To iterate over code points and strings, use a loop like this:
*
* UnicodeSetIterator it(set);
* while (it.next()) {
* processItem(it.getString());
* }
*
* Each item in the set is accessed as a string. Set elements
* consisting of single code points are returned as strings containing
* just the one code point.
*
*
To iterate over code point ranges, instead of individual code points,
* use a loop like this:
*
* UnicodeSetIterator it(set);
* while (it.nextRange()) {
* if (it.isString()) {
* processString(it.getString());
* } else {
* processCodepointRange(it.getCodepoint(), it.getCodepointEnd());
* }
* }
*
* @author M. Davis
* @stable ICU 2.4
*/
class U_COMMON_API UnicodeSetIterator : public UObject {
protected:
/**
* Value of codepoint if the iterator points to a string.
* If codepoint == IS_STRING, then examine
* string for the current iteration result.
* @stable ICU 2.4
*/
enum { IS_STRING = -1 };
/**
* Current code point, or the special value IS_STRING, if
* the iterator points to a string.
* @stable ICU 2.4
*/
UChar32 codepoint;
/**
* When iterating over ranges using nextRange(),
* codepointEnd contains the inclusive end of the
* iteration range, if codepoint != IS_STRING. If
* iterating over code points using next(), or if
* codepoint == IS_STRING, then the value of
* codepointEnd is undefined.
* @stable ICU 2.4
*/
UChar32 codepointEnd;
/**
* If codepoint == IS_STRING, then string points
* to the current string. If codepoint != IS_STRING, the
* value of string is undefined.
* @stable ICU 2.4
*/
const UnicodeString* string;
public:
/**
* Create an iterator over the given set. The iterator is valid
* only so long as set is valid.
* @param set set to iterate over
* @stable ICU 2.4
*/
UnicodeSetIterator(const UnicodeSet& set);
/**
* Create an iterator over nothing. next() and
* nextRange() return false. This is a convenience
* constructor allowing the target to be set later.
* @stable ICU 2.4
*/
UnicodeSetIterator();
/**
* Destructor.
* @stable ICU 2.4
*/
virtual ~UnicodeSetIterator();
/**
* Returns true if the current element is a string. If so, the
* caller can retrieve it with getString(). If this
* method returns false, the current element is a code point or
* code point range, depending on whether next() or
* nextRange() was called.
* Elements of types string and codepoint can both be retrieved
* with the function getString().
* Elements of type codepoint can also be retrieved with
* getCodepoint().
* For ranges, getCodepoint() returns the starting codepoint
* of the range, and getCodepointEnd() returns the end
* of the range.
* @stable ICU 2.4
*/
inline UBool isString() const;
/**
* Returns the current code point, if isString() returned
* false. Otherwise returns an undefined result.
* @stable ICU 2.4
*/
inline UChar32 getCodepoint() const;
/**
* Returns the end of the current code point range, if
* isString() returned false and nextRange() was
* called. Otherwise returns an undefined result.
* @stable ICU 2.4
*/
inline UChar32 getCodepointEnd() const;
/**
* Returns the current string, if isString() returned
* 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
*/
const UnicodeString& getString();
/**
* Advances the iteration position to the next element in the set,
* which can be either a single code point or a string.
* If there are no more elements in the set, return false.
*
*
* If isString() == TRUE, the value is a
* string, otherwise the value is a
* single code point. Elements of either type can be retrieved
* with the function getString(), while elements of
* consisting of a single code point can be retrieved with
* getCodepoint()
*
*
The order of iteration is all code points in sorted order,
* followed by all strings sorted order. Do not mix
* calls to next() and nextRange() without
* calling reset() between them. The results of doing so
* are undefined.
*
* @return true if there was another element in the set.
* @stable ICU 2.4
*/
UBool next();
/**
* Returns the next element in the set, either a code point range
* or a string. If there are no more elements in the set, return
* false. If isString() == TRUE, the value is a
* string and can be accessed with getString(). Otherwise the value is a
* range of one or more code points from getCodepoint() to
* getCodepointeEnd() inclusive.
*
*
The order of iteration is all code points ranges in sorted
* order, followed by all strings sorted order. Ranges are
* disjoint and non-contiguous. The value returned from getString()
* is undefined unless isString() == TRUE. Do not mix calls to
* next() and nextRange() without calling
* reset() between them. The results of doing so are
* undefined.
*
* @return true if there was another element in the set.
* @stable ICU 2.4
*/
UBool nextRange();
/**
* Sets this iterator to visit the elements of the given set and
* resets it to the start of that set. The iterator is valid only
* so long as set is valid.
* @param set the set to iterate over.
* @stable ICU 2.4
*/
void reset(const UnicodeSet& set);
/**
* Resets this iterator to the start of the set.
* @stable ICU 2.4
*/
void reset();
/**
* ICU "poor man's RTTI", returns a UClassID for this class.
*
* @stable ICU 2.4
*/
static UClassID U_EXPORT2 getStaticClassID();
/**
* ICU "poor man's RTTI", returns a UClassID for the actual class.
*
* @stable ICU 2.4
*/
virtual UClassID getDynamicClassID() const;
// ======================= PRIVATES ===========================
protected:
// endElement and nextElements are really UChar32's, but we keep
// them as signed int32_t's so we can do comparisons with
// endElement set to -1. Leave them as int32_t's.
/** The set
* @stable ICU 2.4
*/
const UnicodeSet* set;
/** End range
* @stable ICU 2.4
*/
int32_t endRange;
/** Range
* @stable ICU 2.4
*/
int32_t range;
/** End element
* @stable ICU 2.4
*/
int32_t endElement;
/** Next element
* @stable ICU 2.4
*/
int32_t nextElement;
//UBool abbreviated;
/** Next string
* @stable ICU 2.4
*/
int32_t nextString;
/** String count
* @stable ICU 2.4
*/
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
*/
UnicodeSetIterator(const UnicodeSetIterator&); // disallow
/** Assignment operator. Disallowed.
* @stable ICU 2.4
*/
UnicodeSetIterator& operator=(const UnicodeSetIterator&); // disallow
/** Load range
* @stable ICU 2.4
*/
virtual void loadRange(int32_t range);
};
inline UBool UnicodeSetIterator::isString() const {
return codepoint == (UChar32)IS_STRING;
}
inline UChar32 UnicodeSetIterator::getCodepoint() const {
return codepoint;
}
inline UChar32 UnicodeSetIterator::getCodepointEnd() const {
return codepointEnd;
}
U_NAMESPACE_END
#endif