jpayne@69: // © 2016 and later: Unicode, Inc. and others.
jpayne@69: // License & terms of use: http://www.unicode.org/copyright.html
jpayne@69: /*
jpayne@69:  *******************************************************************************
jpayne@69:  * Copyright (C) 1996-2014, International Business Machines Corporation and
jpayne@69:  * others. All Rights Reserved.
jpayne@69:  *******************************************************************************
jpayne@69:  */
jpayne@69: 
jpayne@69: #ifndef CANITER_H
jpayne@69: #define CANITER_H
jpayne@69: 
jpayne@69: #include "unicode/utypes.h"
jpayne@69: 
jpayne@69: #if U_SHOW_CPLUSPLUS_API
jpayne@69: 
jpayne@69: #if !UCONFIG_NO_NORMALIZATION
jpayne@69: 
jpayne@69: #include "unicode/uobject.h"
jpayne@69: #include "unicode/unistr.h"
jpayne@69: 
jpayne@69: /**
jpayne@69:  * \file
jpayne@69:  * \brief C++ API: Canonical Iterator
jpayne@69:  */
jpayne@69:  
jpayne@69: /** Should permutation skip characters with combining class zero
jpayne@69:  *  Should be either TRUE or FALSE. This is a compile time option
jpayne@69:  *  @stable ICU 2.4
jpayne@69:  */
jpayne@69: #ifndef CANITER_SKIP_ZEROES
jpayne@69: #define CANITER_SKIP_ZEROES TRUE
jpayne@69: #endif
jpayne@69: 
jpayne@69: U_NAMESPACE_BEGIN
jpayne@69: 
jpayne@69: class Hashtable;
jpayne@69: class Normalizer2;
jpayne@69: class Normalizer2Impl;
jpayne@69: 
jpayne@69: /**
jpayne@69:  * This class allows one to iterate through all the strings that are canonically equivalent to a given
jpayne@69:  * string. For example, here are some sample results:
jpayne@69: Results for: {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
jpayne@69: 1: \\u0041\\u030A\\u0064\\u0307\\u0327
jpayne@69:  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
jpayne@69: 2: \\u0041\\u030A\\u0064\\u0327\\u0307
jpayne@69:  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
jpayne@69: 3: \\u0041\\u030A\\u1E0B\\u0327
jpayne@69:  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
jpayne@69: 4: \\u0041\\u030A\\u1E11\\u0307
jpayne@69:  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
jpayne@69: 5: \\u00C5\\u0064\\u0307\\u0327
jpayne@69:  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
jpayne@69: 6: \\u00C5\\u0064\\u0327\\u0307
jpayne@69:  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
jpayne@69: 7: \\u00C5\\u1E0B\\u0327
jpayne@69:  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
jpayne@69: 8: \\u00C5\\u1E11\\u0307
jpayne@69:  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
jpayne@69: 9: \\u212B\\u0064\\u0307\\u0327
jpayne@69:  = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
jpayne@69: 10: \\u212B\\u0064\\u0327\\u0307
jpayne@69:  = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
jpayne@69: 11: \\u212B\\u1E0B\\u0327
jpayne@69:  = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
jpayne@69: 12: \\u212B\\u1E11\\u0307
jpayne@69:  = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
jpayne@69:  *<br>Note: the code is intended for use with small strings, and is not suitable for larger ones,
jpayne@69:  * since it has not been optimized for that situation.
jpayne@69:  * Note, CanonicalIterator is not intended to be subclassed.
jpayne@69:  * @author M. Davis
jpayne@69:  * @author C++ port by V. Weinstein
jpayne@69:  * @stable ICU 2.4
jpayne@69:  */
jpayne@69: class U_COMMON_API CanonicalIterator U_FINAL : public UObject {
jpayne@69: public:
jpayne@69:     /**
jpayne@69:      * Construct a CanonicalIterator object
jpayne@69:      * @param source    string to get results for
jpayne@69:      * @param status    Fill-in parameter which receives the status of this operation.
jpayne@69:      * @stable ICU 2.4
jpayne@69:      */
jpayne@69:     CanonicalIterator(const UnicodeString &source, UErrorCode &status);
jpayne@69: 
jpayne@69:     /** Destructor
jpayne@69:      *  Cleans pieces
jpayne@69:      * @stable ICU 2.4
jpayne@69:      */
jpayne@69:     virtual ~CanonicalIterator();
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * Gets the NFD form of the current source we are iterating over.
jpayne@69:      * @return gets the source: NOTE: it is the NFD form of source
jpayne@69:      * @stable ICU 2.4
jpayne@69:      */
jpayne@69:     UnicodeString getSource();
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * Resets the iterator so that one can start again from the beginning.
jpayne@69:      * @stable ICU 2.4
jpayne@69:      */
jpayne@69:     void reset();
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * Get the next canonically equivalent string.
jpayne@69:      * <br><b>Warning: The strings are not guaranteed to be in any particular order.</b>
jpayne@69:      * @return the next string that is canonically equivalent. A bogus string is returned when
jpayne@69:      * the iteration is done.
