1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/os/textandloc/fontservices/textshaperplugin/IcuSource/common/ucase.h Fri Jun 15 03:10:57 2012 +0200
1.3 @@ -0,0 +1,359 @@
1.4 +/*
1.5 +*******************************************************************************
1.6 +*
1.7 +* Copyright (C) 2004-2005, International Business Machines
1.8 +* Corporation and others. All Rights Reserved.
1.9 +*
1.10 +*******************************************************************************
1.11 +* file name: ucase.h
1.12 +* encoding: US-ASCII
1.13 +* tab size: 8 (not used)
1.14 +* indentation:4
1.15 +*
1.16 +* created on: 2004aug30
1.17 +* created by: Markus W. Scherer
1.18 +*
1.19 +* Low-level Unicode character/string case mapping code.
1.20 +*/
1.21 +
1.22 +#ifndef __UCASE_H__
1.23 +#define __UCASE_H__
1.24 +
1.25 +#include "unicode/utypes.h"
1.26 +#include "unicode/uset.h"
1.27 +#include "uset_imp.h"
1.28 +#include "udataswp.h"
1.29 +
1.30 +U_CDECL_BEGIN
1.31 +
1.32 +/* library API -------------------------------------------------------------- */
1.33 +
1.34 +struct UCaseProps;
1.35 +typedef struct UCaseProps UCaseProps;
1.36 +
1.37 +U_CAPI UCaseProps * U_EXPORT2
1.38 +ucase_open(UErrorCode *pErrorCode);
1.39 +
1.40 +U_CAPI UCaseProps * U_EXPORT2
1.41 +ucase_openBinary(const uint8_t *bin, int32_t length, UErrorCode *pErrorCode);
1.42 +
1.43 +U_CAPI void U_EXPORT2
1.44 +ucase_close(UCaseProps *csp);
1.45 +
1.46 +
1.47 +U_CAPI const UCaseProps * U_EXPORT2
1.48 +ucase_getSingleton(UErrorCode *pErrorCode);
1.49 +
1.50 +/**
1.51 + * Get a singleton dummy object, one that works with no real data.
1.52 + * This can be used when the real data is not available.
1.53 + * Using the dummy can reduce checks for available data after an initial failure.
1.54 + */
1.55 +U_CAPI const UCaseProps * U_EXPORT2
1.56 +ucase_getDummy(UErrorCode *pErrorCode);
1.57 +
1.58 +
1.59 +U_CAPI int32_t U_EXPORT2
1.60 +ucase_swap(const UDataSwapper *ds,
1.61 + const void *inData, int32_t length, void *outData,
1.62 + UErrorCode *pErrorCode);
1.63 +
1.64 +U_CAPI void U_EXPORT2
1.65 +ucase_addPropertyStarts(const UCaseProps *csp, const USetAdder *sa, UErrorCode *pErrorCode);
1.66 +
1.67 +/**
1.68 + * Requires non-NULL locale ID but otherwise does the equivalent of
1.69 + * checking for language codes as if uloc_getLanguage() were called:
1.70 + * Accepts both 2- and 3-letter codes and accepts case variants.
1.71 + */
1.72 +U_CFUNC int32_t
1.73 +ucase_getCaseLocale(const char *locale, int32_t *locCache);
1.74 +
1.75 +/**
1.76 + * Bit mask for getting just the options from a string compare options word
1.77 + * that are relevant for case-insensitive string comparison.
1.78 + * See uchar.h. Also include _STRNCMP_STYLE and U_COMPARE_CODE_POINT_ORDER.
1.79 + * @internal
1.80 + */
1.81 +#define _STRCASECMP_OPTIONS_MASK 0xffff
1.82 +
1.83 +/**
1.84 + * Bit mask for getting just the options from a string compare options word
1.85 + * that are relevant for case folding (of a single string or code point).
1.86 + * See uchar.h.
1.87 + * @internal
1.88 + */
1.89 +#define _FOLD_CASE_OPTIONS_MASK 0xff
1.90 +
1.91 +/* single-code point functions */
1.92 +
1.93 +U_CAPI UChar32 U_EXPORT2
1.94 +ucase_tolower(const UCaseProps *csp, UChar32 c);
1.95 +
1.96 +U_CAPI UChar32 U_EXPORT2
1.97 +ucase_toupper(const UCaseProps *csp, UChar32 c);
1.98 +
1.99 +U_CAPI UChar32 U_EXPORT2
1.100 +ucase_totitle(const UCaseProps *csp, UChar32 c);
1.101 +
1.102 +U_CAPI UChar32 U_EXPORT2
1.103 +ucase_fold(const UCaseProps *csp, UChar32 c, uint32_t options);
1.104 +
1.105 +/**
1.106 + * Adds all simple case mappings and the full case folding for c to sa,
1.107 + * and also adds special case closure mappings.
