ICU-62141.0.1.tar.gz

[apple/icu.git] / icuSources / common / unicode / uchar.h

diff --git a/icuSources/common/unicode/uchar.h b/icuSources/common/unicode/uchar.h
index 8174ca23e6f6fff7c2602a0cc709c1f14f93ab02..6d31083e66ee0de032597aa1d28c229206cbd8e7 100644 (file)
--- a/icuSources/common/unicode/uchar.h
+++ b/icuSources/common/unicode/uchar.h
@@ -26,6 +26,7 @@
 #define UCHAR_H
 
 #include "unicode/utypes.h"
 #define UCHAR_H
 
 #include "unicode/utypes.h"

+#include "unicode/stringoptions.h"

 
 U_CDECL_BEGIN
 
 
 U_CDECL_BEGIN
 


@@ -41,7 +42,7 @@ U_CDECL_BEGIN
  * @see u_getUnicodeVersion
  * @stable ICU 2.0
  */
  * @see u_getUnicodeVersion
  * @stable ICU 2.0
  */

-#define U_UNICODE_VERSION "9.0"
+#define U_UNICODE_VERSION "11.0"

 
 /**
  * \file
 
 /**
  * \file


@@ -111,11 +112,11 @@ U_CDECL_BEGIN
  * Comparison:
  * - u_isUWhiteSpace=UCHAR_WHITE_SPACE: Unicode White_Space property;
  *       most of general categories "Z" (separators) + most whitespace ISO controls
  * Comparison:
  * - u_isUWhiteSpace=UCHAR_WHITE_SPACE: Unicode White_Space property;
  *       most of general categories "Z" (separators) + most whitespace ISO controls

- *       (including no-break spaces, but excluding IS1..IS4 and ZWSP)
+ *       (including no-break spaces, but excluding IS1..IS4)

  * - u_isWhitespace: Java isWhitespace; Z + whitespace ISO controls but excluding no-break spaces
  * - u_isJavaSpaceChar: Java isSpaceChar; just Z (including no-break spaces)
  * - u_isspace: Z + whitespace ISO controls (including no-break spaces)
  * - u_isWhitespace: Java isWhitespace; Z + whitespace ISO controls but excluding no-break spaces
  * - u_isJavaSpaceChar: Java isSpaceChar; just Z (including no-break spaces)
  * - u_isspace: Z + whitespace ISO controls (including no-break spaces)

- * - u_isblank: "horizontal spaces" = TAB + Zs - ZWSP
+ * - u_isblank: "horizontal spaces" = TAB + Zs

  */
 
 /**
  */
 
 /**


@@ -148,8 +149,9 @@ U_CDECL_BEGIN
  *
  * The properties APIs are intended to reflect Unicode properties as defined
  * in the Unicode Character Database (UCD) and Unicode Technical Reports (UTR).
  *
  * The properties APIs are intended to reflect Unicode properties as defined
  * in the Unicode Character Database (UCD) and Unicode Technical Reports (UTR).

- * For details about the properties see http://www.unicode.org/ucd/ .
- * For names of Unicode properties see the UCD file PropertyAliases.txt.
+ *
+ * For details about the properties see
+ * UAX #44: Unicode Character Database (http://www.unicode.org/reports/tr44/).

  *
  * Important: If ICU is built with UCD files from Unicode versions below, e.g., 3.2,
  * then properties marked with "new in Unicode 3.2" are not or not fully available.
  *
  * Important: If ICU is built with UCD files from Unicode versions below, e.g., 3.2,
  * then properties marked with "new in Unicode 3.2" are not or not fully available.


@@ -427,12 +429,36 @@ typedef enum UProperty {
      * @stable ICU 57
      */
     UCHAR_EMOJI_MODIFIER_BASE=60,
      * @stable ICU 57
      */
     UCHAR_EMOJI_MODIFIER_BASE=60,

+    /**
+     * Binary property Emoji_Component.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 60
+     */
+    UCHAR_EMOJI_COMPONENT=61,
+    /**
+     * Binary property Regional_Indicator.
+     * @stable ICU 60
+     */
+    UCHAR_REGIONAL_INDICATOR=62,
+    /**
+     * Binary property Prepended_Concatenation_Mark.
+     * @stable ICU 60
+     */
+    UCHAR_PREPENDED_CONCATENATION_MARK=63,
+    /**
+     * Binary property Extended_Pictographic.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 62
+     */
+    UCHAR_EXTENDED_PICTOGRAPHIC=64,

