1 /*
2 +----------------------------------------------------------------------+
3 | PHP Version 5 |
4 +----------------------------------------------------------------------+
5 | This source file is subject to version 3.01 of the PHP license, |
6 | that is bundled with this package in the file LICENSE, and is |
7 | available through the world-wide-web at the following url: |
8 | http://www.php.net/license/3_01.txt |
9 | If you did not receive a copy of the PHP license and are unable to |
10 | obtain it through the world-wide-web, please send a note to |
11 | license@php.net so we can mail you a copy immediately. |
12 +----------------------------------------------------------------------+
13 | Authors: Gustavo Lopes <cataphract@php.net> |
14 +----------------------------------------------------------------------+
15 */
16
17 #include "codepointiterator_internal.h"
18 #include <unicode/uchriter.h>
19 #include <typeinfo>
20
21 //copied from cmemory.h, which is not public
22 typedef union {
23 long t1;
24 double t2;
25 void *t3;
26 } UAlignedMemory;
27
28 #define U_POINTER_MASK_LSB(ptr, mask) (((ptrdiff_t)(char *)(ptr)) & (mask))
29 #define U_ALIGNMENT_OFFSET(ptr) U_POINTER_MASK_LSB(ptr, sizeof(UAlignedMemory) - 1)
30 #define U_ALIGNMENT_OFFSET_UP(ptr) (sizeof(UAlignedMemory) - U_ALIGNMENT_OFFSET(ptr))
31
32 using namespace PHP;
33
34 UOBJECT_DEFINE_RTTI_IMPLEMENTATION(CodePointBreakIterator);
35
CodePointBreakIterator()36 CodePointBreakIterator::CodePointBreakIterator()
37 : BreakIterator(), fCharIter(NULL), lastCodePoint(U_SENTINEL)
38 {
39 UErrorCode uec = UErrorCode();
40 this->fText = utext_openUChars(NULL, NULL, 0, &uec);
41 }
42
CodePointBreakIterator(const PHP::CodePointBreakIterator & other)43 CodePointBreakIterator::CodePointBreakIterator(const PHP::CodePointBreakIterator &other)
44 : BreakIterator(other), fText(NULL), fCharIter(NULL), lastCodePoint(U_SENTINEL)
45 {
46 *this = other;
47 }
48
operator =(const CodePointBreakIterator & that)49 CodePointBreakIterator& CodePointBreakIterator::operator=(const CodePointBreakIterator& that)
50 {
51 UErrorCode uec = UErrorCode();
52 UText *ut_clone = NULL;
53
54 if (this == &that) {
55 return *this;
56 }
57
58 this->fText = utext_clone(this->fText, that.fText, FALSE, TRUE, &uec);
59
60 //don't bother copying the character iterator, getText() is deprecated
61 clearCurrentCharIter();
62
63 this->lastCodePoint = that.lastCodePoint;
64 return *this;
65 }
66
~CodePointBreakIterator()67 CodePointBreakIterator::~CodePointBreakIterator()
68 {
69 if (this->fText) {
70 utext_close(this->fText);
71 }
72 clearCurrentCharIter();
73 }
74
operator ==(const BreakIterator & that) const75 UBool CodePointBreakIterator::operator==(const BreakIterator& that) const
76 {
77 if (typeid(*this) != typeid(that)) {
78 return FALSE;
79 }
80
81 const CodePointBreakIterator& that2 =
82 static_cast<const CodePointBreakIterator&>(that);
83
84 if (!utext_equals(this->fText, that2.fText)) {
85 return FALSE;
86 }
87
88 return TRUE;
89 }
90
clone(void) const91 CodePointBreakIterator* CodePointBreakIterator::clone(void) const
92 {
93 return new CodePointBreakIterator(*this);
94 }
95
getText(void) const96 CharacterIterator& CodePointBreakIterator::getText(void) const
97 {
98 if (this->fCharIter == NULL) {
99 //this method is deprecated anyway; setup bogus iterator
100 static const UChar c = 0;
101 this->fCharIter = new UCharCharacterIterator(&c, 0);
102 }
103
104 return *this->fCharIter;
105 }
106
getUText(UText * fillIn,UErrorCode & status) const107 UText *CodePointBreakIterator::getUText(UText *fillIn, UErrorCode &status) const
108 {
109 return utext_clone(fillIn, this->fText, FALSE, TRUE, &status);
110 }
111
setText(const UnicodeString & text)112 void CodePointBreakIterator::setText(const UnicodeString &text)
113 {
114 UErrorCode uec = UErrorCode();
115
116 //this closes the previous utext, if any
117 this->fText = utext_openConstUnicodeString(this->fText, &text, &uec);
118
119 clearCurrentCharIter();
120 }
121
setText(UText * text,UErrorCode & status)122 void CodePointBreakIterator::setText(UText *text, UErrorCode &status)
123 {
124 if (U_FAILURE(status)) {
125 return;
126 }
127
128 this->fText = utext_clone(this->fText, text, FALSE, TRUE, &status);
129
130 clearCurrentCharIter();
131 }
132
adoptText(CharacterIterator * it)133 void CodePointBreakIterator::adoptText(CharacterIterator* it)
134 {
135 UErrorCode uec = UErrorCode();
