2 ********************************************************************
4 * Copyright (C) 1997-2011, International Business Machines
5 * Corporation and others. All Rights Reserved.
7 ********************************************************************
13 #include "unicode/utypes.h"
14 #include "unicode/uobject.h"
15 #include "unicode/unistr.h"
18 * \brief C++ API: Character Iterator
23 * Abstract class that defines an API for forward-only iteration
25 * This is a minimal interface for iteration without random access
26 * or backwards iteration. It is especially useful for wrapping
27 * streams with converters into an object for collation or
30 * <p>Characters can be accessed in two ways: as code units or as
32 * Unicode code points are 21-bit integers and are the scalar values
33 * of Unicode characters. ICU uses the type UChar32 for them.
34 * Unicode code units are the storage units of a given
35 * Unicode/UCS Transformation Format (a character encoding scheme).
36 * With UTF-16, all code points can be represented with either one
37 * or two code units ("surrogates").
38 * String storage is typically based on code units, while properties
39 * of characters are typically determined using code point values.
40 * Some processes may be designed to work with sequences of code units,
41 * or it may be known that all characters that are important to an
42 * algorithm can be represented with single code units.
43 * Other processes will need to use the code point access functions.</p>
45 * <p>ForwardCharacterIterator provides nextPostInc() to access
46 * a code unit and advance an internal position into the text object,
47 * similar to a <code>return text[position++]</code>.<br>
48 * It provides next32PostInc() to access a code point and advance an internal
51 * <p>next32PostInc() assumes that the current position is that of
52 * the beginning of a code point, i.e., of its first code unit.
53 * After next32PostInc(), this will be true again.
54 * In general, access to code units and code points in the same
55 * iteration loop should not be mixed. In UTF-16, if the current position
56 * is on a second code unit (Low Surrogate), then only that code unit
57 * is returned even by next32PostInc().</p>
59 * <p>For iteration with either function, there are two ways to
60 * check for the end of the iteration. When there are no more
61 * characters in the text object:
63 * <li>The hasNext() function returns FALSE.</li>
64 * <li>nextPostInc() and next32PostInc() return DONE
65 * when one attempts to read beyond the end of the text object.</li>
70 * void function1(ForwardCharacterIterator &it) {
72 * while(it.hasNext()) {
73 * c=it.next32PostInc();
78 * void function1(ForwardCharacterIterator &it) {
80 * while((c=it.nextPostInc())!=ForwardCharacterIterator::DONE) {
89 class U_COMMON_API ForwardCharacterIterator
: public UObject
{
92 * Value returned by most of ForwardCharacterIterator's functions
93 * when the iterator has reached the limits of its iteration.
96 enum { DONE
= 0xffff };
102 virtual ~ForwardCharacterIterator();
105 * Returns true when both iterators refer to the same
106 * character in the same character-storage object.
107 * @param that The ForwardCharacterIterator to be compared for equality
108 * @return true when both iterators refer to the same
109 * character in the same character-storage object
112 virtual UBool
operator==(const ForwardCharacterIterator
& that
) const = 0;
115 * Returns true when the iterators refer to different
116 * text-storage objects, or to different characters in the
117 * same text-storage object.
118 * @param that The ForwardCharacterIterator to be compared for inequality
119 * @return true when the iterators refer to different
120 * text-storage objects, or to different characters in the
121 * same text-storage object
124 inline UBool
operator!=(const ForwardCharacterIterator
& that
) const;
127 * Generates a hash code for this iterator.
128 * @return the hash code.
131 virtual int32_t hashCode(void) const = 0;
134 * Returns a UClassID for this ForwardCharacterIterator ("poor man's
135 * RTTI").<P> Despite the fact that this function is public,
136 * DO NOT CONSIDER IT PART OF CHARACTERITERATOR'S API!
137 * @return a UClassID for this ForwardCharacterIterator
140 virtual UClassID
getDynamicClassID(void) const = 0;
143 * Gets the current code unit for returning and advances to the next code unit
144 * in the iteration range
145 * (toward endIndex()). If there are
146 * no more code units to return, returns DONE.
147 * @return the current code unit.
150 virtual UChar
nextPostInc(void) = 0;
153 * Gets the current code point for returning and advances to the next code point
154 * in the iteration range
155 * (toward endIndex()). If there are
156 * no more code points to return, returns DONE.
157 * @return the current code point.
160 virtual UChar32
next32PostInc(void) = 0;
163 * Returns FALSE if there are no more code units or code points
164 * at or after the current position in the iteration range.
165 * This is used with nextPostInc() or next32PostInc() in forward
167 * @returns FALSE if there are no more code units or code points
168 * at or after the current position in the iteration range.
