X-Git-Url: https://git.saurik.com/apple/icu.git/blobdiff_plain/46f4442e9a5a4f3b98b7c1083586332f6a8a99a4..2ca993e82fb37b597a3c73ecd1586a139a6579c5:/icuSources/common/unicode/uset.h

diff --git a/icuSources/common/unicode/uset.h b/icuSources/common/unicode/uset.h
index 2bbfd7a5..eb3c9e6a 100644
--- a/icuSources/common/unicode/uset.h
+++ b/icuSources/common/unicode/uset.h
@@ -1,7 +1,7 @@
 /*
 *******************************************************************************
 *
-*   Copyright (C) 2002-2008, International Business Machines
+*   Copyright (C) 2002-2014, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 *
 *******************************************************************************
@@ -29,6 +29,7 @@
 
 #include "unicode/utypes.h"
 #include "unicode/uchar.h"
+#include "unicode/localpointer.h"
 
 #ifndef UCNV_H
 struct USet;
@@ -88,13 +89,7 @@ enum {
      * of each existing element in the set.
      * @stable ICU 3.2
      */
-    USET_ADD_CASE_MAPPINGS = 4,
-    
-    /**
-     * Enough for any single-code point set
-     * @internal
-     */
-    USET_SERIALIZED_STATIC_ARRAY_CAPACITY=8
+    USET_ADD_CASE_MAPPINGS = 4
 };
 
 /**
@@ -103,8 +98,8 @@ enum {
  *
  * The functionality is straightforward for sets with only single code points,
  * without strings (which is the common case):
- * - USET_SPAN_CONTAINED and USET_SPAN_SIMPLE
- *   work the same.
+ * - USET_SPAN_CONTAINED and USET_SPAN_SIMPLE work the same.
+ * - USET_SPAN_CONTAINED and USET_SPAN_SIMPLE are inverses of USET_SPAN_NOT_CONTAINED.
  * - span() and spanBack() partition any string the same way when
  *   alternating between span(USET_SPAN_NOT_CONTAINED) and
  *   span(either "contained" condition).
@@ -150,11 +145,12 @@ enum {
  * (strings with unpaired surrogates which cannot be converted to UTF-8)
  * are ignored.
  *
- * @stable ICU 4.0
+ * @stable ICU 3.8
  */
 typedef enum USetSpanCondition {
     /**
-     * Continue a span() while there is no set element at the current position.
+     * Continues a span() while there is no set element at the current position.
+     * Increments by one code point at a time.
      * Stops before the first set element (character or string).
      * (For code points only, this is like while contains(current)==FALSE).
      *
@@ -162,26 +158,27 @@ typedef enum USetSpanCondition {
      * it returned consists only of characters that are not in the set,
      * and none of its strings overlap with the span.
      *
-     * @stable ICU 4.0
+     * @stable ICU 3.8
      */
     USET_SPAN_NOT_CONTAINED = 0,
     /**
-     * Continue a span() while there is a set element at the current position.
+     * Spans the longest substring that is a concatenation of set elements (characters or strings).
      * (For characters only, this is like while contains(current)==TRUE).
      *
      * When span() returns, the substring between where it started and the position
      * it returned consists only of set elements (characters or strings) that are in the set.
      *
-     * If a set contains strings, then the span will be the longest substring
-     * matching any of the possible concatenations of set elements (characters or strings).
-     * (There must be a single, non-overlapping concatenation of characters or strings.)
-     * This is equivalent to a POSIX regular expression for (OR of each set element)*.
+     * If a set contains strings, then the span will be the longest substring for which there
+     * exists at least one non-overlapping concatenation of set elements (characters or strings).
+     * This is equivalent to a POSIX regular expression for <code>(OR of each set element)*</code>.
+     * (Java/ICU/Perl regex stops at the first match of an OR.)
      *
-     * @stable ICU 4.0
+     * @stable ICU 3.8
      */
     USET_SPAN_CONTAINED = 1,
     /**
-     * Continue a span() while there is a set element at the current position.
+     * Continues a span() while there is a set element at the current position.
+     * Increments by the longest matching element at each position.
      * (For characters only, this is like while contains(current)==TRUE).
      *
      * When span() returns, the substring between where it started and the position
@@ -196,16 +193,26 @@ typedef enum USetSpanCondition {
      * Use this span condition together with other longest-match algorithms,
      * such as ICU converters (ucnv_getUnicodeSet()).
      *
-     * @stable ICU 4.0
+     * @stable ICU 3.8
      */
     USET_SPAN_SIMPLE = 2,
     /**
      * One more than the last span condition.
-     * @stable ICU 4.0
+     * @stable ICU 3.8
      */
     USET_SPAN_CONDITION_COUNT
 } USetSpanCondition;
 