1.108 + * c itself is not added.
1.109 + * For example, the mappings
1.110 + * - for s include long s
1.111 + * - for sharp s include ss
1.112 + * - for k include the Kelvin sign
1.113 + */
1.114 +U_CAPI void U_EXPORT2
1.115 +ucase_addCaseClosure(const UCaseProps *csp, UChar32 c, const USetAdder *sa);
1.116 +
1.117 +/**
1.118 + * Maps the string to single code points and adds the associated case closure
1.119 + * mappings.
1.120 + * The string is mapped to code points if it is their full case folding string.
1.121 + * In other words, this performs a reverse full case folding and then
1.122 + * adds the case closure items of the resulting code points.
1.123 + * If the string is found and its closure applied, then
1.124 + * the string itself is added as well as part of its code points' closure.
1.125 + * It must be length>=0.
1.126 + *
1.127 + * @return TRUE if the string was found
1.128 + */
1.129 +U_CAPI UBool U_EXPORT2
1.130 +ucase_addStringCaseClosure(const UCaseProps *csp, const UChar *s, int32_t length, const USetAdder *sa);
1.131 +
1.132 +/** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */
1.133 +U_CAPI int32_t U_EXPORT2
1.134 +ucase_getType(const UCaseProps *csp, UChar32 c);
1.135 +
1.136 +/** @return same as ucase_getType(), or <0 if c is case-ignorable */
1.137 +U_CAPI int32_t U_EXPORT2
1.138 +ucase_getTypeOrIgnorable(const UCaseProps *csp, UChar32 c);
1.139 +
1.140 +U_CAPI UBool U_EXPORT2
1.141 +ucase_isSoftDotted(const UCaseProps *csp, UChar32 c);
1.142 +
1.143 +U_CAPI UBool U_EXPORT2
1.144 +ucase_isCaseSensitive(const UCaseProps *csp, UChar32 c);
1.145 +
1.146 +/* string case mapping functions */
1.147 +
1.148 +/**
1.149 + * Iterator function for string case mappings, which need to look at the
1.150 + * context (surrounding text) of a given character for conditional mappings.
1.151 + *
1.152 + * The iterator only needs to go backward or forward away from the
1.153 + * character in question. It does not use any indexes on this interface.
1.154 + * It does not support random access or an arbitrary change of
1.155 + * iteration direction.
1.156 + *
1.157 + * The code point being case-mapped itself is never returned by
1.158 + * this iterator.
1.159 + *
1.160 + * @param context A pointer to the iterator's working data.
1.161 + * @param dir If <0 then start iterating backward from the character;
1.162 + * if >0 then start iterating forward from the character;
1.163 + * if 0 then continue iterating in the current direction.
1.164 + * @return Next code point, or <0 when the iteration is done.
1.165 + */
1.166 +typedef UChar32 U_CALLCONV
1.167 +UCaseContextIterator(void *context, int8_t dir);
1.168 +
1.169 +/**
1.170 + * Sample struct which may be used by some implementations of
1.171 + * UCaseContextIterator.
1.172 + */
1.173 +struct UCaseContext {
1.174 + void *p;
1.175 + int32_t start, index, limit;
1.176 + int32_t cpStart, cpLimit;
1.177 + int8_t dir;
1.178 + int8_t b1, b2, b3;
1.179 +};
1.180 +typedef struct UCaseContext UCaseContext;
1.181 +
1.182 +enum {
1.183 + /**
1.184 + * For string case mappings, a single character (a code point) is mapped
1.185 + * either to itself (in which case in-place mapping functions do nothing),
1.186 + * or to another single code point, or to a string.
1.187 + * Aside from the string contents, these are indicated with a single int32_t
1.188 + * value as follows:
1.189 + *
1.190 + * Mapping to self: Negative values (~self instead of -self to support U+0000)
1.191 + *
1.192 + * Mapping to another code point: Positive values >UCASE_MAX_STRING_LENGTH
1.193 + *
1.194 + * Mapping to a string: The string length (0..UCASE_MAX_STRING_LENGTH) is
1.195 + * returned. Note that the string result may indeed have zero length.
