Home | History | Annotate | Download | only in common
      1 /*
      2 *******************************************************************************
      3 *
      4 *   Copyright (C) 2004-2010, 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 
     27 U_CDECL_BEGIN
     28 
     29 /* library API -------------------------------------------------------------- */
     30 
     31 struct UCaseProps;
     32 typedef struct UCaseProps UCaseProps;
     33 
     34 U_CAPI const UCaseProps * U_EXPORT2
     35 ucase_getSingleton(void);
     36 
     37 U_CAPI int32_t U_EXPORT2
     38 ucase_swap(const UDataSwapper *ds,
     39            const void *inData, int32_t length, void *outData,
     40            UErrorCode *pErrorCode);
     41 
     42 U_CFUNC void U_EXPORT2
     43 ucase_addPropertyStarts(const UCaseProps *csp, const USetAdder *sa, UErrorCode *pErrorCode);
     44 
     45 /**
     46  * Requires non-NULL locale ID but otherwise does the equivalent of
     47  * checking for language codes as if uloc_getLanguage() were called:
     48  * Accepts both 2- and 3-letter codes and accepts case variants.
     49  */
     50 U_CFUNC int32_t
     51 ucase_getCaseLocale(const char *locale, int32_t *locCache);
     52 
     53 /* Casing locale types for ucase_getCaseLocale */
     54 enum {
     55     UCASE_LOC_UNKNOWN,
     56     UCASE_LOC_ROOT,
     57     UCASE_LOC_TURKISH,
     58     UCASE_LOC_LITHUANIAN,
     59     UCASE_LOC_DUTCH
     60 };
     61 
     62 /**
     63  * Bit mask for getting just the options from a string compare options word
     64  * that are relevant for case-insensitive string comparison.
     65  * See uchar.h. Also include _STRNCMP_STYLE and U_COMPARE_CODE_POINT_ORDER.
     66  * @internal
     67  */
     68 #define _STRCASECMP_OPTIONS_MASK 0xffff
     69 
     70 /**
     71  * Bit mask for getting just the options from a string compare options word
     72  * that are relevant for case folding (of a single string or code point).
     73  * See uchar.h.
     74  * @internal
     75  */
     76 #define _FOLD_CASE_OPTIONS_MASK 0xff
     77 
     78 /* single-code point functions */
     79 
     80 U_CAPI UChar32 U_EXPORT2
     81 ucase_tolower(const UCaseProps *csp, UChar32 c);
     82 
     83 U_CAPI UChar32 U_EXPORT2
     84 ucase_toupper(const UCaseProps *csp, UChar32 c);
     85 
     86 U_CAPI UChar32 U_EXPORT2
     87 ucase_totitle(const UCaseProps *csp, UChar32 c);
     88 
     89 U_CAPI UChar32 U_EXPORT2
     90 ucase_fold(const UCaseProps *csp, UChar32 c, uint32_t options);
     91 
     92 /**
     93  * Adds all simple case mappings and the full case folding for c to sa,
     94  * and also adds special case closure mappings.
     95  * c itself is not added.
     96  * For example, the mappings
     97  * - for s include long s
     98  * - for sharp s include ss
     99  * - for k include the Kelvin sign
    100  */
    101 U_CFUNC void U_EXPORT2
    102 ucase_addCaseClosure(const UCaseProps *csp, UChar32 c, const USetAdder *sa);
    103 
    104 /**
    105  * Maps the string to single code points and adds the associated case closure
    106  * mappings.
    107  * The string is mapped to code points if it is their full case folding string.
    108  * In other words, this performs a reverse full case folding and then
    109  * adds the case closure items of the resulting code points.
    110  * If the string is found and its closure applied, then
    111  * the string itself is added as well as part of its code points' closure.
    112  * It must be length>=0.
    113  *
    114  * @return TRUE if the string was found
    115  */
    116 U_CFUNC UBool U_EXPORT2
    117 ucase_addStringCaseClosure(const UCaseProps *csp, const UChar *s, int32_t length, const USetAdder *sa);
    118 
    119 /** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */
    120 U_CAPI int32_t U_EXPORT2
    121 ucase_getType(const UCaseProps *csp, UChar32 c);
    122 
    123 /** @return same as ucase_getType(), or <0 if c is case-ignorable */
    124 U_CAPI int32_t U_EXPORT2
    125 ucase_getTypeOrIgnorable(const UCaseProps *csp, UChar32 c);
    126 
    127 U_CAPI UBool U_EXPORT2
    128 ucase_isSoftDotted(const UCaseProps *csp, UChar32 c);
    129 
    130 U_CAPI UBool U_EXPORT2
    131 ucase_isCaseSensitive(const UCaseProps *csp, UChar32 c);
    132 
    133 /* string case mapping functions */
    134 
    135 /**
    136  * Iterator function for string case mappings, which need to look at the
    137  * context (surrounding text) of a given character for conditional mappings.
    138  *
    139  * The iterator only needs to go backward or forward away from the
    140  * character in question. It does not use any indexes on this interface.
