The Tor Browser / annotate

intl/icu/source/common/unicode/caniter.h@6474c204b198 (annotated)

intl/icu/source/common/unicode/caniter.h

Wed, 31 Dec 2014 06:09:35 +0100

author
Michael Schloh von Bennewitz <michael@schloh.com>
date
Wed, 31 Dec 2014 06:09:35 +0100
changeset 0
6474c204b198
permissions
-rw-r--r--

Cloned upstream origin tor-browser at tor-browser-31.3.0esr-4.5-1-build1
revision ID fc1c9ff7c1b2defdbc039f12214767608f46423f for hacking purpose.

michael@0 1 /*
michael@0 2 *******************************************************************************
michael@0 3 * Copyright (C) 1996-2011, International Business Machines Corporation and
michael@0 4 * others. All Rights Reserved.
michael@0 5 *******************************************************************************
michael@0 6 */
michael@0 7
michael@0 8 #ifndef CANITER_H
michael@0 9 #define CANITER_H
michael@0 10
michael@0 11 #include "unicode/utypes.h"
michael@0 12
michael@0 13 #if !UCONFIG_NO_NORMALIZATION
michael@0 14
michael@0 15 #include "unicode/uobject.h"
michael@0 16 #include "unicode/unistr.h"
michael@0 17
michael@0 18 /**
michael@0 19 * \file
michael@0 20 * \brief C++ API: Canonical Iterator
michael@0 21 */
michael@0 22
michael@0 23 /** Should permutation skip characters with combining class zero
michael@0 24 * Should be either TRUE or FALSE. This is a compile time option
michael@0 25 * @stable ICU 2.4
michael@0 26 */
michael@0 27 #ifndef CANITER_SKIP_ZEROES
michael@0 28 #define CANITER_SKIP_ZEROES TRUE
michael@0 29 #endif
michael@0 30
michael@0 31 U_NAMESPACE_BEGIN
michael@0 32
michael@0 33 class Hashtable;
michael@0 34 class Normalizer2;
michael@0 35 class Normalizer2Impl;
michael@0 36
michael@0 37 /**
michael@0 38 * This class allows one to iterate through all the strings that are canonically equivalent to a given
michael@0 39 * string. For example, here are some sample results:
michael@0 40 Results for: {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
michael@0 41 1: \\u0041\\u030A\\u0064\\u0307\\u0327
michael@0 42 = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
michael@0 43 2: \\u0041\\u030A\\u0064\\u0327\\u0307
michael@0 44 = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
michael@0 45 3: \\u0041\\u030A\\u1E0B\\u0327
michael@0 46 = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
michael@0 47 4: \\u0041\\u030A\\u1E11\\u0307
michael@0 48 = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
michael@0 49 5: \\u00C5\\u0064\\u0307\\u0327
michael@0 50 = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
michael@0 51 6: \\u00C5\\u0064\\u0327\\u0307
michael@0 52 = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
michael@0 53 7: \\u00C5\\u1E0B\\u0327
michael@0 54 = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
michael@0 55 8: \\u00C5\\u1E11\\u0307
michael@0 56 = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
michael@0 57 9: \\u212B\\u0064\\u0307\\u0327
michael@0 58 = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
michael@0 59 10: \\u212B\\u0064\\u0327\\u0307
michael@0 60 = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
michael@0 61 11: \\u212B\\u1E0B\\u0327
michael@0 62 = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
michael@0 63 12: \\u212B\\u1E11\\u0307
michael@0 64 = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
michael@0 65 *<br>Note: the code is intended for use with small strings, and is not suitable for larger ones,
michael@0 66 * since it has not been optimized for that situation.
michael@0 67 * Note, CanonicalIterator is not intended to be subclassed.
michael@0 68 * @author M. Davis
michael@0 69 * @author C++ port by V. Weinstein
michael@0 70 * @stable ICU 2.4
michael@0 71 */
michael@0 72 class U_COMMON_API CanonicalIterator : public UObject {
michael@0 73 public:
michael@0 74 /**
michael@0 75 * Construct a CanonicalIterator object
michael@0 76 * @param source string to get results for
michael@0 77 * @param status Fill-in parameter which receives the status of this operation.
michael@0 78 * @stable ICU 2.4
michael@0 79 */
michael@0 80 CanonicalIterator(const UnicodeString &source, UErrorCode &status);
michael@0 81
michael@0 82 /** Destructor
michael@0 83 * Cleans pieces
michael@0 84 * @stable ICU 2.4
michael@0 85 */
michael@0 86 virtual ~CanonicalIterator();
michael@0 87
michael@0 88 /**
michael@0 89 * Gets the NFD form of the current source we are iterating over.
michael@0 90 * @return gets the source: NOTE: it is the NFD form of source
michael@0 91 * @stable ICU 2.4
michael@0 92 */
michael@0 93 UnicodeString getSource();
michael@0 94
michael@0 95 /**
michael@0 96 * Resets the iterator so that one can start again from the beginning.
michael@0 97 * @stable ICU 2.4
michael@0 98 */
michael@0 99 void reset();
michael@0 100
michael@0 101 /**
