Home | History | Annotate | Download | only in common
      1 //  2016 and later: Unicode, Inc. and others.
      2 // License & terms of use: http://www.unicode.org/copyright.html
      3 /*
      4 *******************************************************************************
      5 *
      6 *   Copyright (C) 2004-2012, International Business Machines
      7 *   Corporation and others.  All Rights Reserved.
      8 *
      9 *******************************************************************************
     10 *   file name:  ucase.h
     11 *   encoding:   UTF-8
     12 *   tab size:   8 (not used)
     13 *   indentation:4
     14 *
     15 *   created on: 2004aug30
     16 *   created by: Markus W. Scherer
     17 *
     18 *   Low-level Unicode character/string case mapping code.
     19 */
     20 
     21 #ifndef __UCASE_H__
     22 #define __UCASE_H__
     23 
     24 #include "unicode/utypes.h"
     25 #include "unicode/uset.h"
     26 #include "putilimp.h"
     27 #include "uset_imp.h"
     28 #include "udataswp.h"
     29 
     30 #ifdef __cplusplus
     31 U_NAMESPACE_BEGIN
     32 
     33 class UnicodeString;
     34 
     35 U_NAMESPACE_END
     36 #endif
     37 
     38 /* library API -------------------------------------------------------------- */
     39 
     40 U_CFUNC void U_EXPORT2
     41 ucase_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode);
     42 
     43 /**
     44  * Requires non-NULL locale ID but otherwise does the equivalent of
     45  * checking for language codes as if uloc_getLanguage() were called:
     46  * Accepts both 2- and 3-letter codes and accepts case variants.
     47  */
     48 U_CFUNC int32_t
     49 ucase_getCaseLocale(const char *locale);
     50 
     51 /* Casing locale types for ucase_getCaseLocale */
     52 enum {
     53     UCASE_LOC_UNKNOWN,
     54     UCASE_LOC_ROOT,
     55     UCASE_LOC_TURKISH,
     56     UCASE_LOC_LITHUANIAN,
     57     UCASE_LOC_GREEK,
     58     UCASE_LOC_DUTCH
     59 };
     60 
     61 /**
     62  * Bit mask for getting just the options from a string compare options word
     63  * that are relevant for case-insensitive string comparison.
     64  * See stringoptions.h. Also include _STRNCMP_STYLE and U_COMPARE_CODE_POINT_ORDER.
     65  * @internal
     66  */
     67 #define _STRCASECMP_OPTIONS_MASK 0xffff
     68 
     69 /**
     70  * Bit mask for getting just the options from a string compare options word
     71  * that are relevant for case folding (of a single string or code point).
     72  *
     73  * Currently only bit 0 for U_FOLD_CASE_EXCLUDE_SPECIAL_I.
     74  * It is conceivable that at some point we might use one more bit for using uppercase sharp s.
     75  * It is conceivable that at some point we might want the option to use only simple case foldings
     76  * when operating on strings.
     77  *
     78  * See stringoptions.h.
     79  * @internal
     80  */
     81 #define _FOLD_CASE_OPTIONS_MASK 7
     82 
     83 /* single-code point functions */
     84 
     85 U_CAPI UChar32 U_EXPORT2
     86 ucase_tolower(UChar32 c);
     87 
     88 U_CAPI UChar32 U_EXPORT2
     89 ucase_toupper(UChar32 c);
     90 
     91 U_CAPI UChar32 U_EXPORT2
     92 ucase_totitle(UChar32 c);
     93 
     94 U_CAPI UChar32 U_EXPORT2
     95 ucase_fold(UChar32 c, uint32_t options);
     96 
     97 /**
     98  * Adds all simple case mappings and the full case folding for c to sa,
     99  * and also adds special case closure mappings.
    100  * c itself is not added.
    101  * For example, the mappings
    102  * - for s include long s
    103  * - for sharp s include ss
    104  * - for k include the Kelvin sign
    105  */
    106 U_CFUNC void U_EXPORT2
    107 ucase_addCaseClosure(UChar32 c, const USetAdder *sa);
    108 
    109 /**
    110  * Maps the string to single code points and adds the associated case closure
    111  * mappings.
    112  * The string is mapped to code points if it is their full case folding string.