    141  * It does not support random access or an arbitrary change of
    142  * iteration direction.
    143  *
    144  * The code point being case-mapped itself is never returned by
    145  * this iterator.
    146  *
    147  * @param context A pointer to the iterator's working data.
    148  * @param dir If <0 then start iterating backward from the character;
    149  *            if >0 then start iterating forward from the character;
    150  *            if 0 then continue iterating in the current direction.
    151  * @return Next code point, or <0 when the iteration is done.
    152  */
    153 typedef UChar32 U_CALLCONV
    154 UCaseContextIterator(void *context, int8_t dir);
    155 
    156 /**
    157  * Sample struct which may be used by some implementations of
    158  * UCaseContextIterator.
    159  */
    160 struct UCaseContext {
    161     void *p;
    162     int32_t start, index, limit;
    163     int32_t cpStart, cpLimit;
    164     int8_t dir;
    165     int8_t b1, b2, b3;
    166 };
    167 typedef struct UCaseContext UCaseContext;
    168 
    169 enum {
    170     /**
    171      * For string case mappings, a single character (a code point) is mapped
    172      * either to itself (in which case in-place mapping functions do nothing),
    173      * or to another single code point, or to a string.
    174      * Aside from the string contents, these are indicated with a single int32_t
    175      * value as follows:
    176      *
    177      * Mapping to self: Negative values (~self instead of -self to support U+0000)
    178      *
    179      * Mapping to another code point: Positive values >UCASE_MAX_STRING_LENGTH
    180      *
    181      * Mapping to a string: The string length (0..UCASE_MAX_STRING_LENGTH) is
    182      * returned. Note that the string result may indeed have zero length.
    183      */
    184     UCASE_MAX_STRING_LENGTH=0x1f
    185 };
    186 
    187 /**
    188  * Get the full lowercase mapping for c.
    189  *
    190  * @param csp Case mapping properties.
    191  * @param c Character to be mapped.
    192  * @param iter Character iterator, used for context-sensitive mappings.
    193  *             See UCaseContextIterator for details.
    194  *             If iter==NULL then a context-independent result is returned.
    195  * @param context Pointer to be passed into iter.
    196  * @param pString If the mapping result is a string, then the pointer is
    197  *                written to *pString.
    198  * @param locale Locale ID for locale-dependent mappings.
    199  * @param locCache Initialize to 0; may be used to cache the result of parsing
    200  *                 the locale ID for subsequent calls.
    201  *                 Can be NULL.
    202  * @return Output code point or string length, see UCASE_MAX_STRING_LENGTH.
    203  *
    204  * @see UCaseContextIterator
    205  * @see UCASE_MAX_STRING_LENGTH
    206  * @internal
    207  */
    208 U_CAPI int32_t U_EXPORT2
    209 ucase_toFullLower(const UCaseProps *csp, UChar32 c,
    210                   UCaseContextIterator *iter, void *context,
    211                   const UChar **pString,
    212                   const char *locale, int32_t *locCache);
    213 
    214 U_CAPI int32_t U_EXPORT2
    215 ucase_toFullUpper(const UCaseProps *csp, UChar32 c,
    216                   UCaseContextIterator *iter, void *context,
    217                   const UChar **pString,
    218                   const char *locale, int32_t *locCache);
    219 
    220 U_CAPI int32_t U_EXPORT2
    221 ucase_toFullTitle(const UCaseProps *csp, UChar32 c,
    222                   UCaseContextIterator *iter, void *context,
    223                   const UChar **pString,
    224                   const char *locale, int32_t *locCache);
    225 
    226 U_CAPI int32_t U_EXPORT2
    227 ucase_toFullFolding(const UCaseProps *csp, UChar32 c,
    228                     const UChar **pString,
    229                     uint32_t options);
    230 
    231 U_CFUNC int32_t U_EXPORT2
    232 ucase_hasBinaryProperty(UChar32 c, UProperty which);
    233 
    234 
    235 U_CDECL_BEGIN
    236 
    237 /**
    238  * @internal
    239  */
    240 typedef int32_t U_CALLCONV
    241 UCaseMapFull(const UCaseProps *csp, UChar32 c,
    242              UCaseContextIterator *iter, void *context,
    243              const UChar **pString,