michael@0 102 * Get the next canonically equivalent string.
michael@0 103 * <br><b>Warning: The strings are not guaranteed to be in any particular order.</b>
michael@0 104 * @return the next string that is canonically equivalent. A bogus string is returned when
michael@0 105 * the iteration is done.
michael@0 106 * @stable ICU 2.4
michael@0 107 */
michael@0 108 UnicodeString next();
michael@0 109
michael@0 110 /**
michael@0 111 * Set a new source for this iterator. Allows object reuse.
michael@0 112 * @param newSource the source string to iterate against. This allows the same iterator to be used
michael@0 113 * while changing the source string, saving object creation.
michael@0 114 * @param status Fill-in parameter which receives the status of this operation.
michael@0 115 * @stable ICU 2.4
michael@0 116 */
michael@0 117 void setSource(const UnicodeString &newSource, UErrorCode &status);
michael@0 118
michael@0 119 #ifndef U_HIDE_INTERNAL_API
michael@0 120 /**
michael@0 121 * Dumb recursive implementation of permutation.
michael@0 122 * TODO: optimize
michael@0 123 * @param source the string to find permutations for
michael@0 124 * @param skipZeros determine if skip zeros
michael@0 125 * @param result the results in a set.
michael@0 126 * @param status Fill-in parameter which receives the status of this operation.
michael@0 127 * @internal
michael@0 128 */
michael@0 129 static void U_EXPORT2 permute(UnicodeString &source, UBool skipZeros, Hashtable *result, UErrorCode &status);
michael@0 130 #endif /* U_HIDE_INTERNAL_API */
michael@0 131
michael@0 132 /**
michael@0 133 * ICU "poor man's RTTI", returns a UClassID for this class.
michael@0 134 *
michael@0 135 * @stable ICU 2.2
michael@0 136 */
michael@0 137 static UClassID U_EXPORT2 getStaticClassID();
michael@0 138
michael@0 139 /**
michael@0 140 * ICU "poor man's RTTI", returns a UClassID for the actual class.
michael@0 141 *
michael@0 142 * @stable ICU 2.2
michael@0 143 */
michael@0 144 virtual UClassID getDynamicClassID() const;
michael@0 145
michael@0 146 private:
michael@0 147 // ===================== PRIVATES ==============================
michael@0 148 // private default constructor
michael@0 149 CanonicalIterator();
michael@0 150
michael@0 151
michael@0 152 /**
michael@0 153 * Copy constructor. Private for now.
michael@0 154 * @internal
michael@0 155 */
michael@0 156 CanonicalIterator(const CanonicalIterator& other);
michael@0 157
michael@0 158 /**
michael@0 159 * Assignment operator. Private for now.
michael@0 160 * @internal
michael@0 161 */
michael@0 162 CanonicalIterator& operator=(const CanonicalIterator& other);
michael@0 163
michael@0 164 // fields
michael@0 165 UnicodeString source;
michael@0 166 UBool done;
michael@0 167
michael@0 168 // 2 dimensional array holds the pieces of the string with
michael@0 169 // their different canonically equivalent representations
michael@0 170 UnicodeString **pieces;
michael@0 171 int32_t pieces_length;
michael@0 172 int32_t *pieces_lengths;
michael@0 173
michael@0 174 // current is used in iterating to combine pieces
michael@0 175 int32_t *current;
michael@0 176 int32_t current_length;
michael@0 177
michael@0 178 // transient fields
michael@0 179 UnicodeString buffer;
michael@0 180
michael@0 181 const Normalizer2 &nfd;
michael@0 182 const Normalizer2Impl &nfcImpl;
michael@0 183
michael@0 184 // we have a segment, in NFD. Find all the strings that are canonically equivalent to it.
michael@0 185 UnicodeString *getEquivalents(const UnicodeString &segment, int32_t &result_len, UErrorCode &status); //private String[] getEquivalents(String segment)
michael@0 186
michael@0 187 //Set getEquivalents2(String segment);
michael@0 188 Hashtable *getEquivalents2(Hashtable *fillinResult, const UChar *segment, int32_t segLen, UErrorCode &status);
michael@0 189 //Hashtable *getEquivalents2(const UnicodeString &segment, int32_t segLen, UErrorCode &status);
michael@0 190
michael@0 191 /**
michael@0 192 * See if the decomposition of cp2 is at segment starting at segmentPos
michael@0 193 * (with canonical rearrangment!)
michael@0 194 * If so, take the remainder, and return the equivalents
michael@0 195 */
michael@0 196 //Set extract(int comp, String segment, int segmentPos, StringBuffer buffer);
michael@0 197 Hashtable *extract(Hashtable *fillinResult, UChar32 comp, const UChar *segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
michael@0 198 //Hashtable *extract(UChar32 comp, const UnicodeString &segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
michael@0 199
michael@0 200 void cleanPieces();
michael@0 201
michael@0 202 };
michael@0 203
michael@0 204 U_NAMESPACE_END
michael@0 205
michael@0 206 #endif /* #if !UCONFIG_NO_NORMALIZATION */
michael@0 207
michael@0 208 #endif

Mercurial Repository: The Tor Browser

Europalab SCM Repository Server

mercurial