t)UNORM_MAYBE, getNormQuickCheck, getMaxValueFromShift },
454 { UPROPS_SRC_NFC, 0, 0xff, getLeadCombiningClass, getMaxValueFromShift },
455 { UPROPS_SRC_NFC, 0, 0xff, getTrailCombiningClass, getMaxValueFromShift },
456 { 2, UPROPS_GCB_MASK, UPROPS_GCB_SHIFT, defaultGetValue, defaultGetMaxValue },
457 { 2, UPROPS_SB_MASK, UPROPS_SB_SHIFT, defaultGetValue, defaultGetMaxValue },
458 { 2, UPROPS_WB_MASK, UPROPS_WB_SHIFT, defaultGetValue, defaultGetMaxValue },
459 { UPROPS_SRC_BIDI, 0, 0, getBiDiPairedBracketType, biDiGetMaxValue },
460 };
461
462 U_CAPI int32_t U_EXPORT2
u_getIntPropertyValue(UChar32 c,UProperty which)463 u_getIntPropertyValue(UChar32 c, UProperty which) {
464 if(which<UCHAR_INT_START) {
465 if(UCHAR_BINARY_START<=which && which<UCHAR_BINARY_LIMIT) {
466 const BinaryProperty &prop=binProps[which];
467 return prop.contains(prop, c, which);
468 }
469 } else if(which<UCHAR_INT_LIMIT) {
470 const IntProperty &prop=intProps[which-UCHAR_INT_START];
471 return prop.getValue(prop, c, which);
472 } else if(which==UCHAR_GENERAL_CATEGORY_MASK) {
473 return U_MASK(u_charType(c));
474 }
475 return 0; // undefined
476 }
477
478 U_CAPI int32_t U_EXPORT2
u_getIntPropertyMinValue(UProperty)479 u_getIntPropertyMinValue(UProperty /*which*/) {
480 return 0; /* all binary/enum/int properties have a minimum value of 0 */
481 }
482
483 U_CAPI int32_t U_EXPORT2
u_getIntPropertyMaxValue(UProperty which)484 u_getIntPropertyMaxValue(UProperty which) {
485 if(which<UCHAR_INT_START) {
486 if(UCHAR_BINARY_START<=which && which<UCHAR_BINARY_LIMIT) {
487 return 1; // maximum TRUE for all binary properties
488 }
489 } else if(which<UCHAR_INT_LIMIT) {
490 const IntProperty &prop=intProps[which-UCHAR_INT_START];
491 return prop.getMaxValue(prop, which);
492 }
493 return -1; // undefined
494 }
495
496 U_CFUNC UPropertySource U_EXPORT2
uprops_getSource(UProperty which)497 uprops_getSource(UProperty which) {
498 if(which<UCHAR_BINARY_START) {
499 return UPROPS_SRC_NONE; /* undefined */
500 } else if(which<UCHAR_BINARY_LIMIT) {
501 const BinaryProperty &prop=binProps[which];
502 if(prop.mask!=0) {
503 return UPROPS_SRC_PROPSVEC;
504 } else {
505 return (UPropertySource)prop.column;
506 }
507 } else if(which<UCHAR_INT_START) {
508 return UPROPS_SRC_NONE; /* undefined */
509 } else if(which<UCHAR_INT_LIMIT) {
510 const IntProperty &prop=intProps[which-UCHAR_INT_START];
511 if(prop.mask!=0) {
512 return UPROPS_SRC_PROPSVEC;
513 } else {
514 return (UPropertySource)prop.column;
515 }
516 } else if(which<UCHAR_STRING_START) {
517 switch(which) {
518 case UCHAR_GENERAL_CATEGORY_MASK:
519 case UCHAR_NUMERIC_VALUE:
520 return UPROPS_SRC_CHAR;
521
522 default:
523 return UPROPS_SRC_NONE;
524 }
525 } else if(which<UCHAR_STRING_LIMIT) {
526 switch(which) {
527 case UCHAR_AGE:
528 return UPROPS_SRC_PROPSVEC;
529
530 case UCHAR_BIDI_MIRRORING_GLYPH:
