ICU-4152 UnicodeSetIterator::getString() works with code points

X-SVN-Rev: 18958

icu4c/source/common/unicode/usetiter.h

@@ -1,6 +1,6 @@
 /*
 **********************************************************************
-* Copyright (c) 2002-2005, International Business Machines
+* Copyright (c) 2002-2006, International Business Machines
 * Corporation and others.  All Rights Reserved.
 **********************************************************************
 */
@@ -28,6 +28,10 @@ class UnicodeString;
  * code points or ranges have been returned, it returns the
  * multicharacter strings of the UnicodSet, if any.
  *
+ * This class is not intended to be subclassed.  Consider any fields
+ *  or methods declared as "protected" to be private.  The use of
+ *  protected in this class is an artifact of history.
+ *
  * <p>To iterate over code points, use a loop like this:
  * <pre>
  * UnicodeSetIterator it(set);
@@ -145,10 +149,16 @@ class U_COMMON_API UnicodeSetIterator : public UObject {
 
     /**
      * Returns the current string, if <tt>isString()</tt> returned
-     * true.  Otherwise returns an undefined result.
+     * true.  If the current iteration item is a code point, a UnicodeString
+     * containing that single code point is returned.
+     *
+     * Ownership of the returned string remains with the iterator.
+     * The string is guaranteed to remain valid only until the iterator is
+     *   advanced to the next item, or until the iterator is deleted.
+     * 
      * @stable ICU 2.4
      */
-    inline const UnicodeString& getString() const;
+    const UnicodeString& getString();
 
     /**
      * Returns the next element in the set, either a single code point
@@ -259,6 +269,13 @@ class U_COMMON_API UnicodeSetIterator : public UObject {
      */
     int32_t stringCount;
 
+    /**
+     *  Points to the string to use when the caller asks for a
+     *  string and the current iteration item is a code point, not a string.
+     *  @internal
+     */
+    UnicodeString *cpString;
+
     /** Copy constructor. Disallowed.
      * @stable ICU 2.4
      */
@@ -288,9 +305,6 @@ inline UChar32 UnicodeSetIterator::getCodepointEnd() const {
     return codepointEnd;
 }
 
-inline const UnicodeString& UnicodeSetIterator::getString() const {
-    return *string;
-}
 
 U_NAMESPACE_END
 

icu4c/source/common/usetiter.cpp

@@ -1,6 +1,6 @@
 /*
 **********************************************************************
-* Copyright (c) 2002-2003, International Business Machines
+* Copyright (c) 2002-2006, International Business Machines
 * Corporation and others.  All Rights Reserved.
 **********************************************************************
 */
@@ -18,6 +18,7 @@ UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UnicodeSetIterator)
  * @param set set to iterate over
  */
 UnicodeSetIterator::UnicodeSetIterator(const UnicodeSet& uSet) {
+    cpString  = NULL;
     reset(uSet);
 }
 
@@ -26,11 +27,12 @@ UnicodeSetIterator::UnicodeSetIterator(const UnicodeSet& uSet) {
  */
 UnicodeSetIterator::UnicodeSetIterator() {
     this->set = NULL;
+    cpString  = NULL;
     reset();
 }
 
 UnicodeSetIterator::~UnicodeSetIterator() {
-    // Nothing to do
+    delete cpString;
 }
 
 /**
@@ -45,11 +47,13 @@ UnicodeSetIterator::~UnicodeSetIterator() {
 UBool UnicodeSetIterator::next() {
     if (nextElement <= endElement) {
         codepoint = codepointEnd = nextElement++;
+        string = NULL;
         return TRUE;
     }
     if (range < endRange) {
         loadRange(++range);
         codepoint = codepointEnd = nextElement++;
+        string = NULL;
         return TRUE;
     }
 
@@ -71,6 +75,7 @@ UBool UnicodeSetIterator::next() {
  * <br>Note also that the codepointEnd is undefined after calling this method.
  */
 UBool UnicodeSetIterator::nextRange() {
+    string = NULL;
     if (nextElement <= endElement) {
         codepointEnd = endElement;
         codepoint = nextElement;
@@ -118,6 +123,7 @@ void UnicodeSetIterator::reset() {
         loadRange(range);
     }
     nextString = 0;
+    string = NULL;
 }
 
 void UnicodeSetIterator::loadRange(int32_t iRange) {
@@ -125,6 +131,20 @@ void UnicodeSetIterator::loadRange(int32_t iRange) {
     endElement = set->getRangeEnd(iRange);
 }
 
+
+const UnicodeString& UnicodeSetIterator::UnicodeSetIterator::getString()  {
+    if (string==NULL && codepoint!=(UChar32)IS_STRING) {
+       if (cpString == NULL) {
+          cpString = new UnicodeString();
+       }
+       if (cpString != NULL) {
+          cpString->setTo((UChar32)codepoint);
+       }
+       string = cpString;
+    }
+    return *string;
+}
+
 U_NAMESPACE_END
 
 //eof

icu4c/source/test/intltest/transrt.cpp

@@ -1,6 +1,6 @@
 /*
 **********************************************************************
-*   Copyright (C) 2000-2005, International Business Machines
+*   Copyright (C) 2000-2006, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 **********************************************************************
 *   Date        Name        Description
@@ -271,6 +271,10 @@ UBool LegalGreek::isRho(UChar c) {
 }
 
 // AbbreviatedUnicodeSetIterator Interface ---------------------------------------------
+//
+//      Iterate over a UnicodeSet, only returning a sampling of the contained code points.
+//        density is the approximate total number of code points to returned for the entire set.
+//
 
 class AbbreviatedUnicodeSetIterator : public UnicodeSetIterator {
 public :
@@ -291,7 +295,7 @@ public :
 
 private :
     UBool abbreviated;
-    int32_t perRange;
+    int32_t perRange;           // The maximum number of code points to be returned from each range
     virtual void loadRange(int32_t range);
 
     /**