    244              const char *locale, int32_t *locCache);
    245 
    246 U_CDECL_END
    247 
    248 /* file definitions --------------------------------------------------------- */
    249 
    250 #define UCASE_DATA_NAME "ucase"
    251 #define UCASE_DATA_TYPE "icu"
    252 
    253 /* format "cAsE" */
    254 #define UCASE_FMT_0 0x63
    255 #define UCASE_FMT_1 0x41
    256 #define UCASE_FMT_2 0x53
    257 #define UCASE_FMT_3 0x45
    258 
    259 /* indexes into indexes[] */
    260 enum {
    261     UCASE_IX_INDEX_TOP,
    262     UCASE_IX_LENGTH,
    263     UCASE_IX_TRIE_SIZE,
    264     UCASE_IX_EXC_LENGTH,
    265     UCASE_IX_UNFOLD_LENGTH,
    266 
    267     UCASE_IX_MAX_FULL_LENGTH=15,
    268     UCASE_IX_TOP=16
    269 };
    270 
    271 /* definitions for 16-bit case properties word ------------------------------ */
    272 
    273 /* 2-bit constants for types of cased characters */
    274 #define UCASE_TYPE_MASK     3
    275 enum {
    276     UCASE_NONE,
    277     UCASE_LOWER,
    278     UCASE_UPPER,
    279     UCASE_TITLE
    280 };
    281 
    282 #define UCASE_GET_TYPE(props) ((props)&UCASE_TYPE_MASK)
    283 
    284 #define UCASE_SENSITIVE     4
    285 #define UCASE_EXCEPTION     8
    286 
    287 #define UCASE_DOT_MASK      0x30
    288 enum {
    289     UCASE_NO_DOT=0,         /* normal characters with cc=0 */
    290     UCASE_SOFT_DOTTED=0x10, /* soft-dotted characters with cc=0 */
    291     UCASE_ABOVE=0x20,       /* "above" accents with cc=230 */
    292     UCASE_OTHER_ACCENT=0x30 /* other accent character (0<cc!=230) */
    293 };
    294 
    295 /* no exception: bits 15..6 are a 10-bit signed case mapping delta */
    296 #define UCASE_DELTA_SHIFT   6
    297 #define UCASE_DELTA_MASK    0xffc0
    298 #define UCASE_MAX_DELTA     0x1ff
    299 #define UCASE_MIN_DELTA     (-UCASE_MAX_DELTA-1)
    300 
    301 #define UCASE_GET_DELTA(props) ((int16_t)(props)>>UCASE_DELTA_SHIFT)
    302 
    303 /* case-ignorable uses one of the delta bits, see gencase/store.c */
    304 #define UCASE_CASE_IGNORABLE 0x40
    305 
    306 /* exception: bits 15..4 are an unsigned 12-bit index into the exceptions array */
    307 #define UCASE_EXC_SHIFT     4
    308 #define UCASE_EXC_MASK      0xfff0
    309 #define UCASE_MAX_EXCEPTIONS 0x1000
    310 
    311 /* definitions for 16-bit main exceptions word ------------------------------ */
    312 
    313 /* first 8 bits indicate values in optional slots */
    314 enum {
    315     UCASE_EXC_LOWER,
    316     UCASE_EXC_FOLD,
    317     UCASE_EXC_UPPER,
    318     UCASE_EXC_TITLE,
    319     UCASE_EXC_4,            /* reserved */
    320     UCASE_EXC_5,            /* reserved */
    321     UCASE_EXC_CLOSURE,
    322     UCASE_EXC_FULL_MAPPINGS,
    323     UCASE_EXC_ALL_SLOTS     /* one past the last slot */
    324 };
    325 
    326 /* each slot is 2 uint16_t instead of 1 */
    327 #define UCASE_EXC_DOUBLE_SLOTS      0x100
    328 
    329 /* reserved: exception bits 10..9 */
    330 
    331 #define UCASE_EXC_CASE_IGNORABLE        0x800
    332 
    333 /* UCASE_EXC_DOT_MASK=UCASE_DOT_MASK<<UCASE_EXC_DOT_SHIFT */
    334 #define UCASE_EXC_DOT_SHIFT     8
    335 
    336 /* normally stored in the main word, but pushed out for larger exception indexes */
    337 #define UCASE_EXC_DOT_MASK      0x3000
    338 enum {
    339     UCASE_EXC_NO_DOT=0,
    340     UCASE_EXC_SOFT_DOTTED=0x1000,
    341     UCASE_EXC_ABOVE=0x2000,         /* "above" accents with cc=230 */
    342     UCASE_EXC_OTHER_ACCENT=0x3000   /* other character (0<cc!=230) */
    343 };
    344 
    345 /* complex/conditional mappings */
    346 #define UCASE_EXC_CONDITIONAL_SPECIAL   0x4000
    347 #define UCASE_EXC_CONDITIONAL_FOLD      0x8000
    348 
    349 /* definitions for lengths word for full case mappings */
    350 #define UCASE_FULL_LOWER    0xf
    351 #define UCASE_FULL_FOLDING  0xf0
    352 #define UCASE_FULL_UPPER    0xf00
    353 #define UCASE_FULL_TITLE    0xf000
    354 
    355 /* maximum lengths */
    356 #define UCASE_FULL_MAPPINGS_MAX_LENGTH (4*0xf)
    357 #define UCASE_CLOSURE_MAX_LENGTH 0xf
    358 
    359 /* constants for reverse case folding ("unfold") data */
    360 enum {
    361     UCASE_UNFOLD_ROWS,
    362     UCASE_UNFOLD_ROW_WIDTH,
    363     UCASE_UNFOLD_STRING_WIDTH
    364 };
    365 
    366 U_CDECL_END
    367 
    368 #endif
    369