531 return UPROPS_SRC_BIDI;
532
533 case UCHAR_CASE_FOLDING:
534 case UCHAR_LOWERCASE_MAPPING:
535 case UCHAR_SIMPLE_CASE_FOLDING:
536 case UCHAR_SIMPLE_LOWERCASE_MAPPING:
537 case UCHAR_SIMPLE_TITLECASE_MAPPING:
538 case UCHAR_SIMPLE_UPPERCASE_MAPPING:
539 case UCHAR_TITLECASE_MAPPING:
540 case UCHAR_UPPERCASE_MAPPING:
541 return UPROPS_SRC_CASE;
542
543 case UCHAR_ISO_COMMENT:
544 case UCHAR_NAME:
545 case UCHAR_UNICODE_1_NAME:
546 return UPROPS_SRC_NAMES;
547
548 default:
549 return UPROPS_SRC_NONE;
550 }
551 } else {
552 switch(which) {
553 case UCHAR_SCRIPT_EXTENSIONS:
554 return UPROPS_SRC_PROPSVEC;
555 default:
556 return UPROPS_SRC_NONE; /* undefined */
557 }
558 }
559 }
560
561 #if !UCONFIG_NO_NORMALIZATION
562
563 U_CAPI int32_t U_EXPORT2
u_getFC_NFKC_Closure(UChar32 c,UChar * dest,int32_t destCapacity,UErrorCode * pErrorCode)564 u_getFC_NFKC_Closure(UChar32 c, UChar *dest, int32_t destCapacity, UErrorCode *pErrorCode) {
565 if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
566 return 0;
567 }
568 if(destCapacity<0 || (dest==NULL && destCapacity>0)) {
569 *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
570 return 0;
571 }
572 // Compute the FC_NFKC_Closure on the fly:
573 // We have the API for complete coverage of Unicode properties, although
574 // this value by itself is not useful via API.
575 // (What could be useful is a custom normalization table that combines
576 // case folding and NFKC.)
577 // For the derivation, see Unicode's DerivedNormalizationProps.txt.
578 const Normalizer2 *nfkc=Normalizer2::getNFKCInstance(*pErrorCode);
579 const UCaseProps *csp=ucase_getSingleton();
580 if(U_FAILURE(*pErrorCode)) {
581 return 0;
582 }
583 // first: b = NFKC(Fold(a))
584 UnicodeString folded1String;
585 const UChar *folded1;
586 int32_t folded1Length=ucase_toFullFolding(csp, c, &folded1, U_FOLD_CASE_DEFAULT);
587 if(folded1Length<0) {
588 const Normalizer2Impl *nfkcImpl=Normalizer2Factory::getImpl(nfkc);
589 if(nfkcImpl->getCompQuickCheck(nfkcImpl->getNorm16(c))!=UNORM_NO) {
590 return u_terminateUChars(dest, destCapacity, 0, pErrorCode); // c does not change at all under CaseFolding+NFKC
591 }
592 folded1String.setTo(c);
593 } else {
594 if(folded1Length>UCASE_MAX_STRING_LENGTH) {
595 folded1String.setTo(folded1Length);
596 } else {
597 folded1String.setTo(FALSE, folded1, folded1Length);
598 }
599 }
600 UnicodeString kc1=nfkc->normalize(folded1String, *pErrorCode);
601 // second: c = NFKC(Fold(b))
602 UnicodeString folded2String(kc1);
603 UnicodeString kc2=nfkc->normalize(folded2String.foldCase(), *pErrorCode);
604 // if (c != b) add the mapping from a to c
605 if(U_FAILURE(*pErrorCode) || kc1==kc2) {
606 return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
607 } else {
608 return kc2.extract(dest, destCapacity, *pErrorCode);
609 }
610 }
611
612 #endif
613