ICU-64260.0.1.tar.gz

[apple/icu.git] / icuSources / common / ucase.h

diff --git a/icuSources/common/ucase.h b/icuSources/common/ucase.h
index e3e5be0517daeb3afa4e8ed0718f027d83f9b7bb..b0a453b87e8afa9936f4ef6780548ee391fb5bd0 100644 (file)
--- a/icuSources/common/ucase.h
+++ b/icuSources/common/ucase.h
@@ -1,12 +1,14 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html

 /*
 *******************************************************************************
 *
 /*
 *******************************************************************************
 *

-*   Copyright (C) 2004-2006, International Business Machines
+*   Copyright (C) 2004-2012, International Business Machines

 *   Corporation and others.  All Rights Reserved.
 *
 *******************************************************************************
 *   file name:  ucase.h
 *   Corporation and others.  All Rights Reserved.
 *
 *******************************************************************************
 *   file name:  ucase.h

-*   encoding:   US-ASCII
+*   encoding:   UTF-8

 *   tab size:   8 (not used)
 *   indentation:4
 *
 *   tab size:   8 (not used)
 *   indentation:4
 *


@@ -21,45 +23,23 @@
 
 #include "unicode/utypes.h"
 #include "unicode/uset.h"
 
 #include "unicode/utypes.h"
 #include "unicode/uset.h"

+#include "putilimp.h"

 #include "uset_imp.h"
 #include "udataswp.h"
 #include "uset_imp.h"
 #include "udataswp.h"

+#include "utrie2.h"

 
 

-U_CDECL_BEGIN
-
-/* library API -------------------------------------------------------------- */
-
-struct UCaseProps;
-typedef struct UCaseProps UCaseProps;
-
-U_CAPI UCaseProps * U_EXPORT2
-ucase_open(UErrorCode *pErrorCode);
-
-U_CAPI UCaseProps * U_EXPORT2
-ucase_openBinary(const uint8_t *bin, int32_t length, UErrorCode *pErrorCode);
-
-U_CAPI void U_EXPORT2
-ucase_close(UCaseProps *csp);
-
+#ifdef __cplusplus
+U_NAMESPACE_BEGIN

 
 

-U_CAPI const UCaseProps * U_EXPORT2
-ucase_getSingleton(UErrorCode *pErrorCode);
-
-/**
- * Get a singleton dummy object, one that works with no real data.
- * This can be used when the real data is not available.
- * Using the dummy can reduce checks for available data after an initial failure.
- */
-U_CAPI const UCaseProps * U_EXPORT2
-ucase_getDummy(UErrorCode *pErrorCode);
+class UnicodeString;

 
 

+U_NAMESPACE_END
+#endif

 
 

-U_CAPI int32_t U_EXPORT2
-ucase_swap(const UDataSwapper *ds,
-           const void *inData, int32_t length, void *outData,
-           UErrorCode *pErrorCode);
+/* library API -------------------------------------------------------------- */

 
 

-U_CAPI void U_EXPORT2
-ucase_addPropertyStarts(const UCaseProps *csp, const USetAdder *sa, UErrorCode *pErrorCode);
+U_CFUNC void U_EXPORT2
+ucase_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode);

 
 /**
  * Requires non-NULL locale ID but otherwise does the equivalent of
 
 /**
  * Requires non-NULL locale ID but otherwise does the equivalent of


@@ -67,12 +47,22 @@ ucase_addPropertyStarts(const UCaseProps *csp, const USetAdder *sa, UErrorCode *
  * Accepts both 2- and 3-letter codes and accepts case variants.
  */
 U_CFUNC int32_t
  * Accepts both 2- and 3-letter codes and accepts case variants.
  */
 U_CFUNC int32_t

-ucase_getCaseLocale(const char *locale, int32_t *locCache);
+ucase_getCaseLocale(const char *locale);
+
+/* Casing locale types for ucase_getCaseLocale */
+enum {
+    UCASE_LOC_UNKNOWN,
+    UCASE_LOC_ROOT,
+    UCASE_LOC_TURKISH,
+    UCASE_LOC_LITHUANIAN,
+    UCASE_LOC_GREEK,
+    UCASE_LOC_DUTCH
+};

 
 /**
  * Bit mask for getting just the options from a string compare options word
  * that are relevant for case-insensitive string comparison.
 
 /**
  * Bit mask for getting just the options from a string compare options word
  * that are relevant for case-insensitive string comparison.