 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the last constant for binary Unicode properties.
      * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
      */
 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the last constant for binary Unicode properties.
      * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
      */

-    UCHAR_BINARY_LIMIT=61,
+    UCHAR_BINARY_LIMIT,

 #endif  // U_HIDE_DEPRECATED_API
 
     /** Enumerated property Bidi_Class.
 #endif  // U_HIDE_DEPRECATED_API
 
     /** Enumerated property Bidi_Class.


@@ -1647,6 +1673,48 @@ enum UBlockCode {
     /** @stable ICU 58 */
     UBLOCK_TANGUT_COMPONENTS = 273, /*[18800]*/
 
     /** @stable ICU 58 */
     UBLOCK_TANGUT_COMPONENTS = 273, /*[18800]*/
 

+    // New blocks in Unicode 10.0
+
+    /** @stable ICU 60 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_F = 274, /*[2CEB0]*/
+    /** @stable ICU 60 */
+    UBLOCK_KANA_EXTENDED_A = 275, /*[1B100]*/
+    /** @stable ICU 60 */
+    UBLOCK_MASARAM_GONDI = 276, /*[11D00]*/
+    /** @stable ICU 60 */
+    UBLOCK_NUSHU = 277, /*[1B170]*/
+    /** @stable ICU 60 */
+    UBLOCK_SOYOMBO = 278, /*[11A50]*/
+    /** @stable ICU 60 */
+    UBLOCK_SYRIAC_SUPPLEMENT = 279, /*[0860]*/
+    /** @stable ICU 60 */
+    UBLOCK_ZANABAZAR_SQUARE = 280, /*[11A00]*/
+
+    // New blocks in Unicode 11.0
+
+    /** @stable ICU 62 */
+    UBLOCK_CHESS_SYMBOLS = 281, /*[1FA00]*/
+    /** @stable ICU 62 */
+    UBLOCK_DOGRA = 282, /*[11800]*/
+    /** @stable ICU 62 */
+    UBLOCK_GEORGIAN_EXTENDED = 283, /*[1C90]*/
+    /** @stable ICU 62 */
+    UBLOCK_GUNJALA_GONDI = 284, /*[11D60]*/
+    /** @stable ICU 62 */
+    UBLOCK_HANIFI_ROHINGYA = 285, /*[10D00]*/
+    /** @stable ICU 62 */
+    UBLOCK_INDIC_SIYAQ_NUMBERS = 286, /*[1EC70]*/
+    /** @stable ICU 62 */
+    UBLOCK_MAKASAR = 287, /*[11EE0]*/
+    /** @stable ICU 62 */
+    UBLOCK_MAYAN_NUMERALS = 288, /*[1D2E0]*/
+    /** @stable ICU 62 */
+    UBLOCK_MEDEFAIDRIN = 289, /*[16E40]*/
+    /** @stable ICU 62 */
+    UBLOCK_OLD_SOGDIAN = 290, /*[10F00]*/
+    /** @stable ICU 62 */
+    UBLOCK_SOGDIAN = 291, /*[10F30]*/
+

 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UBlockCode value.
 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UBlockCode value.


@@ -1654,7 +1722,7 @@ enum UBlockCode {
      *
      * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
      */
      *
      * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
      */

-    UBLOCK_COUNT = 274,
+    UBLOCK_COUNT = 292,

 #endif  // U_HIDE_DEPRECATED_API
 
     /** @stable ICU 2.0 */
 #endif  // U_HIDE_DEPRECATED_API
 
     /** @stable ICU 2.0 */


@@ -1930,6 +1998,22 @@ typedef enum UJoiningGroup {
     U_JG_AFRICAN_FEH,  /**< @stable ICU 58 */
     U_JG_AFRICAN_NOON,  /**< @stable ICU 58 */
     U_JG_AFRICAN_QAF,  /**< @stable ICU 58 */
     U_JG_AFRICAN_FEH,  /**< @stable ICU 58 */
     U_JG_AFRICAN_NOON,  /**< @stable ICU 58 */
     U_JG_AFRICAN_QAF,  /**< @stable ICU 58 */

+
+    U_JG_MALAYALAM_BHA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_JA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_LLA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_LLLA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NGA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NNA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NNNA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NYA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_RA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_SSA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_TTA,  /**< @stable ICU 60 */
+
+    U_JG_HANIFI_ROHINGYA_KINNA_YA,  /**< @stable ICU 62 */
+    U_JG_HANIFI_ROHINGYA_PA,  /**< @stable ICU 62 */
+

 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UJoiningGroup value.
 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UJoiningGroup value.