+enum {
+    /**
+     * Capacity of USerializedSet::staticArray.
+     * Enough for any single-code point set.
+     * Also provides padding for nice sizeof(USerializedSet).
+     * @stable ICU 2.4
+     */
+    USET_SERIALIZED_STATIC_ARRAY_CAPACITY=8
+};
+
 /**
  * A serialized form of a Unicode set.  Limited manipulations are
  * possible directly on a serialized set.  See below.
@@ -238,10 +245,20 @@ typedef struct USerializedSet {
  * USet API
  *********************************************************************/
 
+/**
+ * Create an empty USet object.
+ * Equivalent to uset_open(1, 0).
+ * @return a newly created USet.  The caller must call uset_close() on
+ * it when done.
+ * @stable ICU 4.2
+ */
+U_STABLE USet* U_EXPORT2
+uset_openEmpty(void);
+
 /**
  * Creates a USet object that contains the range of characters
  * start..end, inclusive.  If <code>start > end</code> 
- * then an empty set is created.
+ * then an empty set is created (same as using uset_openEmpty()).
  * @param start first character of the range, inclusive
  * @param end last character of the range, inclusive
  * @return a newly created USet.  The caller must call uset_close() on
@@ -289,6 +306,25 @@ uset_openPatternOptions(const UChar* pattern, int32_t patternLength,
 U_STABLE void U_EXPORT2
 uset_close(USet* set);
 
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUSetPointer
+ * "Smart pointer" class, closes a USet via uset_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUSetPointer, USet, uset_close);
+
+U_NAMESPACE_END
+
+#endif
+
 /**
  * Returns a copy of this object.
  * If this set is frozen, then the clone will be frozen as well.
@@ -296,9 +332,9 @@ uset_close(USet* set);
  * @param set the original set
  * @return the newly allocated copy of the set
  * @see uset_cloneAsThawed
- * @stable ICU 4.0
+ * @stable ICU 3.8
  */
-U_DRAFT USet * U_EXPORT2
+U_STABLE USet * U_EXPORT2
 uset_clone(const USet *set);
 
 /**
@@ -308,9 +344,9 @@ uset_clone(const USet *set);
  * @return TRUE/FALSE for whether the set has been frozen
  * @see uset_freeze
  * @see uset_cloneAsThawed
- * @stable ICU 4.0
+ * @stable ICU 3.8
  */
-U_DRAFT UBool U_EXPORT2
+U_STABLE UBool U_EXPORT2
 uset_isFrozen(const USet *set);
 
 /**
@@ -325,9 +361,9 @@ uset_isFrozen(const USet *set);
  * @return the same set, now frozen
  * @see uset_isFrozen
  * @see uset_cloneAsThawed
- * @stable ICU 4.0
+ * @stable ICU 3.8
  */
-U_DRAFT void U_EXPORT2
+U_STABLE void U_EXPORT2
 uset_freeze(USet *set);
 
 /**
@@ -338,9 +374,9 @@ uset_freeze(USet *set);
  * @see uset_freeze
  * @see uset_isFrozen
  * @see uset_clone
- * @stable ICU 4.0
+ * @stable ICU 3.8
  */
-U_DRAFT USet * U_EXPORT2
+U_STABLE USet * U_EXPORT2
 uset_cloneAsThawed(const USet *set);
 
 /**
@@ -673,6 +709,44 @@ uset_complementAll(USet* set, const USet* complement);
 U_STABLE void U_EXPORT2
 uset_clear(USet* set);
 