    113  * In other words, this performs a reverse full case folding and then
    114  * adds the case closure items of the resulting code points.
    115  * If the string is found and its closure applied, then
    116  * the string itself is added as well as part of its code points' closure.
    117  * It must be length>=0.
    118  *
    119  * @return TRUE if the string was found
    120  */
    121 U_CFUNC UBool U_EXPORT2
    122 ucase_addStringCaseClosure(const UChar *s, int32_t length, const USetAdder *sa);
    123 
    124 #ifdef __cplusplus
    125 U_NAMESPACE_BEGIN
    126 
    127 /**
    128  * Iterator over characters with more than one code point in the full default Case_Folding.
    129  */
    130 class U_COMMON_API FullCaseFoldingIterator {
    131 public:
    132     /** Constructor. */
    133     FullCaseFoldingIterator();
    134     /**
    135      * Returns the next (cp, full) pair where "full" is cp's full default Case_Folding.
    136      * Returns a negative cp value at the end of the iteration.
    137      */
    138     UChar32 next(UnicodeString &full);
    139 private:
    140     FullCaseFoldingIterator(const FullCaseFoldingIterator &);  // no copy
    141     FullCaseFoldingIterator &operator=(const FullCaseFoldingIterator &);  // no assignment
    142 
    143     const UChar *unfold;
    144     int32_t unfoldRows;
    145     int32_t unfoldRowWidth;
    146     int32_t unfoldStringWidth;
    147     int32_t currentRow;
    148     int32_t rowCpIndex;
    149 };
    150 
    151 U_NAMESPACE_END
    152 #endif
    153 
    154 /** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */
    155 U_CAPI int32_t U_EXPORT2
    156 ucase_getType(UChar32 c);
    157 
    158 /** @return like ucase_getType() but also sets UCASE_IGNORABLE if c is case-ignorable */
    159 U_CAPI int32_t U_EXPORT2
    160 ucase_getTypeOrIgnorable(UChar32 c);
    161 
    162 U_CAPI UBool U_EXPORT2
    163 ucase_isSoftDotted(UChar32 c);
    164 
    165 U_CAPI UBool U_EXPORT2
    166 ucase_isCaseSensitive(UChar32 c);
    167 
    168 /* string case mapping functions */
    169 
    170 U_CDECL_BEGIN
    171 
    172 /**
    173  * Iterator function for string case mappings, which need to look at the
    174  * context (surrounding text) of a given character for conditional mappings.
    175  *
    176  * The iterator only needs to go backward or forward away from the
    177  * character in question. It does not use any indexes on this interface.
    178  * It does not support random access or an arbitrary change of
    179  * iteration direction.
    180  *
    181  * The code point being case-mapped itself is never returned by
    182  * this iterator.
    183  *
    184  * @param context A pointer to the iterator's working data.
    185  * @param dir If <0 then start iterating backward from the character;
    186  *            if >0 then start iterating forward from the character;
    187  *            if 0 then continue iterating in the current direction.
    188  * @return Next code point, or <0 when the iteration is done.
    189  */
    190 typedef UChar32 U_CALLCONV
    191 UCaseContextIterator(void *context, int8_t dir);
    192 
    193 /**
    194  * Sample struct which may be used by some implementations of
    195  * UCaseContextIterator.
    196  */
    197 struct UCaseContext {
    198     void *p;
    199     int32_t start, index, limit;
    200     int32_t cpStart, cpLimit;
    201     int8_t dir;
    202     int8_t b1, b2, b3;
    203 };
    204 typedef struct UCaseContext UCaseContext;
    205 
    206 U_CDECL_END
    207 
    208 #define UCASECONTEXT_INITIALIZER { NULL,  0, 0, 0,  0, 0,  0,  0, 0, 0 }
    209 
    210 enum {
    211     /**
    212      * For string case mappings, a single character (a code point) is mapped
    213      * either to itself (in which case in-place mapping functions do nothing),
    214      * or to another single code point, or to a string.
    215      * Aside from the string contents, these are indicated with a single int32_t
    216      * value as follows:
    217      *
    218      * Mapping to self: Negative values (~self instead of -self to support U+0000)
    219      *
    220      * Mapping to another code point: Positive values >UCASE_MAX_STRING_LENGTH
    221      *
    222      * Mapping to a string: The string length (0..UCASE_MAX_STRING_LENGTH) is
    223      * returned. Note that the string result may indeed have zero length.