1.196 + */
1.197 + UCASE_MAX_STRING_LENGTH=0x1f
1.198 +};
1.199 +
1.200 +/**
1.201 + * Get the full lowercase mapping for c.
1.202 + *
1.203 + * @param csp Case mapping properties.
1.204 + * @param c Character to be mapped.
1.205 + * @param iter Character iterator, used for context-sensitive mappings.
1.206 + * See UCaseContextIterator for details.
1.207 + * If iter==NULL then a context-independent result is returned.
1.208 + * @param context Pointer to be passed into iter.
1.209 + * @param pString If the mapping result is a string, then the pointer is
1.210 + * written to *pString.
1.211 + * @param locale Locale ID for locale-dependent mappings.
1.212 + * @param locCache Initialize to 0; may be used to cache the result of parsing
1.213 + * the locale ID for subsequent calls.
1.214 + * Can be NULL.
1.215 + * @return Output code point or string length, see UCASE_MAX_STRING_LENGTH.
1.216 + *
1.217 + * @see UCaseContextIterator
1.218 + * @see UCASE_MAX_STRING_LENGTH
1.219 + * @internal
1.220 + */
1.221 +U_CAPI int32_t U_EXPORT2
1.222 +ucase_toFullLower(const UCaseProps *csp, UChar32 c,
1.223 + UCaseContextIterator *iter, void *context,
1.224 + const UChar **pString,
1.225 + const char *locale, int32_t *locCache);
1.226 +
1.227 +U_CAPI int32_t U_EXPORT2
1.228 +ucase_toFullUpper(const UCaseProps *csp, UChar32 c,
1.229 + UCaseContextIterator *iter, void *context,
1.230 + const UChar **pString,
1.231 + const char *locale, int32_t *locCache);
1.232 +
1.233 +U_CAPI int32_t U_EXPORT2
1.234 +ucase_toFullTitle(const UCaseProps *csp, UChar32 c,
1.235 + UCaseContextIterator *iter, void *context,
1.236 + const UChar **pString,
1.237 + const char *locale, int32_t *locCache);
1.238 +
1.239 +U_CAPI int32_t U_EXPORT2
1.240 +ucase_toFullFolding(const UCaseProps *csp, UChar32 c,
1.241 + const UChar **pString,
1.242 + uint32_t options);
1.243 +
1.244 +/* file definitions --------------------------------------------------------- */
1.245 +
1.246 +#define UCASE_DATA_NAME "ucase"
1.247 +#define UCASE_DATA_TYPE "icu"
1.248 +
1.249 +/* format "cAsE" */
1.250 +#define UCASE_FMT_0 0x63
1.251 +#define UCASE_FMT_1 0x41
1.252 +#define UCASE_FMT_2 0x53
1.253 +#define UCASE_FMT_3 0x45
1.254 +
1.255 +/* indexes into indexes[] */
1.256 +enum {
1.257 + UCASE_IX_INDEX_TOP,
1.258 + UCASE_IX_LENGTH,
1.259 + UCASE_IX_TRIE_SIZE,
1.260 + UCASE_IX_EXC_LENGTH,
1.261 + UCASE_IX_UNFOLD_LENGTH,
1.262 +
1.263 + UCASE_IX_MAX_FULL_LENGTH=15,
1.264 + UCASE_IX_TOP=16
1.265 +};
1.266 +
1.267 +/* definitions for 16-bit case properties word ------------------------------ */
1.268 +
1.269 +/* 2-bit constants for types of cased characters */
1.270 +#define UCASE_TYPE_MASK 3
1.271 +enum {
1.272 + UCASE_NONE,
1.273 + UCASE_LOWER,
1.274 + UCASE_UPPER,
1.275 + UCASE_TITLE
1.276 +};
1.277 +
1.278 +#define UCASE_GET_TYPE(props) ((props)&UCASE_TYPE_MASK)
1.279 +
1.280 +#define UCASE_SENSITIVE 4