+/**
+ * Close this set over the given attribute.  For the attribute
+ * USET_CASE, the result is to modify this set so that:
+ *
+ * 1. For each character or string 'a' in this set, all strings or
+ * characters 'b' such that foldCase(a) == foldCase(b) are added
+ * to this set.
+ *
+ * 2. For each string 'e' in the resulting set, if e !=
+ * foldCase(e), 'e' will be removed.
+ *
+ * Example: [aq\\u00DF{Bc}{bC}{Fi}] => [aAqQ\\u00DF\\uFB01{ss}{bc}{fi}]
+ *
+ * (Here foldCase(x) refers to the operation u_strFoldCase, and a
+ * == b denotes that the contents are the same, not pointer
+ * comparison.)
+ *
+ * A frozen set will not be modified.
+ *
+ * @param set the set
+ *
+ * @param attributes bitmask for attributes to close over.
+ * Currently only the USET_CASE bit is supported.  Any undefined bits
+ * are ignored.
+ * @stable ICU 4.2
+ */
+U_STABLE void U_EXPORT2
+uset_closeOver(USet* set, int32_t attributes);
+
+/**
+ * Remove all strings from this set.
+ *
+ * @param set the set
+ * @stable ICU 4.2
+ */
+U_STABLE void U_EXPORT2
+uset_removeAllStrings(USet* set);
+
 /**
  * Returns TRUE if the given USet contains no characters and no
  * strings.
@@ -736,12 +810,12 @@ uset_indexOf(const USet* set, UChar32 c);
  * out of range, return (UChar32)-1.  The inverse of this method is
  * <code>indexOf()</code>.
  * @param set the set
- * @param index an index from 0..size()-1 to obtain the char for
+ * @param charIndex an index from 0..size()-1 to obtain the char for
  * @return the character at the given index, or (UChar32)-1.
  * @stable ICU 3.2
  */
 U_STABLE UChar32 U_EXPORT2
-uset_charAt(const USet* set, int32_t index);
+uset_charAt(const USet* set, int32_t charIndex);
 
 /**
  * Returns the number of characters and strings contained in the given
@@ -851,10 +925,10 @@ uset_containsSome(const USet* set1, const USet* set2);
  * @param spanCondition specifies the containment condition
  * @return the length of the initial substring according to the spanCondition;
  *         0 if the start of the string does not fit the spanCondition
- * @stable ICU 4.0
+ * @stable ICU 3.8
  * @see USetSpanCondition
  */
-U_DRAFT int32_t U_EXPORT2
+U_STABLE int32_t U_EXPORT2
 uset_span(const USet *set, const UChar *s, int32_t length, USetSpanCondition spanCondition);
 
 /**
@@ -872,10 +946,10 @@ uset_span(const USet *set, const UChar *s, int32_t length, USetSpanCondition spa
  * @param spanCondition specifies the containment condition
  * @return the start of the trailing substring according to the spanCondition;
  *         the string length if the end of the string does not fit the spanCondition
- * @stable ICU 4.0
+ * @stable ICU 3.8
  * @see USetSpanCondition
  */
-U_DRAFT int32_t U_EXPORT2
+U_STABLE int32_t U_EXPORT2
 uset_spanBack(const USet *set, const UChar *s, int32_t length, USetSpanCondition spanCondition);
 
 /**
@@ -894,10 +968,10 @@ uset_spanBack(const USet *set, const UChar *s, int32_t length, USetSpanCondition
  * @param spanCondition specifies the containment condition
  * @return the length of the initial substring according to the spanCondition;
  *         0 if the start of the string does not fit the spanCondition
- * @stable ICU 4.0
+ * @stable ICU 3.8
  * @see USetSpanCondition
  */
-U_DRAFT int32_t U_EXPORT2
+U_STABLE int32_t U_EXPORT2
 uset_spanUTF8(const USet *set, const char *s, int32_t length, USetSpanCondition spanCondition);
 
 /**
@@ -915,10 +989,10 @@ uset_spanUTF8(const USet *set, const char *s, int32_t length, USetSpanCondition
  * @param spanCondition specifies the containment condition
  * @return the start of the trailing substring according to the spanCondition;
  *         the string length if the end of the string does not fit the spanCondition
- * @stable ICU 4.0
+ * @stable ICU 3.8
  * @see USetSpanCondition
  */
-U_DRAFT int32_t U_EXPORT2
+U_STABLE int32_t U_EXPORT2
 uset_spanBackUTF8(const USet *set, const char *s, int32_t length, USetSpanCondition spanCondition);
 
 /**