171 virtual UBool
hasNext() = 0;
174 /** Default constructor to be overridden in the implementing class. @stable ICU 2.0*/
175 ForwardCharacterIterator();
177 /** Copy constructor to be overridden in the implementing class. @stable ICU 2.0*/
178 ForwardCharacterIterator(const ForwardCharacterIterator
&other
);
181 * Assignment operator to be overridden in the implementing class.
184 ForwardCharacterIterator
&operator=(const ForwardCharacterIterator
&) { return *this; }
188 * Abstract class that defines an API for iteration
190 * This is an interface for forward and backward iteration
191 * and random access into a text object.
193 * <p>The API provides backward compatibility to the Java and older ICU
194 * CharacterIterator classes but extends them significantly:
196 * <li>CharacterIterator is now a subclass of ForwardCharacterIterator.</li>
197 * <li>While the old API functions provided forward iteration with
198 * "pre-increment" semantics, the new one also provides functions
199 * with "post-increment" semantics. They are more efficient and should
200 * be the preferred iterator functions for new implementations.
201 * The backward iteration always had "pre-decrement" semantics, which
202 * are efficient.</li>
203 * <li>Just like ForwardCharacterIterator, it provides access to
204 * both code units and code points. Code point access versions are available
205 * for the old and the new iteration semantics.</li>
206 * <li>There are new functions for setting and moving the current position
207 * without returning a character, for efficiency.</li>
210 * See ForwardCharacterIterator for examples for using the new forward iteration
211 * functions. For backward iteration, there is also a hasPrevious() function
212 * that can be used analogously to hasNext().
213 * The old functions work as before and are shown below.</p>
215 * <p>Examples for some of the new functions:</p>
217 * Forward iteration with hasNext():
219 * void forward1(CharacterIterator &it) {
221 * for(it.setToStart(); it.hasNext();) {
222 * c=it.next32PostInc();
227 * Forward iteration more similar to loops with the old forward iteration,
228 * showing a way to convert simple for() loops:
230 * void forward2(CharacterIterator &it) {
232 * for(c=it.firstPostInc(); c!=CharacterIterator::DONE; c=it.nextPostInc()) {
237 * Backward iteration with setToEnd() and hasPrevious():
239 * void backward1(CharacterIterator &it) {
241 * for(it.setToEnd(); it.hasPrevious();) {
247 * Backward iteration with a more traditional for() loop:
249 * void backward2(CharacterIterator &it) {
251 * for(c=it.last(); c!=CharacterIterator::DONE; c=it.previous()) {
257 * Example for random access:
259 * void random(CharacterIterator &it) {
260 * // set to the third code point from the beginning
261 * it.move32(3, CharacterIterator::kStart);
262 * // get a code point from here without moving the position
263 * UChar32 c=it.current32();
264 * // get the position
265 * int32_t pos=it.getIndex();
266 * // get the previous code unit
267 * UChar u=it.previous();
268 * // move back one more code unit
269 * it.move(-1, CharacterIterator::kCurrent);
270 * // set the position back to where it was
271 * // and read the same code point c and move beyond it
273 * if(c!=it.next32PostInc()) {
274 * exit(1); // CharacterIterator inconsistent
279 * <p>Examples, especially for the old API:</p>
281 * Function processing characters, in this example simple output
284 * void processChar( UChar c )
290 * Traverse the text from start to finish
293 * void traverseForward(CharacterIterator& iter)
295 * for(UChar c = iter.first(); c != CharacterIterator.DONE; c = iter.next()) {
301 * Traverse the text backwards, from end to start
304 * void traverseBackward(CharacterIterator& iter)
306 * for(UChar c = iter.last(); c != CharacterIterator.DONE; c = iter.previous()) {
312 * Traverse both forward and backward from a given position in the text.
313 * Calls to notBoundary() in this example represents some additional stopping criteria.
316 * void traverseOut(CharacterIterator& iter, int32_t pos)
319 * for (c = iter.setIndex(pos);
320 * c != CharacterIterator.DONE && (Unicode::isLetter(c) || Unicode::isDigit(c));
321 * c = iter.next()) {}
322 * int32_t end = iter.getIndex();
323 * for (c = iter.setIndex(pos);
324 * c != CharacterIterator.DONE && (Unicode::isLetter(c) || Unicode::isDigit(c));
325 * c = iter.previous()) {}
326 * int32_t start = iter.getIndex() + 1;
328 * cout << "start: " << start << " end: " << end << endl;
329 * for (c = iter.setIndex(start); iter.getIndex() < end; c = iter.next() ) {
335 * Creating a StringCharacterIterator and calling the test functions
338 * void CharacterIterator_Example( void )
340 * cout << endl << "===== CharacterIterator_Example: =====" << endl;
341 * UnicodeString text("Ein kleiner Satz.");
342 * StringCharacterIterator iterator(text);
343 * cout << "----- traverseForward: -----------" << endl;
344 * traverseForward( iterator );
345 * cout << endl << endl << "----- traverseBackward: ----------" << endl;
346 * traverseBackward( iterator );
347 * cout << endl << endl << "----- traverseOut: ---------------" << endl;
348 * traverseOut( iterator, 7 );
349 * cout << endl << endl << "-----" << endl;
356 class U_COMMON_API CharacterIterator
: public ForwardCharacterIterator
{
359 * Origin enumeration for the move() and move32() functions.
