michael@0: /*
michael@0: *******************************************************************************
michael@0: *
michael@0: *   Copyright (C) 2002-2011 International Business Machines
michael@0: *   Corporation and others.  All Rights Reserved.
michael@0: *
michael@0: *******************************************************************************
michael@0: *   file name:  uiter.h
michael@0: *   encoding:   US-ASCII
michael@0: *   tab size:   8 (not used)
michael@0: *   indentation:4
michael@0: *
michael@0: *   created on: 2002jan18
michael@0: *   created by: Markus W. Scherer
michael@0: */
michael@0: 
michael@0: #ifndef __UITER_H__
michael@0: #define __UITER_H__
michael@0: 
michael@0: /**
michael@0:  * \file
michael@0:  * \brief C API: Unicode Character Iteration
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  */
michael@0: 
michael@0: #include "unicode/utypes.h"
michael@0: 
michael@0: #if U_SHOW_CPLUSPLUS_API
michael@0:     U_NAMESPACE_BEGIN
michael@0: 
michael@0:     class CharacterIterator;
michael@0:     class Replaceable;
michael@0: 
michael@0:     U_NAMESPACE_END
michael@0: #endif
michael@0: 
michael@0: U_CDECL_BEGIN
michael@0: 
michael@0: struct UCharIterator;
michael@0: typedef struct UCharIterator UCharIterator; /**< C typedef for struct UCharIterator. @stable ICU 2.1 */
michael@0: 
michael@0: /**
michael@0:  * Origin constants for UCharIterator.getIndex() and UCharIterator.move().
michael@0:  * @see UCharIteratorMove
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef enum UCharIteratorOrigin {
michael@0:     UITER_START, UITER_CURRENT, UITER_LIMIT, UITER_ZERO, UITER_LENGTH
michael@0: } UCharIteratorOrigin;
michael@0: 
michael@0: /** Constants for UCharIterator. @stable ICU 2.6 */
michael@0: enum {
michael@0:     /**
michael@0:      * Constant value that may be returned by UCharIteratorMove
michael@0:      * indicating that the final UTF-16 index is not known, but that the move succeeded.
michael@0:      * This can occur when moving relative to limit or length, or
michael@0:      * when moving relative to the current index after a setState()
michael@0:      * when the current UTF-16 index is not known.
michael@0:      *
michael@0:      * It would be very inefficient to have to count from the beginning of the text
michael@0:      * just to get the current/limit/length index after moving relative to it.
michael@0:      * The actual index can be determined with getIndex(UITER_CURRENT)
michael@0:      * which will count the UChars if necessary.
michael@0:      *
michael@0:      * @stable ICU 2.6
michael@0:      */
michael@0:     UITER_UNKNOWN_INDEX=-2
michael@0: };
michael@0: 
michael@0: 
michael@0: /**
michael@0:  * Constant for UCharIterator getState() indicating an error or
michael@0:  * an unknown state.
michael@0:  * Returned by uiter_getState()/UCharIteratorGetState
michael@0:  * when an error occurs.
michael@0:  * Also, some UCharIterator implementations may not be able to return
michael@0:  * a valid state for each position. This will be clearly documented
michael@0:  * for each such iterator (none of the public ones here).
michael@0:  *
michael@0:  * @stable ICU 2.6
michael@0:  */
michael@0: #define UITER_NO_STATE ((uint32_t)0xffffffff)
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.getIndex().
michael@0:  *
michael@0:  * Gets the current position, or the start or limit of the
michael@0:  * iteration range.
michael@0:  *
michael@0:  * This function may perform slowly for UITER_CURRENT after setState() was called,
michael@0:  * or for UITER_LENGTH, because an iterator implementation may have to count
michael@0:  * UChars if the underlying storage is not UTF-16.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @param origin get the 0, start, limit, length, or current index
michael@0:  * @return the requested index, or U_SENTINEL in an error condition
michael@0:  *
michael@0:  * @see UCharIteratorOrigin
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef int32_t U_CALLCONV
michael@0: UCharIteratorGetIndex(UCharIterator *iter, UCharIteratorOrigin origin);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.move().
michael@0:  *
michael@0:  * Use iter->move(iter, index, UITER_ZERO) like CharacterIterator::setIndex(index).
michael@0:  *
michael@0:  * Moves the current position relative to the start or limit of the
michael@0:  * iteration range, or relative to the current position itself.