136 clearCurrentCharIter();
137
138 this->fCharIter = it;
139 this->fText = utext_openCharacterIterator(this->fText, it, &uec);
140 }
141
first(void)142 int32_t CodePointBreakIterator::first(void)
143 {
144 UTEXT_SETNATIVEINDEX(this->fText, 0);
145 this->lastCodePoint = U_SENTINEL;
146
147 return 0;
148 }
149
last(void)150 int32_t CodePointBreakIterator::last(void)
151 {
152 int32_t pos = (int32_t)utext_nativeLength(this->fText);
153 UTEXT_SETNATIVEINDEX(this->fText, pos);
154 this->lastCodePoint = U_SENTINEL;
155
156 return pos;
157 }
158
previous(void)159 int32_t CodePointBreakIterator::previous(void)
160 {
161 this->lastCodePoint = UTEXT_PREVIOUS32(this->fText);
162 if (this->lastCodePoint == U_SENTINEL) {
163 return BreakIterator::DONE;
164 }
165
166 return (int32_t)UTEXT_GETNATIVEINDEX(this->fText);
167 }
168
next(void)169 int32_t CodePointBreakIterator::next(void)
170 {
171 this->lastCodePoint = UTEXT_NEXT32(this->fText);
172 if (this->lastCodePoint == U_SENTINEL) {
173 return BreakIterator::DONE;
174 }
175
176 return (int32_t)UTEXT_GETNATIVEINDEX(this->fText);
177 }
178
current(void) const179 int32_t CodePointBreakIterator::current(void) const
180 {
181 return (int32_t)UTEXT_GETNATIVEINDEX(this->fText);
182 }
183
following(int32_t offset)184 int32_t CodePointBreakIterator::following(int32_t offset)
185 {
186 this->lastCodePoint = utext_next32From(this->fText, offset);
187 if (this->lastCodePoint == U_SENTINEL) {
188 return BreakIterator::DONE;
189 }
190
191 return (int32_t)UTEXT_GETNATIVEINDEX(this->fText);
192 }
193
preceding(int32_t offset)194 int32_t CodePointBreakIterator::preceding(int32_t offset)
195 {
196 this->lastCodePoint = utext_previous32From(this->fText, offset);
197 if (this->lastCodePoint == U_SENTINEL) {
198 return BreakIterator::DONE;
199 }
200
201 return (int32_t)UTEXT_GETNATIVEINDEX(this->fText);
202 }
203
isBoundary(int32_t offset)204 UBool CodePointBreakIterator::isBoundary(int32_t offset)
205 {
206 //this function has side effects, and it's supposed to
207 utext_setNativeIndex(this->fText, offset);
208 return (offset == utext_getNativeIndex(this->fText));
209 }
210
next(int32_t n)211 int32_t CodePointBreakIterator::next(int32_t n)
212 {
213 UBool res = utext_moveIndex32(this->fText, n);
214
215 #ifndef UTEXT_CURRENT32
216 #define UTEXT_CURRENT32 utext_current32
217 #endif
218
219 if (res) {
220 this->lastCodePoint = UTEXT_CURRENT32(this->fText);
221 return (int32_t)UTEXT_GETNATIVEINDEX(this->fText);
222 } else {
223 this->lastCodePoint = U_SENTINEL;
224 return BreakIterator::DONE;
225 }
226 }
227
createBufferClone(void * stackBuffer,int32_t & bufferSize,UErrorCode & status)228 CodePointBreakIterator *CodePointBreakIterator::createBufferClone(
229 void *stackBuffer, int32_t &bufferSize, UErrorCode &status)
230 {
231 //see implementation of RuleBasedBreakIterator::createBufferClone()
232 if (U_FAILURE(status)) {
233 return NULL;
234 }
235
236 if (bufferSize <= 0) {
237 bufferSize = sizeof(CodePointBreakIterator) + U_ALIGNMENT_OFFSET_UP(0);
238 return NULL;
239 }
240
241 char *buf = (char*)stackBuffer;
242 uint32_t s = bufferSize;
243
244 if (stackBuffer == NULL) {
245 s = 0;
246 }
247
248 if (U_ALIGNMENT_OFFSET(stackBuffer) != 0) {
249 uint32_t offsetUp = (uint32_t)U_ALIGNMENT_OFFSET_UP(buf);
250 s -= offsetUp;
251 buf += offsetUp;
252 }
253
254 if (s < sizeof(CodePointBreakIterator)) {
255 CodePointBreakIterator *clonedBI = new CodePointBreakIterator(*this);
256 if (clonedBI == NULL) {
257 status = U_MEMORY_ALLOCATION_ERROR;
258 } else {
259 status = U_SAFECLONE_ALLOCATED_WARNING;
260 }
261
262 return clonedBI;
263 }
264
265 return new(buf) CodePointBreakIterator(*this);
266 }
267
refreshInputText(UText * input,UErrorCode & status)268 CodePointBreakIterator &CodePointBreakIterator::refreshInputText(UText *input, UErrorCode &status)
269 {
270 //see implementation of RuleBasedBreakIterator::createBufferClone()
271 if (U_FAILURE(status)) {
272 return *this;
273 }
274 if (input == NULL) {
275 status = U_ILLEGAL_ARGUMENT_ERROR;
276 return *this;
277 }
278
279 int64_t pos = utext_getNativeIndex(this->fText);
280 this->fText = utext_clone(this->fText, input, FALSE, TRUE, &status);
281 if (U_FAILURE(status)) {
282 return *this;
283 }
284
285 utext_setNativeIndex(this->fText, pos);
286 if (utext_getNativeIndex(fText) != pos) {
287 status = U_ILLEGAL_ARGUMENT_ERROR;
288 }
289
290 return *this;
291 }
292