ucase.h revision c69afcec261fc345fda8daf46f0ea6b4351dc777
1/* 2******************************************************************************* 3* 4* Copyright (C) 2004-2008, International Business Machines 5* Corporation and others. All Rights Reserved. 6* 7******************************************************************************* 8* file name: ucase.h 9* encoding: US-ASCII 10* tab size: 8 (not used) 11* indentation:4 12* 13* created on: 2004aug30 14* created by: Markus W. Scherer 15* 16* Low-level Unicode character/string case mapping code. 17*/ 18 19#ifndef __UCASE_H__ 20#define __UCASE_H__ 21 22#include "unicode/utypes.h" 23#include "unicode/uset.h" 24#include "uset_imp.h" 25#include "udataswp.h" 26 27U_CDECL_BEGIN 28 29/* library API -------------------------------------------------------------- */ 30 31struct UCaseProps; 32typedef struct UCaseProps UCaseProps; 33 34U_CAPI UCaseProps * U_EXPORT2 35ucase_open(UErrorCode *pErrorCode); 36 37U_CAPI UCaseProps * U_EXPORT2 38ucase_openBinary(const uint8_t *bin, int32_t length, UErrorCode *pErrorCode); 39 40U_CAPI void U_EXPORT2 41ucase_close(UCaseProps *csp); 42 43 44U_CAPI const UCaseProps * U_EXPORT2 45ucase_getSingleton(UErrorCode *pErrorCode); 46 47#define UCASE_HARDCODE_DATA 1 48 49#if !UCASE_HARDCODE_DATA 50/** 51 * Get a singleton dummy object, one that works with no real data. 52 * This can be used when the real data is not available. 53 * Using the dummy can reduce checks for available data after an initial failure. 54 */ 55U_CAPI const UCaseProps * U_EXPORT2 56ucase_getDummy(UErrorCode *pErrorCode); 57#endif 58 59 60U_CAPI int32_t U_EXPORT2 61ucase_swap(const UDataSwapper *ds, 62 const void *inData, int32_t length, void *outData, 63 UErrorCode *pErrorCode); 64 65U_CFUNC void U_EXPORT2 66ucase_addPropertyStarts(const UCaseProps *csp, const USetAdder *sa, UErrorCode *pErrorCode); 67 68/** 69 * Requires non-NULL locale ID but otherwise does the equivalent of 70 * checking for language codes as if uloc_getLanguage() were called: 71 * Accepts both 2- and 3-letter codes and accepts case variants. 72 */ 73U_CFUNC int32_t 74ucase_getCaseLocale(const char *locale, int32_t *locCache); 75 76/* Casing locale types for ucase_getCaseLocale */ 77enum { 78 UCASE_LOC_UNKNOWN, 79 UCASE_LOC_ROOT, 80 UCASE_LOC_TURKISH, 81 UCASE_LOC_LITHUANIAN, 82 UCASE_LOC_DUTCH 83}; 84 85/** 86 * Bit mask for getting just the options from a string compare options word 87 * that are relevant for case-insensitive string comparison. 88 * See uchar.h. Also include _STRNCMP_STYLE and U_COMPARE_CODE_POINT_ORDER. 89 * @internal 90 */ 91#define _STRCASECMP_OPTIONS_MASK 0xffff 92 93/** 94 * Bit mask for getting just the options from a string compare options word 95 * that are relevant for case folding (of a single string or code point). 96 * See uchar.h. 97 * @internal 98 */ 99#define _FOLD_CASE_OPTIONS_MASK 0xff 100 101/* single-code point functions */ 102 103U_CAPI UChar32 U_EXPORT2 104ucase_tolower(const UCaseProps *csp, UChar32 c); 105 106U_CAPI UChar32 U_EXPORT2 107ucase_toupper(const UCaseProps *csp, UChar32 c); 108 109U_CAPI UChar32 U_EXPORT2 110ucase_totitle(const UCaseProps *csp, UChar32 c); 111 112U_CAPI UChar32 U_EXPORT2 113ucase_fold(const UCaseProps *csp, UChar32 c, uint32_t options); 114 115/** 116 * Adds all simple case mappings and the full case folding for c to sa, 117 * and also adds special case closure mappings. 118 * c itself is not added. 119 * For example, the mappings 120 * - for s include long s 121 * - for sharp s include ss 122 * - for k include the Kelvin sign 123 */ 124U_CFUNC void U_EXPORT2 125ucase_addCaseClosure(const UCaseProps *csp, UChar32 c, const USetAdder *sa); 126 127/** 128 * Maps the string to single code points and adds the associated case closure 129 * mappings. 130 * The string is mapped to code points if it is their full case folding string. 131 * In other words, this performs a reverse full case folding and then 132 * adds the case closure items of the resulting code points. 133 * If the string is found and its closure applied, then 134 * the string itself is added as well as part of its code points' closure. 