michael@0:  * The movement is expressed in numbers of code units forward
michael@0:  * or backward by specifying a positive or negative delta.
michael@0:  * Out of bounds movement will be pinned to the start or limit.
michael@0:  *
michael@0:  * This function may perform slowly for moving relative to UITER_LENGTH
michael@0:  * because an iterator implementation may have to count the rest of the
michael@0:  * UChars if the native storage is not UTF-16.
michael@0:  *
michael@0:  * When moving relative to the limit or length, or
michael@0:  * relative to the current position after setState() was called,
michael@0:  * move() may return UITER_UNKNOWN_INDEX (-2) to avoid an inefficient
michael@0:  * determination of the actual UTF-16 index.
michael@0:  * The actual index can be determined with getIndex(UITER_CURRENT)
michael@0:  * which will count the UChars if necessary.
michael@0:  * See UITER_UNKNOWN_INDEX for details.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @param delta can be positive, zero, or negative
michael@0:  * @param origin move relative to the 0, start, limit, length, or current index
michael@0:  * @return the new index, or U_SENTINEL on an error condition,
michael@0:  *         or UITER_UNKNOWN_INDEX when the index is not known.
michael@0:  *
michael@0:  * @see UCharIteratorOrigin
michael@0:  * @see UCharIterator
michael@0:  * @see UITER_UNKNOWN_INDEX
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef int32_t U_CALLCONV
michael@0: UCharIteratorMove(UCharIterator *iter, int32_t delta, UCharIteratorOrigin origin);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.hasNext().
michael@0:  *
michael@0:  * Check if current() and next() can still
michael@0:  * return another code unit.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return boolean value for whether current() and next() can still return another code unit
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef UBool U_CALLCONV
michael@0: UCharIteratorHasNext(UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.hasPrevious().
michael@0:  *
michael@0:  * Check if previous() can still return another code unit.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return boolean value for whether previous() can still return another code unit
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef UBool U_CALLCONV
michael@0: UCharIteratorHasPrevious(UCharIterator *iter);
michael@0:  
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.current().
michael@0:  *
michael@0:  * Return the code unit at the current position,
michael@0:  * or U_SENTINEL if there is none (index is at the limit).
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the current code unit
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef UChar32 U_CALLCONV
michael@0: UCharIteratorCurrent(UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.next().
michael@0:  *
michael@0:  * Return the code unit at the current index and increment
michael@0:  * the index (post-increment, like s[i++]),
michael@0:  * or return U_SENTINEL if there is none (index is at the limit).
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the current code unit (and post-increment the current index)
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef UChar32 U_CALLCONV
michael@0: UCharIteratorNext(UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.previous().
michael@0:  *
michael@0:  * Decrement the index and return the code unit from there
michael@0:  * (pre-decrement, like s[--i]),
michael@0:  * or return U_SENTINEL if there is none (index is at the start).
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the previous code unit (after pre-decrementing the current index)
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef UChar32 U_CALLCONV
michael@0: UCharIteratorPrevious(UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.reservedFn().
michael@0:  * Reserved for future use.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @param something some integer argument
michael@0:  * @return some integer
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: typedef int32_t U_CALLCONV
michael@0: UCharIteratorReserved(UCharIterator *iter, int32_t something);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.getState().
michael@0:  *
michael@0:  * Get the "state" of the iterator in the form of a single 32-bit word.
michael@0:  * It is recommended that the state value be calculated to be as small as
michael@0:  * is feasible. For strings with limited lengths, fewer than 32 bits may
michael@0:  * be sufficient.
michael@0:  *
michael@0:  * This is used together with setState()/UCharIteratorSetState
michael@0:  * to save and restore the iterator position more efficiently than with
michael@0:  * getIndex()/move().
michael@0:  *
michael@0:  * The iterator state is defined as a uint32_t value because it is designed
michael@0:  * for use in ucol_nextSortKeyPart() which provides 32 bits to store the state
michael@0:  * of the character iterator.
michael@0:  *
michael@0:  * With some UCharIterator implementations (e.g., UTF-8),
michael@0:  * getting and setting the UTF-16 index with existing functions
michael@0:  * (getIndex(UITER_CURRENT) followed by move(pos, UITER_ZERO)) is possible but
michael@0:  * relatively slow because the iterator has to "walk" from a known index
michael@0:  * to the requested one.