jpayne@69:      * @stable ICU 2.4
jpayne@69:      */
jpayne@69:     UnicodeString next();
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * Set a new source for this iterator. Allows object reuse.
jpayne@69:      * @param newSource     the source string to iterate against. This allows the same iterator to be used
jpayne@69:      *                     while changing the source string, saving object creation.
jpayne@69:      * @param status        Fill-in parameter which receives the status of this operation.
jpayne@69:      * @stable ICU 2.4
jpayne@69:      */
jpayne@69:     void setSource(const UnicodeString &newSource, UErrorCode &status);
jpayne@69: 
jpayne@69: #ifndef U_HIDE_INTERNAL_API
jpayne@69:     /**
jpayne@69:      * Dumb recursive implementation of permutation.
jpayne@69:      * TODO: optimize
jpayne@69:      * @param source     the string to find permutations for
jpayne@69:      * @param skipZeros  determine if skip zeros
jpayne@69:      * @param result     the results in a set.
jpayne@69:      * @param status       Fill-in parameter which receives the status of this operation.
jpayne@69:      * @internal
jpayne@69:      */
jpayne@69:     static void U_EXPORT2 permute(UnicodeString &source, UBool skipZeros, Hashtable *result, UErrorCode &status);
jpayne@69: #endif  /* U_HIDE_INTERNAL_API */
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * ICU "poor man's RTTI", returns a UClassID for this class.
jpayne@69:      *
jpayne@69:      * @stable ICU 2.2
jpayne@69:      */
jpayne@69:     static UClassID U_EXPORT2 getStaticClassID();
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * ICU "poor man's RTTI", returns a UClassID for the actual class.
jpayne@69:      *
jpayne@69:      * @stable ICU 2.2
jpayne@69:      */
jpayne@69:     virtual UClassID getDynamicClassID() const;
jpayne@69: 
jpayne@69: private:
jpayne@69:     // ===================== PRIVATES ==============================
jpayne@69:     // private default constructor
jpayne@69:     CanonicalIterator();
jpayne@69: 
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * Copy constructor. Private for now.
jpayne@69:      * @internal (private)
jpayne@69:      */
jpayne@69:     CanonicalIterator(const CanonicalIterator& other);
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * Assignment operator. Private for now.
jpayne@69:      * @internal (private)
jpayne@69:      */
jpayne@69:     CanonicalIterator& operator=(const CanonicalIterator& other);
jpayne@69: 
jpayne@69:     // fields
jpayne@69:     UnicodeString source;
jpayne@69:     UBool done;
jpayne@69: 
jpayne@69:     // 2 dimensional array holds the pieces of the string with
jpayne@69:     // their different canonically equivalent representations
jpayne@69:     UnicodeString **pieces;
jpayne@69:     int32_t pieces_length;
jpayne@69:     int32_t *pieces_lengths;
jpayne@69: 
jpayne@69:     // current is used in iterating to combine pieces
jpayne@69:     int32_t *current;
jpayne@69:     int32_t current_length;
jpayne@69: 
jpayne@69:     // transient fields
jpayne@69:     UnicodeString buffer;
jpayne@69: 
jpayne@69:     const Normalizer2 &nfd;
jpayne@69:     const Normalizer2Impl &nfcImpl;
jpayne@69: 
jpayne@69:     // we have a segment, in NFD. Find all the strings that are canonically equivalent to it.
jpayne@69:     UnicodeString *getEquivalents(const UnicodeString &segment, int32_t &result_len, UErrorCode &status); //private String[] getEquivalents(String segment)
jpayne@69: 
jpayne@69:     //Set getEquivalents2(String segment);
jpayne@69:     Hashtable *getEquivalents2(Hashtable *fillinResult, const char16_t *segment, int32_t segLen, UErrorCode &status);
jpayne@69:     //Hashtable *getEquivalents2(const UnicodeString &segment, int32_t segLen, UErrorCode &status);
jpayne@69: 
jpayne@69:     /**
jpayne@69:      * See if the decomposition of cp2 is at segment starting at segmentPos
jpayne@69:      * (with canonical rearrangment!)
jpayne@69:      * If so, take the remainder, and return the equivalents
jpayne@69:      */
jpayne@69:     //Set extract(int comp, String segment, int segmentPos, StringBuffer buffer);
jpayne@69:     Hashtable *extract(Hashtable *fillinResult, UChar32 comp, const char16_t *segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
jpayne@69:     //Hashtable *extract(UChar32 comp, const UnicodeString &segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
jpayne@69: 
jpayne@69:     void cleanPieces();
jpayne@69: 
jpayne@69: };
jpayne@69: 
jpayne@69: U_NAMESPACE_END
jpayne@69: 
jpayne@69: #endif /* #if !UCONFIG_NO_NORMALIZATION */
jpayne@69: 
jpayne@69: #endif /* U_SHOW_CPLUSPLUS_API */
jpayne@69: 
jpayne@69: #endif