| 1 | // © 2016 and later: Unicode, Inc. and others. |
| 2 | // License & terms of use: http://www.unicode.org/copyright.html |
| 3 | /* |
| 4 | ******************************************************************************* |
| 5 | * |
| 6 | * Copyright (C) 2001-2011, International Business Machines |
| 7 | * Corporation and others. All Rights Reserved. |
| 8 | * |
| 9 | ******************************************************************************* |
| 10 | * file name: casetrn.cpp |
| 11 | * encoding: UTF-8 |
| 12 | * tab size: 8 (not used) |
| 13 | * indentation:4 |
| 14 | * |
| 15 | * created on: 2004sep03 |
| 16 | * created by: Markus W. Scherer |
| 17 | * |
| 18 | * Implementation class for lower-/upper-/title-casing transliterators. |
| 19 | */ |
| 20 | |
| 21 | #include "unicode/utypes.h" |
| 22 | |
| 23 | #if !UCONFIG_NO_TRANSLITERATION |
| 24 | |
| 25 | #include "unicode/uchar.h" |
| 26 | #include "unicode/ustring.h" |
| 27 | #include "unicode/utf.h" |
| 28 | #include "unicode/utf16.h" |
| 29 | #include "tolowtrn.h" |
| 30 | #include "ucase.h" |
| 31 | #include "cpputils.h" |
| 32 | |
| 33 | /* case context iterator using a Replaceable */ |
| 34 | U_CFUNC UChar32 U_CALLCONV |
| 35 | utrans_rep_caseContextIterator(void *context, int8_t dir) |
| 36 | { |
| 37 | U_NAMESPACE_USE |
| 38 | |
| 39 | UCaseContext *csc=(UCaseContext *)context; |
| 40 | Replaceable *rep=(Replaceable *)csc->p; |
| 41 | UChar32 c; |
| 42 | |
| 43 | if(dir<0) { |
| 44 | /* reset for backward iteration */ |
| 45 | csc->index=csc->cpStart; |
| 46 | csc->dir=dir; |
| 47 | } else if(dir>0) { |
| 48 | /* reset for forward iteration */ |
| 49 | csc->index=csc->cpLimit; |
| 50 | csc->dir=dir; |
| 51 | } else { |
| 52 | /* continue current iteration direction */ |
| 53 | dir=csc->dir; |
| 54 | } |
| 55 | |
| 56 | // automatically adjust start and limit if the Replaceable disagrees |
| 57 | // with the original values |
| 58 | if(dir<0) { |
| 59 | if(csc->start<csc->index) { |
| 60 | c=rep->char32At(csc->index-1); |
| 61 | if(c<0) { |
| 62 | csc->start=csc->index; |
| 63 | } else { |
| 64 | csc->index-=U16_LENGTH(c); |
| 65 | return c; |
| 66 | } |
| 67 | } |
| 68 | } else { |
| 69 | // detect, and store in csc->b1, if we hit the limit |
| 70 | if(csc->index<csc->limit) { |
| 71 | c=rep->char32At(csc->index); |
| 72 | if(c<0) { |
| 73 | csc->limit=csc->index; |
| 74 | csc->b1=TRUE; |
| 75 | } else { |
| 76 | csc->index+=U16_LENGTH(c); |
| 77 | return c; |
| 78 | } |
| 79 | } else { |
| 80 | csc->b1=TRUE; |
| 81 | } |
| 82 | } |
| 83 | return U_SENTINEL; |
| 84 | } |
| 85 | |
| 86 | U_NAMESPACE_BEGIN |
| 87 | |
| 88 | UOBJECT_DEFINE_ABSTRACT_RTTI_IMPLEMENTATION(CaseMapTransliterator) |
| 89 | |
| 90 | /** |
| 91 | * Constructs a transliterator. |
| 92 | */ |
| 93 | CaseMapTransliterator::CaseMapTransliterator(const UnicodeString &id, UCaseMapFull *map) : |
| 94 | Transliterator(id, 0), |
| 95 | fMap(map) |
| 96 | { |