- * See uchar.h. Also include _STRNCMP_STYLE and U_COMPARE_CODE_POINT_ORDER.
+ * See stringoptions.h. Also include _STRNCMP_STYLE and U_COMPARE_CODE_POINT_ORDER.

  * @internal
  */
 #define _STRCASECMP_OPTIONS_MASK 0xffff
  * @internal
  */
 #define _STRCASECMP_OPTIONS_MASK 0xffff


@@ -80,24 +70,30 @@ ucase_getCaseLocale(const char *locale, int32_t *locCache);
 /**
  * Bit mask for getting just the options from a string compare options word
  * that are relevant for case folding (of a single string or code point).
 /**
  * Bit mask for getting just the options from a string compare options word
  * that are relevant for case folding (of a single string or code point).

- * See uchar.h.
+ *
+ * Currently only bit 0 for U_FOLD_CASE_EXCLUDE_SPECIAL_I.
+ * It is conceivable that at some point we might use one more bit for using uppercase sharp s.
+ * It is conceivable that at some point we might want the option to use only simple case foldings
+ * when operating on strings.
+ *
+ * See stringoptions.h.

  * @internal
  */
  * @internal
  */

-#define _FOLD_CASE_OPTIONS_MASK 0xff
+#define _FOLD_CASE_OPTIONS_MASK 7

 
 /* single-code point functions */
 
 U_CAPI UChar32 U_EXPORT2
 
 /* single-code point functions */
 
 U_CAPI UChar32 U_EXPORT2

-ucase_tolower(const UCaseProps *csp, UChar32 c);
+ucase_tolower(UChar32 c);

 
 U_CAPI UChar32 U_EXPORT2
 
 U_CAPI UChar32 U_EXPORT2

-ucase_toupper(const UCaseProps *csp, UChar32 c);
+ucase_toupper(UChar32 c);

 
 U_CAPI UChar32 U_EXPORT2
 
 U_CAPI UChar32 U_EXPORT2

-ucase_totitle(const UCaseProps *csp, UChar32 c);
+ucase_totitle(UChar32 c);

 
 U_CAPI UChar32 U_EXPORT2
 
 U_CAPI UChar32 U_EXPORT2

-ucase_fold(const UCaseProps *csp, UChar32 c, uint32_t options);
+ucase_fold(UChar32 c, uint32_t options);

 
 /**
  * Adds all simple case mappings and the full case folding for c to sa,
 
 /**
  * Adds all simple case mappings and the full case folding for c to sa,


@@ -108,8 +104,8 @@ ucase_fold(const UCaseProps *csp, UChar32 c, uint32_t options);
  * - for sharp s include ss
  * - for k include the Kelvin sign
  */
  * - for sharp s include ss
  * - for k include the Kelvin sign
  */

-U_CAPI void U_EXPORT2
-ucase_addCaseClosure(const UCaseProps *csp, UChar32 c, const USetAdder *sa);
+U_CFUNC void U_EXPORT2
+ucase_addCaseClosure(UChar32 c, const USetAdder *sa);

 
 /**
  * Maps the string to single code points and adds the associated case closure
 
 /**
  * Maps the string to single code points and adds the associated case closure


@@ -123,25 +119,84 @@ ucase_addCaseClosure(const UCaseProps *csp, UChar32 c, const USetAdder *sa);
  *
  * @return TRUE if the string was found
  */
  *
  * @return TRUE if the string was found
  */