michael@0:  * This takes more time the farther it needs to go.
michael@0:  *
michael@0:  * An opaque state value allows an iterator implementation to provide
michael@0:  * an internal index (UTF-8: the source byte array index) for
michael@0:  * fast, constant-time restoration.
michael@0:  *
michael@0:  * After calling setState(), a getIndex(UITER_CURRENT) may be slow because
michael@0:  * the UTF-16 index may not be restored as well, but the iterator can deliver
michael@0:  * the correct text contents and move relative to the current position
michael@0:  * without performance degradation.
michael@0:  *
michael@0:  * Some UCharIterator implementations may not be able to return
michael@0:  * a valid state for each position, in which case they return UITER_NO_STATE instead.
michael@0:  * This will be clearly documented for each such iterator (none of the public ones here).
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the state word
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see UCharIteratorSetState
michael@0:  * @see UITER_NO_STATE
michael@0:  * @stable ICU 2.6
michael@0:  */
michael@0: typedef uint32_t U_CALLCONV
michael@0: UCharIteratorGetState(const UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Function type declaration for UCharIterator.setState().
michael@0:  *
michael@0:  * Restore the "state" of the iterator using a state word from a getState() call.
michael@0:  * The iterator object need not be the same one as for which getState() was called,
michael@0:  * but it must be of the same type (set up using the same uiter_setXYZ function)
michael@0:  * and it must iterate over the same string
michael@0:  * (binary identical regardless of memory address).
michael@0:  * For more about the state word see UCharIteratorGetState.
michael@0:  *
michael@0:  * After calling setState(), a getIndex(UITER_CURRENT) may be slow because
michael@0:  * the UTF-16 index may not be restored as well, but the iterator can deliver
michael@0:  * the correct text contents and move relative to the current position
michael@0:  * without performance degradation.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @param state the state word from a getState() call
michael@0:  *              on a same-type, same-string iterator
michael@0:  * @param pErrorCode Must be a valid pointer to an error code value,
michael@0:  *                   which must not indicate a failure before the function call.
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see UCharIteratorGetState
michael@0:  * @stable ICU 2.6
michael@0:  */
michael@0: typedef void U_CALLCONV
michael@0: UCharIteratorSetState(UCharIterator *iter, uint32_t state, UErrorCode *pErrorCode);
michael@0: 
michael@0: 
michael@0: /**
michael@0:  * C API for code unit iteration.
michael@0:  * This can be used as a C wrapper around
michael@0:  * CharacterIterator, Replaceable, or implemented using simple strings, etc.
michael@0:  *
michael@0:  * There are two roles for using UCharIterator:
michael@0:  *
michael@0:  * A "provider" sets the necessary function pointers and controls the "protected"
michael@0:  * fields of the UCharIterator structure. A "provider" passes a UCharIterator
michael@0:  * into C APIs that need a UCharIterator as an abstract, flexible string interface.
michael@0:  *
michael@0:  * Implementations of such C APIs are "callers" of UCharIterator functions;
michael@0:  * they only use the "public" function pointers and never access the "protected"
michael@0:  * fields directly.
michael@0:  *
michael@0:  * The current() and next() functions only check the current index against the
michael@0:  * limit, and previous() only checks the current index against the start,
michael@0:  * to see if the iterator already reached the end of the iteration range.
michael@0:  *
michael@0:  * The assumption - in all iterators - is that the index is moved via the API,
michael@0:  * which means it won't go out of bounds, or the index is modified by
michael@0:  * user code that knows enough about the iterator implementation to set valid
michael@0:  * index values.
michael@0:  *
michael@0:  * UCharIterator functions return code unit values 0..0xffff,
michael@0:  * or U_SENTINEL if the iteration bounds are reached.
michael@0:  *
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: struct UCharIterator {
michael@0:     /**
michael@0:      * (protected) Pointer to string or wrapped object or similar.
michael@0:      * Not used by caller.
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     const void *context;
michael@0: 
michael@0:     /**
michael@0:      * (protected) Length of string or similar.
michael@0:      * Not used by caller.
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     int32_t length;
michael@0: 
michael@0:     /**
michael@0:      * (protected) Start index or similar.
michael@0:      * Not used by caller.
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     int32_t start;
michael@0: 
michael@0:     /**
michael@0:      * (protected) Current index or similar.
michael@0:      * Not used by caller.