1.281 +#define UCASE_EXCEPTION 8
1.282 +
1.283 +#define UCASE_DOT_MASK 0x30
1.284 +enum {
1.285 + UCASE_NO_DOT=0, /* normal characters with cc=0 */
1.286 + UCASE_SOFT_DOTTED=0x10, /* soft-dotted characters with cc=0 */
1.287 + UCASE_ABOVE=0x20, /* "above" accents with cc=230 */
1.288 + UCASE_OTHER_ACCENT=0x30 /* other accent character (0<cc!=230) */
1.289 +};
1.290 +
1.291 +/* no exception: bits 15..6 are a 10-bit signed case mapping delta */
1.292 +#define UCASE_DELTA_SHIFT 6
1.293 +#define UCASE_DELTA_MASK 0xffc0
1.294 +#define UCASE_MAX_DELTA 0x1ff
1.295 +#define UCASE_MIN_DELTA (-UCASE_MAX_DELTA-1)
1.296 +
1.297 +#define UCASE_GET_DELTA(props) ((int16_t)(props)>>UCASE_DELTA_SHIFT)
1.298 +
1.299 +/* case-ignorable uses one of the delta bits, see gencase/store.c */
1.300 +#define UCASE_CASE_IGNORABLE 0x40
1.301 +
1.302 +/* exception: bits 15..4 are an unsigned 12-bit index into the exceptions array */
1.303 +#define UCASE_EXC_SHIFT 4
1.304 +#define UCASE_EXC_MASK 0xfff0
1.305 +#define UCASE_MAX_EXCEPTIONS 0x1000
1.306 +
1.307 +/* definitions for 16-bit main exceptions word ------------------------------ */
1.308 +
1.309 +/* first 8 bits indicate values in optional slots */
1.310 +enum {
1.311 + UCASE_EXC_LOWER,
1.312 + UCASE_EXC_FOLD,
1.313 + UCASE_EXC_UPPER,
1.314 + UCASE_EXC_TITLE,
1.315 + UCASE_EXC_4, /* reserved */
1.316 + UCASE_EXC_5, /* reserved */
1.317 + UCASE_EXC_CLOSURE,
1.318 + UCASE_EXC_FULL_MAPPINGS,
1.319 + UCASE_EXC_ALL_SLOTS /* one past the last slot */
1.320 +};
1.321 +
1.322 +/* each slot is 2 uint16_t instead of 1 */
1.323 +#define UCASE_EXC_DOUBLE_SLOTS 0x100
1.324 +
1.325 +/* reserved: exception bits 11..9 */
1.326 +
1.327 +/* UCASE_EXC_DOT_MASK=UCASE_DOT_MASK<<UCASE_EXC_DOT_SHIFT */
1.328 +#define UCASE_EXC_DOT_SHIFT 8
1.329 +
1.330 +/* normally stored in the main word, but pushed out for larger exception indexes */
1.331 +#define UCASE_EXC_DOT_MASK 0x3000
1.332 +enum {
1.333 + UCASE_EXC_NO_DOT=0,
1.334 + UCASE_EXC_SOFT_DOTTED=0x1000,
1.335 + UCASE_EXC_ABOVE=0x2000, /* "above" accents with cc=230 */
1.336 + UCASE_EXC_OTHER_ACCENT=0x3000 /* other character (0<cc!=230) */
1.337 +};
1.338 +
1.339 +/* complex/conditional mappings */
1.340 +#define UCASE_EXC_CONDITIONAL_SPECIAL 0x4000
1.341 +#define UCASE_EXC_CONDITIONAL_FOLD 0x8000
1.342 +
1.343 +/* definitions for lengths word for full case mappings */
1.344 +#define UCASE_FULL_LOWER 0xf
1.345 +#define UCASE_FULL_FOLDING 0xf0
1.346 +#define UCASE_FULL_UPPER 0xf00
1.347 +#define UCASE_FULL_TITLE 0xf000
1.348 +
1.349 +/* maximum lengths */
1.350 +#define UCASE_FULL_MAPPINGS_MAX_LENGTH (4*0xf)
1.351 +#define UCASE_CLOSURE_MAX_LENGTH 0xf
1.352 +
1.353 +/* constants for reverse case folding ("unfold") data */
1.354 +enum {
1.355 + UCASE_UNFOLD_ROWS,
1.356 + UCASE_UNFOLD_ROW_WIDTH,
1.357 + UCASE_UNFOLD_STRING_WIDTH
1.358 +};
1.359 +
1.360 +U_CDECL_END
1.361 +
1.362 +#endif