135 * It must be length>=0. 136 * 137 * @return TRUE if the string was found 138 */ 139U_CFUNC UBool U_EXPORT2 140ucase_addStringCaseClosure(const UCaseProps *csp, const UChar *s, int32_t length, const USetAdder *sa); 141 142/** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */ 143U_CAPI int32_t U_EXPORT2 144ucase_getType(const UCaseProps *csp, UChar32 c); 145 146/** @return same as ucase_getType(), or <0 if c is case-ignorable */ 147U_CAPI int32_t U_EXPORT2 148ucase_getTypeOrIgnorable(const UCaseProps *csp, UChar32 c); 149 150U_CAPI UBool U_EXPORT2 151ucase_isSoftDotted(const UCaseProps *csp, UChar32 c); 152 153U_CAPI UBool U_EXPORT2 154ucase_isCaseSensitive(const UCaseProps *csp, UChar32 c); 155 156/* string case mapping functions */ 157 158/** 159 * Iterator function for string case mappings, which need to look at the 160 * context (surrounding text) of a given character for conditional mappings. 161 * 162 * The iterator only needs to go backward or forward away from the 163 * character in question. It does not use any indexes on this interface. 164 * It does not support random access or an arbitrary change of 165 * iteration direction. 166 * 167 * The code point being case-mapped itself is never returned by 168 * this iterator. 169 * 170 * @param context A pointer to the iterator's working data. 171 * @param dir If <0 then start iterating backward from the character; 172 * if >0 then start iterating forward from the character; 173 * if 0 then continue iterating in the current direction. 174 * @return Next code point, or <0 when the iteration is done. 175 */ 176typedef UChar32 U_CALLCONV 177UCaseContextIterator(void *context, int8_t dir); 178 179/** 180 * Sample struct which may be used by some implementations of 181 * UCaseContextIterator. 182 */ 183struct UCaseContext { 184 void *p; 185 int32_t start, index, limit; 186 int32_t cpStart, cpLimit; 187 int8_t dir; 188 int8_t b1, b2, b3; 189}; 190typedef struct UCaseContext UCaseContext; 191 192enum { 193 /** 194 * For string case mappings, a single character (a code point) is mapped 195 * either to itself (in which case in-place mapping functions do nothing), 196 * or to another single code point, or to a string. 197 * Aside from the string contents, these are indicated with a single int32_t 198 * value as follows: 199 * 200 * Mapping to self: Negative values (~self instead of -self to support U+0000) 201 * 202 * Mapping to another code point: Positive values >UCASE_MAX_STRING_LENGTH 203 * 204 * Mapping to a string: The string length (0..UCASE_MAX_STRING_LENGTH) is 205 * returned. Note that the string result may indeed have zero length. 206 */ 207 UCASE_MAX_STRING_LENGTH=0x1f 208}; 209 210/** 211 * Get the full lowercase mapping for c. 212 * 213 * @param csp Case mapping properties. 214 * @param c Character to be mapped. 215 * @param iter Character iterator, used for context-sensitive mappings. 216 * See UCaseContextIterator for details. 217 * If iter==NULL then a context-independent result is returned. 218 * @param context Pointer to be passed into iter. 219 * @param pString If the mapping result is a string, then the pointer is 220 * written to *pString. 221 * @param locale Locale ID for locale-dependent mappings. 222 * @param locCache Initialize to 0; may be used to cache the result of parsing 223 * the locale ID for subsequent calls. 224 * Can be NULL. 225 * @return Output code point or string length, see UCASE_MAX_STRING_LENGTH. 