    224      */
    225     UCASE_MAX_STRING_LENGTH=0x1f
    226 };
    227 
    228 /**
    229  * Get the full lowercase mapping for c.
    230  *
    231  * @param csp Case mapping properties.
    232  * @param c Character to be mapped.
    233  * @param iter Character iterator, used for context-sensitive mappings.
    234  *             See UCaseContextIterator for details.
    235  *             If iter==NULL then a context-independent result is returned.
    236  * @param context Pointer to be passed into iter.
    237  * @param pString If the mapping result is a string, then the pointer is
    238  *                written to *pString.
    239  * @param caseLocale Case locale value from ucase_getCaseLocale().
    240  * @return Output code point or string length, see UCASE_MAX_STRING_LENGTH.
    241  *
    242  * @see UCaseContextIterator
    243  * @see UCASE_MAX_STRING_LENGTH
    244  * @internal
    245  */
    246 U_CAPI int32_t U_EXPORT2
    247 ucase_toFullLower(UChar32 c,
    248                   UCaseContextIterator *iter, void *context,
    249                   const UChar **pString,
    250                   int32_t caseLocale);
    251 
    252 U_CAPI int32_t U_EXPORT2
    253 ucase_toFullUpper(UChar32 c,
    254                   UCaseContextIterator *iter, void *context,
    255                   const UChar **pString,
    256                   int32_t caseLocale);
    257 
    258 U_CAPI int32_t U_EXPORT2
    259 ucase_toFullTitle(UChar32 c,
    260                   UCaseContextIterator *iter, void *context,
    261                   const UChar **pString,
    262                   int32_t caseLocale);
    263 
    264 U_CAPI int32_t U_EXPORT2
    265 ucase_toFullFolding(UChar32 c,
    266                     const UChar **pString,
    267                     uint32_t options);
    268 
    269 U_CFUNC int32_t U_EXPORT2
    270 ucase_hasBinaryProperty(UChar32 c, UProperty which);
    271 
    272 
    273 U_CDECL_BEGIN
    274 
    275 /**
    276  * @internal
    277  */
    278 typedef int32_t U_CALLCONV
    279 UCaseMapFull(UChar32 c,
    280              UCaseContextIterator *iter, void *context,
    281              const UChar **pString,
    282              int32_t caseLocale);
    283 
    284 U_CDECL_END
    285 
    286 /* file definitions --------------------------------------------------------- */
    287 
    288 #define UCASE_DATA_NAME "ucase"
    289 #define UCASE_DATA_TYPE "icu"
    290 
    291 /* format "cAsE" */
    292 #define UCASE_FMT_0 0x63
    293 #define UCASE_FMT_1 0x41
    294 #define UCASE_FMT_2 0x53
    295 #define UCASE_FMT_3 0x45
    296 
    297 /* indexes into indexes[] */
    298 enum {
    299     UCASE_IX_INDEX_TOP,
    300     UCASE_IX_LENGTH,
    301     UCASE_IX_TRIE_SIZE,
    302     UCASE_IX_EXC_LENGTH,
    303     UCASE_IX_UNFOLD_LENGTH,
    304 
    305     UCASE_IX_MAX_FULL_LENGTH=15,
    306     UCASE_IX_TOP=16
    307 };
    308 
    309 /* definitions for 16-bit case properties word ------------------------------ */
    310 
    311 /* 2-bit constants for types of cased characters */
    312 #define UCASE_TYPE_MASK     3
    313 enum {
    314     UCASE_NONE,
    315     UCASE_LOWER,
    316     UCASE_UPPER,
    317     UCASE_TITLE
    318 };
    319 
    320 #define UCASE_GET_TYPE(props) ((props)&UCASE_TYPE_MASK)
    321 #define UCASE_GET_TYPE_AND_IGNORABLE(props) ((props)&7)