-U_CAPI UBool U_EXPORT2
-ucase_addStringCaseClosure(const UCaseProps *csp, const UChar *s, int32_t length, const USetAdder *sa);
+U_CFUNC UBool U_EXPORT2
+ucase_addStringCaseClosure(const UChar *s, int32_t length, const USetAdder *sa);
+
+#ifdef __cplusplus
+U_NAMESPACE_BEGIN
+
+/**
+ * Iterator over characters with more than one code point in the full default Case_Folding.
+ */
+class U_COMMON_API FullCaseFoldingIterator {
+public:
+    /** Constructor. */
+    FullCaseFoldingIterator();
+    /**
+     * Returns the next (cp, full) pair where "full" is cp's full default Case_Folding.
+     * Returns a negative cp value at the end of the iteration.
+     */
+    UChar32 next(UnicodeString &full);
+private:
+    FullCaseFoldingIterator(const FullCaseFoldingIterator &);  // no copy
+    FullCaseFoldingIterator &operator=(const FullCaseFoldingIterator &);  // no assignment
+
+    const UChar *unfold;
+    int32_t unfoldRows;
+    int32_t unfoldRowWidth;
+    int32_t unfoldStringWidth;
+    int32_t currentRow;
+    int32_t rowCpIndex;
+};
+
+/**
+ * Fast case mapping data for ASCII/Latin.
+ * Linear arrays of delta bytes: 0=no mapping; EXC=exception.
+ * Deltas must not cross the ASCII boundary, or else they cannot be easily used
+ * in simple UTF-8 code.
+ */
+namespace LatinCase {
+
+/** Case mapping/folding data for code points up to U+017F. */
+constexpr UChar LIMIT = 0x180;
+/** U+017F case-folds and uppercases crossing the ASCII boundary. */
+constexpr UChar LONG_S = 0x17f;
+/** Exception: Complex mapping, or too-large delta. */
+constexpr int8_t EXC = -0x80;
+
+/** Deltas for lowercasing for most locales, and default case folding. */
+extern const int8_t TO_LOWER_NORMAL[LIMIT];
+/** Deltas for lowercasing for tr/az/lt, and Turkic case folding. */
+extern const int8_t TO_LOWER_TR_LT[LIMIT];
+
+/** Deltas for uppercasing for most locales. */
+extern const int8_t TO_UPPER_NORMAL[LIMIT];
+/** Deltas for uppercasing for tr/az. */
+extern const int8_t TO_UPPER_TR[LIMIT];
+
+}  // namespace LatinCase
+
+U_NAMESPACE_END
+#endif

 
 /** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */
 U_CAPI int32_t U_EXPORT2
 
 /** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */
 U_CAPI int32_t U_EXPORT2

-ucase_getType(const UCaseProps *csp, UChar32 c);
+ucase_getType(UChar32 c);

 
 

-/** @return same as ucase_getType(), or <0 if c is case-ignorable */
+/** @return like ucase_getType() but also sets UCASE_IGNORABLE if c is case-ignorable */

 U_CAPI int32_t U_EXPORT2
 U_CAPI int32_t U_EXPORT2

-ucase_getTypeOrIgnorable(const UCaseProps *csp, UChar32 c);
+ucase_getTypeOrIgnorable(UChar32 c);

 
 U_CAPI UBool U_EXPORT2
 
 U_CAPI UBool U_EXPORT2

-ucase_isSoftDotted(const UCaseProps *csp, UChar32 c);
+ucase_isSoftDotted(UChar32 c);

 
 U_CAPI UBool U_EXPORT2
 
 U_CAPI UBool U_EXPORT2

-ucase_isCaseSensitive(const UCaseProps *csp, UChar32 c);
+ucase_isCaseSensitive(UChar32 c);

 
 /* string case mapping functions */
 
 
 /* string case mapping functions */
 

+U_CDECL_BEGIN
+

 /**
  * Iterator function for string case mappings, which need to look at the
  * context (surrounding text) of a given character for conditional mappings.
 /**
  * Iterator function for string case mappings, which need to look at the
  * context (surrounding text) of a given character for conditional mappings.


@@ -176,6 +231,10 @@ struct UCaseContext {
 };
 typedef struct UCaseContext UCaseContext;
 
 };
 typedef struct UCaseContext UCaseContext;
 

+U_CDECL_END
+
+#define UCASECONTEXT_INITIALIZER { NULL,  0, 0, 0,  0, 0,  0,  0, 0, 0 }
+

 enum {
     /**
      * For string case mappings, a single character (a code point) is mapped
 enum {
     /**
      * For string case mappings, a single character (a code point) is mapped


@@ -205,10 +264,7 @@ enum {
  * @param context Pointer to be passed into iter.
  * @param pString If the mapping result is a string, then the pointer is
  *                written to *pString.
  * @param context Pointer to be passed into iter.
  * @param pString If the mapping result is a string, then the pointer is
  *                written to *pString.

- * @param locale Locale ID for locale-dependent mappings.
- * @param locCache Initialize to 0; may be used to cache the result of parsing
- *                 the locale ID for subsequent calls.
- *                 Can be NULL.
+ * @param caseLocale Case locale value from ucase_getCaseLocale().