226 * 227 * @see UCaseContextIterator 228 * @see UCASE_MAX_STRING_LENGTH 229 * @internal 230 */ 231U_CAPI int32_t U_EXPORT2 232ucase_toFullLower(const UCaseProps *csp, UChar32 c, 233 UCaseContextIterator *iter, void *context, 234 const UChar **pString, 235 const char *locale, int32_t *locCache); 236 237U_CAPI int32_t U_EXPORT2 238ucase_toFullUpper(const UCaseProps *csp, UChar32 c, 239 UCaseContextIterator *iter, void *context, 240 const UChar **pString, 241 const char *locale, int32_t *locCache); 242 243U_CAPI int32_t U_EXPORT2 244ucase_toFullTitle(const UCaseProps *csp, UChar32 c, 245 UCaseContextIterator *iter, void *context, 246 const UChar **pString, 247 const char *locale, int32_t *locCache); 248 249U_CAPI int32_t U_EXPORT2 250ucase_toFullFolding(const UCaseProps *csp, UChar32 c, 251 const UChar **pString, 252 uint32_t options); 253 254U_CFUNC int32_t U_EXPORT2 255ucase_hasBinaryProperty(UChar32 c, UProperty which); 256 257 258U_CDECL_BEGIN 259 260/** 261 * @internal 262 */ 263typedef int32_t U_CALLCONV 264UCaseMapFull(const UCaseProps *csp, UChar32 c, 265 UCaseContextIterator *iter, void *context, 266 const UChar **pString, 267 const char *locale, int32_t *locCache); 268 269U_CDECL_END 270 271/* file definitions --------------------------------------------------------- */ 272 273#define UCASE_DATA_NAME "ucase" 274#define UCASE_DATA_TYPE "icu" 275 276/* format "cAsE" */ 277#define UCASE_FMT_0 0x63 278#define UCASE_FMT_1 0x41 279#define UCASE_FMT_2 0x53 280#define UCASE_FMT_3 0x45 281 282/* indexes into indexes[] */ 283enum { 284 UCASE_IX_INDEX_TOP, 285 UCASE_IX_LENGTH, 286 UCASE_IX_TRIE_SIZE, 287 UCASE_IX_EXC_LENGTH, 288 UCASE_IX_UNFOLD_LENGTH, 289 290 UCASE_IX_MAX_FULL_LENGTH=15, 291 UCASE_IX_TOP=16 292}; 293 294/* definitions for 16-bit case properties word ------------------------------ */ 295 296/* 2-bit constants for types of cased characters */ 297#define UCASE_TYPE_MASK 3 298enum { 299 UCASE_NONE, 300 UCASE_LOWER, 301 UCASE_UPPER, 302 UCASE_TITLE 303}; 304 305#define UCASE_GET_TYPE(props) ((props)&UCASE_TYPE_MASK) 306 307#define UCASE_SENSITIVE 4 308#define UCASE_EXCEPTION 8 309 310#define UCASE_DOT_MASK 0x30 311enum { 312 UCASE_NO_DOT=0, /* normal characters with cc=0 */ 313 UCASE_SOFT_DOTTED=0x10, /* soft-dotted characters with cc=0 */ 314 UCASE_ABOVE=0x20, /* "above" accents with cc=230 */ 315 UCASE_OTHER_ACCENT=0x30 /* other accent character (0<cc!=230) */ 316}; 317 318/* no exception: bits 15..6 are a 10-bit signed case mapping delta */ 319#define UCASE_DELTA_SHIFT 6 320#define UCASE_DELTA_MASK 0xffc0 321#define UCASE_MAX_DELTA 0x1ff 322#define UCASE_MIN_DELTA (-UCASE_MAX_DELTA-1) 323 324#define UCASE_GET_DELTA(props) ((int16_t)(props)>>UCASE_DELTA_SHIFT) 325 326/* case-ignorable uses one of the delta bits, see gencase/store.c */ 327#define UCASE_CASE_IGNORABLE 0x40 328 329/* exception: bits 15..4 are an unsigned 12-bit index into the exceptions array */ 330#define UCASE_EXC_SHIFT 4 331#define UCASE_EXC_MASK 0xfff0 332#define UCASE_MAX_EXCEPTIONS 0x1000 333 334/* definitions for 16-bit main exceptions word ------------------------------ */ 335 336/* first 8 bits indicate values in optional slots */ 337enum { 338 UCASE_EXC_LOWER, 339 UCASE_EXC_FOLD, 340 UCASE_EXC_UPPER, 341 UCASE_EXC_TITLE, 342 UCASE_EXC_4, /* reserved */ 343 UCASE_EXC_5, /* reserved */ 344 UCASE_EXC_CLOSURE, 345 UCASE_EXC_FULL_MAPPINGS, 346 UCASE_EXC_ALL_SLOTS /* one past the last slot */ 347}; 348 349/* each slot is 2 uint16_t instead of 1 */ 350#define UCASE_EXC_DOUBLE_SLOTS 0x100 351 352/* reserved: exception bits 11..9 */ 353 354/* UCASE_EXC_DOT_MASK=UCASE_DOT_MASK<<UCASE_EXC_DOT_SHIFT */ 355#define UCASE_EXC_DOT_SHIFT 8 356 357/* normally stored in the main word, but pushed out for larger exception indexes */ 358#define UCASE_EXC_DOT_MASK 0x3000 359enum { 360 UCASE_EXC_NO_DOT=0, 361 UCASE_EXC_SOFT_DOTTED=0x1000, 362 UCASE_EXC_ABOVE=0x2000, /* "above" accents with cc=230 */ 363 UCASE_EXC_OTHER_ACCENT=0x3000 /* other character (0<cc!=230) */ 364}; 365 366/* complex/conditional mappings */ 367#define UCASE_EXC_CONDITIONAL_SPECIAL 0x4000 368#define UCASE_EXC_CONDITIONAL_FOLD 0x8000 369 370/* definitions for lengths word for full case mappings */ 371#define UCASE_FULL_LOWER 0xf 372#define UCASE_FULL_FOLDING 0xf0 373#define UCASE_FULL_UPPER 0xf00 374#define UCASE_FULL_TITLE 0xf000 375 376/* maximum lengths */ 377#define UCASE_FULL_MAPPINGS_MAX_LENGTH (4*0xf) 378#define UCASE_CLOSURE_MAX_LENGTH 0xf 379 380/* constants for reverse case folding ("unfold") data */ 381enum { 382 UCASE_UNFOLD_ROWS, 383 UCASE_UNFOLD_ROW_WIDTH, 384 UCASE_UNFOLD_STRING_WIDTH 385}; 386 387U_CDECL_END 388 389#endif 390