    322 
    323 #define UCASE_IGNORABLE         4
    324 #define UCASE_SENSITIVE         8
    325 #define UCASE_EXCEPTION         0x10
    326 
    327 #define UCASE_DOT_MASK      0x60
    328 enum {
    329     UCASE_NO_DOT=0,         /* normal characters with cc=0 */
    330     UCASE_SOFT_DOTTED=0x20, /* soft-dotted characters with cc=0 */
    331     UCASE_ABOVE=0x40,       /* "above" accents with cc=230 */
    332     UCASE_OTHER_ACCENT=0x60 /* other accent character (0<cc!=230) */
    333 };
    334 
    335 /* no exception: bits 15..7 are a 9-bit signed case mapping delta */
    336 #define UCASE_DELTA_SHIFT   7
    337 #define UCASE_DELTA_MASK    0xff80
    338 #define UCASE_MAX_DELTA     0xff
    339 #define UCASE_MIN_DELTA     (-UCASE_MAX_DELTA-1)
    340 
    341 #if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
    342 #   define UCASE_GET_DELTA(props) ((int16_t)(props)>>UCASE_DELTA_SHIFT)
    343 #else
    344 #   define UCASE_GET_DELTA(props) (int16_t)(((props)&0x8000) ? (((props)>>UCASE_DELTA_SHIFT)|0xfe00) : ((uint16_t)(props)>>UCASE_DELTA_SHIFT))
    345 #endif
    346 
    347 /* exception: bits 15..5 are an unsigned 11-bit index into the exceptions array */
    348 #define UCASE_EXC_SHIFT     5
    349 #define UCASE_EXC_MASK      0xffe0
    350 #define UCASE_MAX_EXCEPTIONS ((UCASE_EXC_MASK>>UCASE_EXC_SHIFT)+1)
    351 
    352 /* definitions for 16-bit main exceptions word ------------------------------ */
    353 
    354 /* first 8 bits indicate values in optional slots */
    355 enum {
    356     UCASE_EXC_LOWER,
    357     UCASE_EXC_FOLD,
    358     UCASE_EXC_UPPER,
    359     UCASE_EXC_TITLE,
    360     UCASE_EXC_4,            /* reserved */
    361     UCASE_EXC_5,            /* reserved */
    362     UCASE_EXC_CLOSURE,
    363     UCASE_EXC_FULL_MAPPINGS,
    364     UCASE_EXC_ALL_SLOTS     /* one past the last slot */
    365 };
    366 
    367 /* each slot is 2 uint16_t instead of 1 */
    368 #define UCASE_EXC_DOUBLE_SLOTS      0x100
    369 
    370 /* reserved: exception bits 11..9 */
    371 
    372 /* UCASE_EXC_DOT_MASK=UCASE_DOT_MASK<<UCASE_EXC_DOT_SHIFT */
    373 #define UCASE_EXC_DOT_SHIFT     7
    374 
    375 /* normally stored in the main word, but pushed out for larger exception indexes */
    376 #define UCASE_EXC_DOT_MASK      0x3000
    377 enum {
    378     UCASE_EXC_NO_DOT=0,
    379     UCASE_EXC_SOFT_DOTTED=0x1000,
    380     UCASE_EXC_ABOVE=0x2000,         /* "above" accents with cc=230 */
    381     UCASE_EXC_OTHER_ACCENT=0x3000   /* other character (0<cc!=230) */
    382 };
    383 
    384 /* complex/conditional mappings */
    385 #define UCASE_EXC_CONDITIONAL_SPECIAL   0x4000
    386 #define UCASE_EXC_CONDITIONAL_FOLD      0x8000
    387 
    388 /* definitions for lengths word for full case mappings */
    389 #define UCASE_FULL_LOWER    0xf
    390 #define UCASE_FULL_FOLDING  0xf0
    391 #define UCASE_FULL_UPPER    0xf00
    392 #define UCASE_FULL_TITLE    0xf000
    393 
    394 /* maximum lengths */
    395 #define UCASE_FULL_MAPPINGS_MAX_LENGTH (4*0xf)
    396 #define UCASE_CLOSURE_MAX_LENGTH 0xf
    397 
    398 /* constants for reverse case folding ("unfold") data */
    399 enum {
    400     UCASE_UNFOLD_ROWS,
    401     UCASE_UNFOLD_ROW_WIDTH,
    402     UCASE_UNFOLD_STRING_WIDTH
    403 };
    404 
    405 #endif
    406