  * @return Output code point or string length, see UCASE_MAX_STRING_LENGTH.
  *
  * @see UCaseContextIterator
  * @return Output code point or string length, see UCASE_MAX_STRING_LENGTH.
  *
  * @see UCaseContextIterator


@@ -216,31 +272,45 @@ enum {
  * @internal
  */
 U_CAPI int32_t U_EXPORT2
  * @internal
  */
 U_CAPI int32_t U_EXPORT2

-ucase_toFullLower(const UCaseProps *csp, UChar32 c,
+ucase_toFullLower(UChar32 c,

                   UCaseContextIterator *iter, void *context,
                   const UChar **pString,
                   UCaseContextIterator *iter, void *context,
                   const UChar **pString,

-                  const char *locale, int32_t *locCache);
+                  int32_t caseLocale);

 
 U_CAPI int32_t U_EXPORT2
 
 U_CAPI int32_t U_EXPORT2

-ucase_toFullUpper(const UCaseProps *csp, UChar32 c,
+ucase_toFullUpper(UChar32 c,

                   UCaseContextIterator *iter, void *context,
                   const UChar **pString,
                   UCaseContextIterator *iter, void *context,
                   const UChar **pString,

-                  const char *locale, int32_t *locCache);
+                  int32_t caseLocale);

 
 U_CAPI int32_t U_EXPORT2
 
 U_CAPI int32_t U_EXPORT2

-ucase_toFullTitle(const UCaseProps *csp, UChar32 c,
+ucase_toFullTitle(UChar32 c,

                   UCaseContextIterator *iter, void *context,
                   const UChar **pString,
                   UCaseContextIterator *iter, void *context,
                   const UChar **pString,

-                  const char *locale, int32_t *locCache);
+                  int32_t caseLocale);

 
 U_CAPI int32_t U_EXPORT2
 
 U_CAPI int32_t U_EXPORT2

-ucase_toFullFolding(const UCaseProps *csp, UChar32 c,
+ucase_toFullFolding(UChar32 c,

                     const UChar **pString,
                     uint32_t options);
 
 U_CFUNC int32_t U_EXPORT2
 ucase_hasBinaryProperty(UChar32 c, UProperty which);
 
                     const UChar **pString,
                     uint32_t options);
 
 U_CFUNC int32_t U_EXPORT2
 ucase_hasBinaryProperty(UChar32 c, UProperty which);
 

+
+U_CDECL_BEGIN
+
+/**
+ * @internal
+ */
+typedef int32_t U_CALLCONV
+UCaseMapFull(UChar32 c,
+             UCaseContextIterator *iter, void *context,
+             const UChar **pString,
+             int32_t caseLocale);
+
+U_CDECL_END
+

 /* file definitions --------------------------------------------------------- */
 
 #define UCASE_DATA_NAME "ucase"
 /* file definitions --------------------------------------------------------- */
 
 #define UCASE_DATA_NAME "ucase"


@@ -266,6 +336,9 @@ enum {
 
 /* definitions for 16-bit case properties word ------------------------------ */
 
 
 /* definitions for 16-bit case properties word ------------------------------ */
 

+U_CFUNC const UTrie2 * U_EXPORT2
+ucase_getTrie();
+

 /* 2-bit constants for types of cased characters */
 #define UCASE_TYPE_MASK     3
 enum {
 /* 2-bit constants for types of cased characters */
 #define UCASE_TYPE_MASK     3
 enum {


@@ -276,33 +349,40 @@ enum {
 };
 
 #define UCASE_GET_TYPE(props) ((props)&UCASE_TYPE_MASK)
 };
 
 #define UCASE_GET_TYPE(props) ((props)&UCASE_TYPE_MASK)

+#define UCASE_GET_TYPE_AND_IGNORABLE(props) ((props)&7)
+
+#define UCASE_IS_UPPER_OR_TITLE(props) ((props)&2)

 
 

-#define UCASE_SENSITIVE     4
-#define UCASE_EXCEPTION     8
+#define UCASE_IGNORABLE         4
+#define UCASE_EXCEPTION         8
+#define UCASE_SENSITIVE         0x10

 
 

-#define UCASE_DOT_MASK      0x30
+#define UCASE_HAS_EXCEPTION(props) ((props)&UCASE_EXCEPTION)
+
+#define UCASE_DOT_MASK      0x60

 enum {
     UCASE_NO_DOT=0,         /* normal characters with cc=0 */
 enum {
     UCASE_NO_DOT=0,         /* normal characters with cc=0 */

-    UCASE_SOFT_DOTTED=0x10, /* soft-dotted characters with cc=0 */
-    UCASE_ABOVE=0x20,       /* "above" accents with cc=230 */
-    UCASE_OTHER_ACCENT=0x30 /* other accent character (0<cc!=230) */
+    UCASE_SOFT_DOTTED=0x20, /* soft-dotted characters with cc=0 */
+    UCASE_ABOVE=0x40,       /* "above" accents with cc=230 */
+    UCASE_OTHER_ACCENT=0x60 /* other accent character (0<cc!=230) */