362 enum EOrigin
{ kStart
, kCurrent
, kEnd
};
368 virtual ~CharacterIterator();
371 * Returns a pointer to a new CharacterIterator of the same
372 * concrete class as this one, and referring to the same
373 * character in the same text-storage object as this one. The
374 * caller is responsible for deleting the new clone.
375 * @return a pointer to a new CharacterIterator
378 virtual CharacterIterator
* clone(void) const = 0;
381 * Sets the iterator to refer to the first code unit in its
382 * iteration range, and returns that code unit.
383 * This can be used to begin an iteration with next().
384 * @return the first code unit in its iteration range.
387 virtual UChar
first(void) = 0;
390 * Sets the iterator to refer to the first code unit in its
391 * iteration range, returns that code unit, and moves the position
392 * to the second code unit. This is an alternative to setToStart()
393 * for forward iteration with nextPostInc().
394 * @return the first code unit in its iteration range.
397 virtual UChar
firstPostInc(void);
400 * Sets the iterator to refer to the first code point in its
401 * iteration range, and returns that code unit,
402 * This can be used to begin an iteration with next32().
403 * Note that an iteration with next32PostInc(), beginning with,
404 * e.g., setToStart() or firstPostInc(), is more efficient.
405 * @return the first code point in its iteration range.
408 virtual UChar32
first32(void) = 0;
411 * Sets the iterator to refer to the first code point in its
412 * iteration range, returns that code point, and moves the position
413 * to the second code point. This is an alternative to setToStart()
414 * for forward iteration with next32PostInc().
415 * @return the first code point in its iteration range.
418 virtual UChar32
first32PostInc(void);
421 * Sets the iterator to refer to the first code unit or code point in its
422 * iteration range. This can be used to begin a forward
423 * iteration with nextPostInc() or next32PostInc().
424 * @return the start position of the iteration range
427 inline int32_t setToStart();
430 * Sets the iterator to refer to the last code unit in its
431 * iteration range, and returns that code unit.
432 * This can be used to begin an iteration with previous().
433 * @return the last code unit.
436 virtual UChar
last(void) = 0;
439 * Sets the iterator to refer to the last code point in its
440 * iteration range, and returns that code unit.
441 * This can be used to begin an iteration with previous32().
442 * @return the last code point.
445 virtual UChar32
last32(void) = 0;
448 * Sets the iterator to the end of its iteration range, just behind
449 * the last code unit or code point. This can be used to begin a backward
450 * iteration with previous() or previous32().
451 * @return the end position of the iteration range
454 inline int32_t setToEnd();
457 * Sets the iterator to refer to the "position"-th code unit
458 * in the text-storage object the iterator refers to, and
459 * returns that code unit.
460 * @param position the "position"-th code unit in the text-storage object
461 * @return the "position"-th code unit.
464 virtual UChar
setIndex(int32_t position
) = 0;
467 * Sets the iterator to refer to the beginning of the code point
468 * that contains the "position"-th code unit
469 * in the text-storage object the iterator refers to, and
470 * returns that code point.
471 * The current position is adjusted to the beginning of the code point
472 * (its first code unit).
473 * @param position the "position"-th code unit in the text-storage object
474 * @return the "position"-th code point.
477 virtual UChar32
setIndex32(int32_t position
) = 0;
480 * Returns the code unit the iterator currently refers to.
481 * @return the current code unit.
484 virtual UChar
current(void) const = 0;
487 * Returns the code point the iterator currently refers to.
488 * @return the current code point.
491 virtual UChar32
current32(void) const = 0;
494 * Advances to the next code unit in the iteration range
495 * (toward endIndex()), and returns that code unit. If there are
496 * no more code units to return, returns DONE.
497 * @return the next code unit.
500 virtual UChar
next(void) = 0;
503 * Advances to the next code point in the iteration range
504 * (toward endIndex()), and returns that code point. If there are
505 * no more code points to return, returns DONE.