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     int32_t index;
michael@0: 
michael@0:     /**
michael@0:      * (protected) Limit index or similar.
michael@0:      * Not used by caller.
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     int32_t limit;
michael@0: 
michael@0:     /**
michael@0:      * (protected) Used by UTF-8 iterators and possibly others.
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     int32_t reservedField;
michael@0: 
michael@0:     /**
michael@0:      * (public) Returns the current position or the
michael@0:      * start or limit index of the iteration range.
michael@0:      *
michael@0:      * @see UCharIteratorGetIndex
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorGetIndex *getIndex;
michael@0: 
michael@0:     /**
michael@0:      * (public) Moves the current position relative to the start or limit of the
michael@0:      * iteration range, or relative to the current position itself.
michael@0:      * The movement is expressed in numbers of code units forward
michael@0:      * or backward by specifying a positive or negative delta.
michael@0:      *
michael@0:      * @see UCharIteratorMove
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorMove *move;
michael@0: 
michael@0:     /**
michael@0:      * (public) Check if current() and next() can still
michael@0:      * return another code unit.
michael@0:      *
michael@0:      * @see UCharIteratorHasNext
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorHasNext *hasNext;
michael@0: 
michael@0:     /**
michael@0:      * (public) Check if previous() can still return another code unit.
michael@0:      *
michael@0:      * @see UCharIteratorHasPrevious
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorHasPrevious *hasPrevious;
michael@0: 
michael@0:     /**
michael@0:      * (public) Return the code unit at the current position,
michael@0:      * or U_SENTINEL if there is none (index is at the limit).
michael@0:      *
michael@0:      * @see UCharIteratorCurrent
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorCurrent *current;
michael@0: 
michael@0:     /**
michael@0:      * (public) Return the code unit at the current index and increment
michael@0:      * the index (post-increment, like s[i++]),
michael@0:      * or return U_SENTINEL if there is none (index is at the limit).
michael@0:      *
michael@0:      * @see UCharIteratorNext
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorNext *next;
michael@0: 
michael@0:     /**
michael@0:      * (public) Decrement the index and return the code unit from there
michael@0:      * (pre-decrement, like s[--i]),
michael@0:      * or return U_SENTINEL if there is none (index is at the start).
michael@0:      *
michael@0:      * @see UCharIteratorPrevious
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorPrevious *previous;
michael@0: 
michael@0:     /**
michael@0:      * (public) Reserved for future use. Currently NULL.
michael@0:      *
michael@0:      * @see UCharIteratorReserved
michael@0:      * @stable ICU 2.1
michael@0:      */
michael@0:     UCharIteratorReserved *reservedFn;
michael@0: 
michael@0:     /**
michael@0:      * (public) Return the state of the iterator, to be restored later with setState().
michael@0:      * This function pointer is NULL if the iterator does not implement it.
michael@0:      *
michael@0:      * @see UCharIteratorGet
michael@0:      * @stable ICU 2.6
michael@0:      */
michael@0:     UCharIteratorGetState *getState;
michael@0: 
michael@0:     /**
michael@0:      * (public) Restore the iterator state from the state word from a call
michael@0:      * to getState().
michael@0:      * This function pointer is NULL if the iterator does not implement it.
michael@0:      *
michael@0:      * @see UCharIteratorSet
michael@0:      * @stable ICU 2.6
michael@0:      */
michael@0:     UCharIteratorSetState *setState;
michael@0: };
michael@0: 
michael@0: /**
michael@0:  * Helper function for UCharIterator to get the code point
michael@0:  * at the current index.
michael@0:  *
michael@0:  * Return the code point that includes the code unit at the current position,
michael@0:  * or U_SENTINEL if there is none (index is at the limit).
michael@0:  * If the current code unit is a lead or trail surrogate,
michael@0:  * then the following or preceding surrogate is used to form
michael@0:  * the code point value.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the current code point
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see U16_GET
michael@0:  * @see UnicodeString::char32At()
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: U_STABLE UChar32 U_EXPORT2
michael@0: uiter_current32(UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Helper function for UCharIterator to get the next code point.
michael@0:  *
michael@0:  * Return the code point at the current index and increment
michael@0:  * the index (post-increment, like s[i++]),
michael@0:  * or return U_SENTINEL if there is none (index is at the limit).