| 97 | // TODO test incremental mode with context-sensitive text (e.g. greek sigma) |
| 98 | // TODO need to call setMaximumContextLength()?! |
| 99 | } |
| 100 | |
| 101 | /** |
| 102 | * Destructor. |
| 103 | */ |
| 104 | CaseMapTransliterator::~CaseMapTransliterator() { |
| 105 | } |
| 106 | |
| 107 | /** |
| 108 | * Copy constructor. |
| 109 | */ |
| 110 | CaseMapTransliterator::CaseMapTransliterator(const CaseMapTransliterator& o) : |
| 111 | Transliterator(o), |
| 112 | fMap(o.fMap) |
| 113 | { |
| 114 | } |
| 115 | |
| 116 | /** |
| 117 | * Assignment operator. |
| 118 | */ |
| 119 | /*CaseMapTransliterator& CaseMapTransliterator::operator=(const CaseMapTransliterator& o) { |
| 120 | Transliterator::operator=(o); |
| 121 | fMap = o.fMap; |
| 122 | return *this; |
| 123 | }*/ |
| 124 | |
| 125 | /** |
| 126 | * Transliterator API. |
| 127 | */ |
| 128 | /*CaseMapTransliterator* CaseMapTransliterator::clone() const { |
| 129 | return new CaseMapTransliterator(*this); |
| 130 | }*/ |
| 131 | |
| 132 | /** |
| 133 | * Implements {@link Transliterator#handleTransliterate}. |
| 134 | */ |
| 135 | void CaseMapTransliterator::handleTransliterate(Replaceable& text, |
| 136 | UTransPosition& offsets, |
| 137 | UBool isIncremental) const |
| 138 | { |
| 139 | if (offsets.start >= offsets.limit) { |
| 140 | return; |
| 141 | } |
| 142 | |
| 143 | UCaseContext csc; |
| 144 | uprv_memset(&csc, 0, sizeof(csc)); |
| 145 | csc.p = &text; |
| 146 | csc.start = offsets.contextStart; |
| 147 | csc.limit = offsets.contextLimit; |
| 148 | |
| 149 | UnicodeString tmp; |
| 150 | const UChar *s; |
| 151 | UChar32 c; |
| 152 | int32_t textPos, delta, result; |
| 153 | |
| 154 | for(textPos=offsets.start; textPos<offsets.limit;) { |
| 155 | csc.cpStart=textPos; |
| 156 | c=text.char32At(textPos); |
| 157 | csc.cpLimit=textPos+=U16_LENGTH(c); |
| 158 | |
| 159 | result=fMap(c, utrans_rep_caseContextIterator, &csc, &s, UCASE_LOC_ROOT); |
| 160 | |
| 161 | if(csc.b1 && isIncremental) { |
| 162 | // fMap() tried to look beyond the context limit |
| 163 | // wait for more input |
| 164 | offsets.start=csc.cpStart; |
| 165 | return; |
| 166 | } |
| 167 | |
| 168 | if(result>=0) { |
| 169 | // replace the current code point with its full case mapping result |
| 170 | // see UCASE_MAX_STRING_LENGTH |
| 171 | if(result<=UCASE_MAX_STRING_LENGTH) { |
| 172 | // string s[result] |
| 173 | tmp.setTo(FALSE, s, result); |
| 174 | delta=result-U16_LENGTH(c); |
| 175 | } else { |
| 176 | // single code point |
| 177 | tmp.setTo(result); |
| 178 | delta=tmp.length()-U16_LENGTH(c); |
| 179 | } |
| 180 | text.handleReplaceBetween(csc.cpStart, textPos, tmp); |
| 181 | if(delta!=0) { |
| 182 | textPos+=delta; |
| 183 | csc.limit=offsets.contextLimit+=delta; |
| 184 | offsets.limit+=delta; |
| 185 | } |
| 186 | } |
| 187 | } |
| 188 | offsets.start=textPos; |
| 189 | } |
| 190 | |
| 191 | U_NAMESPACE_END |
| 192 | |
| 193 | #endif /* #if !UCONFIG_NO_TRANSLITERATION */ |
| 194 | |