1 /*
2  *******************************************************************************
3  * Copyright (C) 1996-2014, International Business Machines Corporation and
4  * others. All Rights Reserved.
5  *******************************************************************************
6  */
7 
8 #ifndef CANITER_H
9 #define CANITER_H
10 
11 #include "unicode/utypes.h"
12 
13 #if !UCONFIG_NO_NORMALIZATION
14 
15 #include "unicode/uobject.h"
16 #include "unicode/unistr.h"
17 
18 /**
19  * \file
20  * \brief C++ API: Canonical Iterator
21  */
22 
23 /** Should permutation skip characters with combining class zero
24  *  Should be either TRUE or FALSE. This is a compile time option
25  *  @stable ICU 2.4
26  */
27 #ifndef CANITER_SKIP_ZEROES
28 #define CANITER_SKIP_ZEROES TRUE
29 #endif
30 
31 U_NAMESPACE_BEGIN
32 
33 class Hashtable;
34 class Normalizer2;
35 class Normalizer2Impl;
36 
37 /**
38  * This class allows one to iterate through all the strings that are canonically equivalent to a given
39  * string. For example, here are some sample results:
40 Results for: {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
41 1: \\u0041\\u030A\\u0064\\u0307\\u0327
42  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
43 2: \\u0041\\u030A\\u0064\\u0327\\u0307
44  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
45 3: \\u0041\\u030A\\u1E0B\\u0327
46  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
47 4: \\u0041\\u030A\\u1E11\\u0307
48  = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
49 5: \\u00C5\\u0064\\u0307\\u0327
50  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
51 6: \\u00C5\\u0064\\u0327\\u0307
52  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
53 7: \\u00C5\\u1E0B\\u0327
54  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
55 8: \\u00C5\\u1E11\\u0307
56  = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
57 9: \\u212B\\u0064\\u0307\\u0327
58  = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
59 10: \\u212B\\u0064\\u0327\\u0307
60  = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
61 11: \\u212B\\u1E0B\\u0327
62  = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
63 12: \\u212B\\u1E11\\u0307
64  = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
65  *<br>Note: the code is intended for use with small strings, and is not suitable for larger ones,
66  * since it has not been optimized for that situation.
67  * Note, CanonicalIterator is not intended to be subclassed.
68  * @author M. Davis
69  * @author C++ port by V. Weinstein
70  * @stable ICU 2.4
71  */
72 class U_COMMON_API CanonicalIterator U_FINAL : public UObject {
73 public:
74     /**
75      * Construct a CanonicalIterator object
76      * @param source    string to get results for
77      * @param status    Fill-in parameter which receives the status of this operation.
78      * @stable ICU 2.4
79      */
80     CanonicalIterator(const UnicodeString &source, UErrorCode &status);
81 
82     /** Destructor
83      *  Cleans pieces
84      * @stable ICU 2.4
85      */
86     virtual ~CanonicalIterator();
87 
88     /**
89      * Gets the NFD form of the current source we are iterating over.
90      * @return gets the source: NOTE: it is the NFD form of source
91      * @stable ICU 2.4
92      */
93     UnicodeString getSource();
94 
95     /**
96      * Resets the iterator so that one can start again from the beginning.
97      * @stable ICU 2.4
98      */
99     void reset();
100 
101     /**
102      * Get the next canonically equivalent string.
103      * <br><b>Warning: The strings are not guaranteed to be in any particular order.</b>
104      * @return the next string that is canonically equivalent. A bogus string is returned when
105      * the iteration is done.
106      * @stable ICU 2.4
107      */
108     UnicodeString next();
109 
110     /**
111      * Set a new source for this iterator. Allows object reuse.
112      * @param newSource     the source string to iterate against. This allows the same iterator to be used
113      *                     while changing the source string, saving object creation.
114      * @param status        Fill-in parameter which receives the status of this operation.
115      * @stable ICU 2.4
116      */
117     void setSource(const UnicodeString &newSource, UErrorCode &status);
118 
119 #ifndef U_HIDE_INTERNAL_API
120     /**
121      * Dumb recursive implementation of permutation.
122      * TODO: optimize
123      * @param source     the string to find permutations for
124      * @param skipZeros  determine if skip zeros
125      * @param result     the results in a set.
126      * @param status       Fill-in parameter which receives the status of this operation.
127      * @internal
128      */
129     static void U_EXPORT2 permute(UnicodeString &source, UBool skipZeros, Hashtable *result, UErrorCode &status);
130 #endif  /* U_HIDE_INTERNAL_API */
131 
132     /**
133      * ICU "poor man's RTTI", returns a UClassID for this class.
134      *
135      * @stable ICU 2.2
136      */
137     static UClassID U_EXPORT2 getStaticClassID();
138 
139     /**
140      * ICU "poor man's RTTI", returns a UClassID for the actual class.
141      *
142      * @stable ICU 2.2
143      */
144     virtual UClassID getDynamicClassID() const;
145 
146 private:
147     // ===================== PRIVATES ==============================
148     // private default constructor
149     CanonicalIterator();
150 
151 
152     /**
153      * Copy constructor. Private for now.
154      * @internal
155      */
156     CanonicalIterator(const CanonicalIterator& other);
157 
158     /**
159      * Assignment operator. Private for now.
160      * @internal
161      */
162     CanonicalIterator& operator=(const CanonicalIterator& other);
163 
164     // fields
165     UnicodeString source;
166     UBool done;
167 
168     // 2 dimensional array holds the pieces of the string with
169     // their different canonically equivalent representations
170     UnicodeString **pieces;
171     int32_t pieces_length;
172     int32_t *pieces_lengths;
173 
174     // current is used in iterating to combine pieces
175     int32_t *current;
176     int32_t current_length;
177 
178     // transient fields
179     UnicodeString buffer;
180 
181     const Normalizer2 &nfd;
182     const Normalizer2Impl &nfcImpl;
183 
184     // we have a segment, in NFD. Find all the strings that are canonically equivalent to it.
185     UnicodeString *getEquivalents(const UnicodeString &segment, int32_t &result_len, UErrorCode &status); //private String[] getEquivalents(String segment)
186 
187     //Set getEquivalents2(String segment);
188     Hashtable *getEquivalents2(Hashtable *fillinResult, const UChar *segment, int32_t segLen, UErrorCode &status);
189     //Hashtable *getEquivalents2(const UnicodeString &segment, int32_t segLen, UErrorCode &status);
190 
191     /**
192      * See if the decomposition of cp2 is at segment starting at segmentPos
193      * (with canonical rearrangment!)
194      * If so, take the remainder, and return the equivalents
195      */
196     //Set extract(int comp, String segment, int segmentPos, StringBuffer buffer);
197     Hashtable *extract(Hashtable *fillinResult, UChar32 comp, const UChar *segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
198     //Hashtable *extract(UChar32 comp, const UnicodeString &segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
199 
200     void cleanPieces();
201 
202 };
203 
204 U_NAMESPACE_END
205 
206 #endif /* #if !UCONFIG_NO_NORMALIZATION */
207 
208 #endif
209