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the current code point (and post-increment the current index)
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see U16_NEXT
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: U_STABLE UChar32 U_EXPORT2
michael@0: uiter_next32(UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Helper function for UCharIterator to get the previous code point.
michael@0:  *
michael@0:  * Decrement the index and return the code point from there
michael@0:  * (pre-decrement, like s[--i]),
michael@0:  * or return U_SENTINEL if there is none (index is at the start).
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the previous code point (after pre-decrementing the current index)
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see U16_PREV
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: U_STABLE UChar32 U_EXPORT2
michael@0: uiter_previous32(UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Get the "state" of the iterator in the form of a single 32-bit word.
michael@0:  * This is a convenience function that calls iter->getState(iter)
michael@0:  * if iter->getState is not NULL;
michael@0:  * if it is NULL or any other error occurs, then UITER_NO_STATE is returned.
michael@0:  *
michael@0:  * Some UCharIterator implementations may not be able to return
michael@0:  * a valid state for each position, in which case they return UITER_NO_STATE instead.
michael@0:  * This will be clearly documented for each such iterator (none of the public ones here).
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @return the state word
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see UCharIteratorGetState
michael@0:  * @see UITER_NO_STATE
michael@0:  * @stable ICU 2.6
michael@0:  */
michael@0: U_STABLE uint32_t U_EXPORT2
michael@0: uiter_getState(const UCharIterator *iter);
michael@0: 
michael@0: /**
michael@0:  * Restore the "state" of the iterator using a state word from a getState() call.
michael@0:  * This is a convenience function that calls iter->setState(iter, state, pErrorCode)
michael@0:  * if iter->setState is not NULL; if it is NULL, then U_UNSUPPORTED_ERROR is set.
michael@0:  *
michael@0:  * @param iter the UCharIterator structure ("this pointer")
michael@0:  * @param state the state word from a getState() call
michael@0:  *              on a same-type, same-string iterator
michael@0:  * @param pErrorCode Must be a valid pointer to an error code value,
michael@0:  *                   which must not indicate a failure before the function call.
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see UCharIteratorSetState
michael@0:  * @stable ICU 2.6
michael@0:  */
michael@0: U_STABLE void U_EXPORT2
michael@0: uiter_setState(UCharIterator *iter, uint32_t state, UErrorCode *pErrorCode);
michael@0: 
michael@0: /**
michael@0:  * Set up a UCharIterator to iterate over a string.
michael@0:  *
michael@0:  * Sets the UCharIterator function pointers for iteration over the string s
michael@0:  * with iteration boundaries start=index=0 and length=limit=string length.
michael@0:  * The "provider" may set the start, index, and limit values at any time
michael@0:  * within the range 0..length.
michael@0:  * The length field will be ignored.
michael@0:  *
michael@0:  * The string pointer s is set into UCharIterator.context without copying
michael@0:  * or reallocating the string contents.
michael@0:  *
michael@0:  * getState() simply returns the current index.
michael@0:  * move() will always return the final index.
michael@0:  *
michael@0:  * @param iter UCharIterator structure to be set for iteration
michael@0:  * @param s String to iterate over
michael@0:  * @param length Length of s, or -1 if NUL-terminated
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: U_STABLE void U_EXPORT2
michael@0: uiter_setString(UCharIterator *iter, const UChar *s, int32_t length);
michael@0: 
michael@0: /**
michael@0:  * Set up a UCharIterator to iterate over a UTF-16BE string
michael@0:  * (byte vector with a big-endian pair of bytes per UChar).
michael@0:  *
michael@0:  * Everything works just like with a normal UChar iterator (uiter_setString),
michael@0:  * except that UChars are assembled from byte pairs,
michael@0:  * and that the length argument here indicates an even number of bytes.
michael@0:  *
michael@0:  * getState() simply returns the current index.
michael@0:  * move() will always return the final index.
michael@0:  *
michael@0:  * @param iter UCharIterator structure to be set for iteration
michael@0:  * @param s UTF-16BE string to iterate over
michael@0:  * @param length Length of s as an even number of bytes, or -1 if NUL-terminated
michael@0:  *               (NUL means pair of 0 bytes at even index from s)
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @see uiter_setString
michael@0:  * @stable ICU 2.6
michael@0:  */
michael@0: U_STABLE void U_EXPORT2
michael@0: uiter_setUTF16BE(UCharIterator *iter, const char *s, int32_t length);
michael@0: 
michael@0: /**
michael@0:  * Set up a UCharIterator to iterate over a UTF-8 string.