506 * Note that iteration with "pre-increment" semantics is less
507 * efficient than iteration with "post-increment" semantics
508 * that is provided by next32PostInc().
509 * @return the next code point.
512 virtual UChar32
next32(void) = 0;
515 * Advances to the previous code unit in the iteration range
516 * (toward startIndex()), and returns that code unit. If there are
517 * no more code units to return, returns DONE.
518 * @return the previous code unit.
521 virtual UChar
previous(void) = 0;
524 * Advances to the previous code point in the iteration range
525 * (toward startIndex()), and returns that code point. If there are
526 * no more code points to return, returns DONE.
527 * @return the previous code point.
530 virtual UChar32
previous32(void) = 0;
533 * Returns FALSE if there are no more code units or code points
534 * before the current position in the iteration range.
535 * This is used with previous() or previous32() in backward
537 * @return FALSE if there are no more code units or code points
538 * before the current position in the iteration range, return TRUE otherwise.
541 virtual UBool
hasPrevious() = 0;
544 * Returns the numeric index in the underlying text-storage
545 * object of the character returned by first(). Since it's
546 * possible to create an iterator that iterates across only
547 * part of a text-storage object, this number isn't
549 * @returns the numeric index in the underlying text-storage
550 * object of the character returned by first().
553 inline int32_t startIndex(void) const;
556 * Returns the numeric index in the underlying text-storage
557 * object of the position immediately BEYOND the character
558 * returned by last().
559 * @return the numeric index in the underlying text-storage
560 * object of the position immediately BEYOND the character
561 * returned by last().
564 inline int32_t endIndex(void) const;
567 * Returns the numeric index in the underlying text-storage
568 * object of the character the iterator currently refers to
569 * (i.e., the character returned by current()).
570 * @return the numberic index in the text-storage object of
571 * the character the iterator currently refers to
574 inline int32_t getIndex(void) const;
577 * Returns the length of the entire text in the underlying
578 * text-storage object.
579 * @return the length of the entire text in the text-storage object
582 inline int32_t getLength() const;
585 * Moves the current position relative to the start or end of the
586 * iteration range, or relative to the current position itself.
587 * The movement is expressed in numbers of code units forward
588 * or backward by specifying a positive or negative delta.
589 * @param delta the position relative to origin. A positive delta means forward;
590 * a negative delta means backward.
591 * @param origin Origin enumeration {kStart, kCurrent, kEnd}
592 * @return the new position
595 virtual int32_t move(int32_t delta
, EOrigin origin
) = 0;
598 * Moves the current position relative to the start or end of the
599 * iteration range, or relative to the current position itself.
600 * The movement is expressed in numbers of code points forward
601 * or backward by specifying a positive or negative delta.
602 * @param delta the position relative to origin. A positive delta means forward;
603 * a negative delta means backward.
604 * @param origin Origin enumeration {kStart, kCurrent, kEnd}
605 * @return the new position
608 virtual int32_t move32(int32_t delta
, EOrigin origin
) = 0;
611 * Copies the text under iteration into the UnicodeString
612 * referred to by "result".
613 * @param result Receives a copy of the text under iteration.
616 virtual void getText(UnicodeString
& result
) = 0;
626 * Constructor, just setting the length field in this base class.
629 CharacterIterator(int32_t length
);
632 * Constructor, just setting the length and position fields in this base class.
635 CharacterIterator(int32_t length
, int32_t position
);
638 * Constructor, just setting the length, start, end, and position fields in this base class.
641 CharacterIterator(int32_t length
, int32_t textBegin
, int32_t textEnd
, int32_t position
);
646 * @param that The CharacterIterator to be copied
649 CharacterIterator(const CharacterIterator
&that
);
652 * Assignment operator. Sets this CharacterIterator to have the same behavior,
653 * as the one passed in.
654 * @param that The CharacterIterator passed in.
655 * @return the newly set CharacterIterator.
658 CharacterIterator
&operator=(const CharacterIterator
&that
);
661 * Base class text length field.
662 * Necessary this for correct getText() and hashCode().
668 * Base class field for the current position.
674 * Base class field for the start of the iteration range.
680 * Base class field for the end of the iteration range.
687 ForwardCharacterIterator::operator!=(const ForwardCharacterIterator
& that
) const {
688 return !operator==(that
);
692 CharacterIterator::setToStart() {
693 return move(0, kStart
);
697 CharacterIterator::setToEnd() {
698 return move(0, kEnd
);
702 CharacterIterator::startIndex(void) const {
707 CharacterIterator::endIndex(void) const {
712 CharacterIterator::getIndex(void) const {
717 CharacterIterator::getLength(void) const {