@@ -1980,6 +2064,7 @@ typedef enum UGraphemeClusterBreak {
     U_GCB_GLUE_AFTER_ZWJ = 16,  /*[GAZ]*/
     /** @stable ICU 58 */
     U_GCB_ZWJ = 17,             /*[ZWJ]*/
     U_GCB_GLUE_AFTER_ZWJ = 16,  /*[GAZ]*/
     /** @stable ICU 58 */
     U_GCB_ZWJ = 17,             /*[ZWJ]*/

+

 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UGraphemeClusterBreak value.
 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UGraphemeClusterBreak value.


@@ -2041,6 +2126,9 @@ typedef enum UWordBreakValues {
     U_WB_GLUE_AFTER_ZWJ = 20,   /*[GAZ]*/
     /** @stable ICU 58 */
     U_WB_ZWJ = 21,              /*[ZWJ]*/
     U_WB_GLUE_AFTER_ZWJ = 20,   /*[GAZ]*/
     /** @stable ICU 58 */
     U_WB_ZWJ = 21,              /*[ZWJ]*/

+    /** @stable ICU 62 */
+    U_WB_WSEGSPACE = 22,        /*[WSEGSPACE]*/
+

 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UWordBreakValues value.
 #ifndef U_HIDE_DEPRECATED_API
     /**
      * One more than the highest normal UWordBreakValues value.


@@ -2048,7 +2136,7 @@ typedef enum UWordBreakValues {
      *
      * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
      */
      *
      * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
      */

-    U_WB_COUNT = 22
+    U_WB_COUNT = 23

 #endif  // U_HIDE_DEPRECATED_API
 } UWordBreakValues;
 
 #endif  // U_HIDE_DEPRECATED_API
 } UWordBreakValues;
 


@@ -2653,8 +2741,7 @@ u_isgraph(UChar32 c);
  *
  * same as
  *
  *
  * same as
  *

- * TRUE for U+0009 (TAB) and characters with general category "Zs" (space separators)
- * except Zero Width Space (ZWSP, U+200B).
+ * TRUE for U+0009 (TAB) and characters with general category "Zs" (space separators).

  *
  * Note: There are several ICU whitespace functions; please see the uchar.h
  * file documentation for a detailed comparison.
  *
  * Note: There are several ICU whitespace functions; please see the uchar.h
  * file documentation for a detailed comparison.


@@ -3521,27 +3608,6 @@ u_toupper(UChar32 c);
 U_STABLE UChar32 U_EXPORT2
 u_totitle(UChar32 c);
 
 U_STABLE UChar32 U_EXPORT2
 u_totitle(UChar32 c);
 

-/** Option value for case folding: use default mappings defined in CaseFolding.txt. @stable ICU 2.0 */
-#define U_FOLD_CASE_DEFAULT 0
-
-/**
- * Option value for case folding:
- *
- * Use the modified set of mappings provided in CaseFolding.txt to handle dotted I
- * and dotless i appropriately for Turkic languages (tr, az).
- *
- * Before Unicode 3.2, CaseFolding.txt contains mappings marked with 'I' that
- * are to be included for default mappings and
- * excluded for the Turkic-specific mappings.
- *
- * Unicode 3.2 CaseFolding.txt instead contains mappings marked with 'T' that
- * are to be excluded for default mappings and
- * included for the Turkic-specific mappings.
- *
- * @stable ICU 2.0
- */
-#define U_FOLD_CASE_EXCLUDE_SPECIAL_I 1
-

 /**
  * The given character is mapped to its case folding equivalent according to
  * UnicodeData.txt and CaseFolding.txt;
 /**
  * The given character is mapped to its case folding equivalent according to
  * UnicodeData.txt and CaseFolding.txt;