michael@0:  *
michael@0:  * Sets the UCharIterator function pointers for iteration over the UTF-8 string s
michael@0:  * with UTF-8 iteration boundaries 0 and length.
michael@0:  * The implementation counts the UTF-16 index on the fly and
michael@0:  * lazily evaluates the UTF-16 length of the text.
michael@0:  *
michael@0:  * The start field is used as the UTF-8 offset, the limit field as the UTF-8 length.
michael@0:  * When the reservedField is not 0, then it contains a supplementary code point
michael@0:  * and the UTF-16 index is between the two corresponding surrogates.
michael@0:  * At that point, the UTF-8 index is behind that code point.
michael@0:  *
michael@0:  * The UTF-8 string pointer s is set into UCharIterator.context without copying
michael@0:  * or reallocating the string contents.
michael@0:  *
michael@0:  * getState() returns a state value consisting of
michael@0:  * - the current UTF-8 source byte index (bits 31..1)
michael@0:  * - a flag (bit 0) that indicates whether the UChar position is in the middle
michael@0:  *   of a surrogate pair
michael@0:  *   (from a 4-byte UTF-8 sequence for the corresponding supplementary code point)
michael@0:  *
michael@0:  * getState() cannot also encode the UTF-16 index in the state value.
michael@0:  * move(relative to limit or length), or
michael@0:  * move(relative to current) after setState(), may return UITER_UNKNOWN_INDEX.
michael@0:  *
michael@0:  * @param iter UCharIterator structure to be set for iteration
michael@0:  * @param s UTF-8 string to iterate over
michael@0:  * @param length Length of s in bytes, or -1 if NUL-terminated
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.6
michael@0:  */
michael@0: U_STABLE void U_EXPORT2
michael@0: uiter_setUTF8(UCharIterator *iter, const char *s, int32_t length);
michael@0: 
michael@0: #if U_SHOW_CPLUSPLUS_API
michael@0: 
michael@0: /**
michael@0:  * Set up a UCharIterator to wrap around a C++ CharacterIterator.
michael@0:  *
michael@0:  * Sets the UCharIterator function pointers for iteration using the
michael@0:  * CharacterIterator charIter.
michael@0:  *
michael@0:  * The CharacterIterator pointer charIter is set into UCharIterator.context
michael@0:  * without copying or cloning the CharacterIterator object.
michael@0:  * The other "protected" UCharIterator fields are set to 0 and will be ignored.
michael@0:  * The iteration index and boundaries are controlled by the CharacterIterator.
michael@0:  *
michael@0:  * getState() simply returns the current index.
michael@0:  * move() will always return the final index.
michael@0:  *
michael@0:  * @param iter UCharIterator structure to be set for iteration
michael@0:  * @param charIter CharacterIterator to wrap
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: U_STABLE void U_EXPORT2
michael@0: uiter_setCharacterIterator(UCharIterator *iter, icu::CharacterIterator *charIter);
michael@0: 
michael@0: /**
michael@0:  * Set up a UCharIterator to iterate over a C++ Replaceable.
michael@0:  *
michael@0:  * Sets the UCharIterator function pointers for iteration over the
michael@0:  * Replaceable rep with iteration boundaries start=index=0 and
michael@0:  * length=limit=rep->length().
michael@0:  * The "provider" may set the start, index, and limit values at any time
michael@0:  * within the range 0..length=rep->length().
michael@0:  * The length field will be ignored.
michael@0:  *
michael@0:  * The Replaceable pointer rep is set into UCharIterator.context without copying
michael@0:  * or cloning/reallocating the Replaceable object.
michael@0:  *
michael@0:  * getState() simply returns the current index.
michael@0:  * move() will always return the final index.
michael@0:  *
michael@0:  * @param iter UCharIterator structure to be set for iteration
michael@0:  * @param rep Replaceable to iterate over
michael@0:  *
michael@0:  * @see UCharIterator
michael@0:  * @stable ICU 2.1
michael@0:  */
michael@0: U_STABLE void U_EXPORT2
michael@0: uiter_setReplaceable(UCharIterator *iter, const icu::Replaceable *rep);
michael@0: 
michael@0: #endif
michael@0: 
michael@0: U_CDECL_END
michael@0: 
michael@0: #endif