 };
 
 };
 

-/* no exception: bits 15..6 are a 10-bit signed case mapping delta */
-#define UCASE_DELTA_SHIFT   6
-#define UCASE_DELTA_MASK    0xffc0
-#define UCASE_MAX_DELTA     0x1ff
+/* no exception: bits 15..7 are a 9-bit signed case mapping delta */
+#define UCASE_DELTA_SHIFT   7
+#define UCASE_DELTA_MASK    0xff80
+#define UCASE_MAX_DELTA     0xff

 #define UCASE_MIN_DELTA     (-UCASE_MAX_DELTA-1)
 
 #define UCASE_MIN_DELTA     (-UCASE_MAX_DELTA-1)
 

-#define UCASE_GET_DELTA(props) ((int16_t)(props)>>UCASE_DELTA_SHIFT)
-
-/* case-ignorable uses one of the delta bits, see gencase/store.c */
-#define UCASE_CASE_IGNORABLE 0x40
+#if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
+#   define UCASE_GET_DELTA(props) ((int16_t)(props)>>UCASE_DELTA_SHIFT)
+#else
+#   define UCASE_GET_DELTA(props) (int16_t)(((props)&0x8000) ? (((props)>>UCASE_DELTA_SHIFT)|0xfe00) : ((uint16_t)(props)>>UCASE_DELTA_SHIFT))
+#endif

 
 /* exception: bits 15..4 are an unsigned 12-bit index into the exceptions array */
 #define UCASE_EXC_SHIFT     4
 #define UCASE_EXC_MASK      0xfff0
 
 /* exception: bits 15..4 are an unsigned 12-bit index into the exceptions array */
 #define UCASE_EXC_SHIFT     4
 #define UCASE_EXC_MASK      0xfff0

-#define UCASE_MAX_EXCEPTIONS 0x1000
+#define UCASE_MAX_EXCEPTIONS ((UCASE_EXC_MASK>>UCASE_EXC_SHIFT)+1)

 
 /* definitions for 16-bit main exceptions word ------------------------------ */
 
 
 /* definitions for 16-bit main exceptions word ------------------------------ */
 


@@ -312,7 +392,7 @@ enum {
     UCASE_EXC_FOLD,
     UCASE_EXC_UPPER,
     UCASE_EXC_TITLE,
     UCASE_EXC_FOLD,
     UCASE_EXC_UPPER,
     UCASE_EXC_TITLE,

-    UCASE_EXC_4,            /* reserved */
+    UCASE_EXC_DELTA,

     UCASE_EXC_5,            /* reserved */
     UCASE_EXC_CLOSURE,
     UCASE_EXC_FULL_MAPPINGS,
     UCASE_EXC_5,            /* reserved */
     UCASE_EXC_CLOSURE,
     UCASE_EXC_FULL_MAPPINGS,


@@ -322,10 +402,14 @@ enum {
 /* each slot is 2 uint16_t instead of 1 */
 #define UCASE_EXC_DOUBLE_SLOTS      0x100
 
 /* each slot is 2 uint16_t instead of 1 */
 #define UCASE_EXC_DOUBLE_SLOTS      0x100
 

-/* reserved: exception bits 11..9 */
+enum {
+    UCASE_EXC_NO_SIMPLE_CASE_FOLDING=0x200,
+    UCASE_EXC_DELTA_IS_NEGATIVE=0x400,
+    UCASE_EXC_SENSITIVE=0x800
+};

 
 /* UCASE_EXC_DOT_MASK=UCASE_DOT_MASK<<UCASE_EXC_DOT_SHIFT */
 
 /* UCASE_EXC_DOT_MASK=UCASE_DOT_MASK<<UCASE_EXC_DOT_SHIFT */

-#define UCASE_EXC_DOT_SHIFT     8
+#define UCASE_EXC_DOT_SHIFT     7

 
 /* normally stored in the main word, but pushed out for larger exception indexes */
 #define UCASE_EXC_DOT_MASK      0x3000
 
 /* normally stored in the main word, but pushed out for larger exception indexes */
 #define UCASE_EXC_DOT_MASK      0x3000


@@ -357,6 +441,4 @@ enum {
     UCASE_UNFOLD_STRING_WIDTH
 };
 
     UCASE_UNFOLD_STRING_WIDTH
 };
 

-U_CDECL_END
-

 #endif
 #endif