Bug Summary

File:out/../deps/icu-small/source/common/unicode/unistr.h
Warning:line 4542, column 74
The result of the left shift is undefined because the left operand is negative

Annotated Source Code

Press '?' to see keyboard shortcuts

clang -cc1 -cc1 -triple x86_64-unknown-linux-gnu -analyze -disable-free -clear-ast-before-backend -disable-llvm-verifier -discard-value-names -main-file-name unistr.cpp -analyzer-checker=core -analyzer-checker=apiModeling -analyzer-checker=unix -analyzer-checker=deadcode -analyzer-checker=cplusplus -analyzer-checker=security.insecureAPI.UncheckedReturn -analyzer-checker=security.insecureAPI.getpw -analyzer-checker=security.insecureAPI.gets -analyzer-checker=security.insecureAPI.mktemp -analyzer-checker=security.insecureAPI.mkstemp -analyzer-checker=security.insecureAPI.vfork -analyzer-checker=nullability.NullPassedToNonnull -analyzer-checker=nullability.NullReturnedFromNonnull -analyzer-output plist -w -setup-static-analyzer -mrelocation-model pic -pic-level 2 -pic-is-pie -mframe-pointer=all -fmath-errno -ffp-contract=on -fno-rounding-math -mconstructor-aliases -funwind-tables=2 -target-cpu x86-64 -tune-cpu generic -debugger-tuning=gdb -fcoverage-compilation-dir=/home/maurizio/node-v18.6.0/out -resource-dir /usr/local/lib/clang/16.0.0 -D V8_DEPRECATION_WARNINGS -D V8_IMMINENT_DEPRECATION_WARNINGS -D _GLIBCXX_USE_CXX11_ABI=1 -D NODE_OPENSSL_CONF_NAME=nodejs_conf -D NODE_OPENSSL_HAS_QUIC -D __STDC_FORMAT_MACROS -D OPENSSL_NO_PINSHARED -D OPENSSL_THREADS -D U_COMMON_IMPLEMENTATION=1 -D U_ATTRIBUTE_DEPRECATED= -D _CRT_SECURE_NO_DEPRECATE= -D U_STATIC_IMPLEMENTATION=1 -D UCONFIG_NO_SERVICE=1 -D U_ENABLE_DYLOAD=0 -D U_HAVE_STD_STRING=1 -D UCONFIG_NO_BREAK_ITERATION=0 -I ../deps/icu-small/source/common -internal-isystem /usr/lib/gcc/x86_64-redhat-linux/8/../../../../include/c++/8 -internal-isystem /usr/lib/gcc/x86_64-redhat-linux/8/../../../../include/c++/8/x86_64-redhat-linux -internal-isystem /usr/lib/gcc/x86_64-redhat-linux/8/../../../../include/c++/8/backward -internal-isystem /usr/local/lib/clang/16.0.0/include -internal-isystem /usr/local/include -internal-isystem /usr/lib/gcc/x86_64-redhat-linux/8/../../../../x86_64-redhat-linux/include -internal-externc-isystem /include -internal-externc-isystem /usr/include -O3 -Wno-unused-parameter -Wno-deprecated-declarations -Wno-strict-aliasing -std=gnu++17 -fdeprecated-macro -fdebug-compilation-dir=/home/maurizio/node-v18.6.0/out -ferror-limit 19 -fgnuc-version=4.2.1 -vectorize-loops -vectorize-slp -analyzer-output=html -faddrsig -D__GCC_HAVE_DWARF2_CFI_ASM=1 -o /tmp/scan-build-2022-08-22-142216-507842-1 -x c++ ../deps/icu-small/source/common/unistr.cpp

../deps/icu-small/source/common/unistr.cpp

1// © 2016 and later: Unicode, Inc. and others.
2// License & terms of use: http://www.unicode.org/copyright.html
3/*
4******************************************************************************
5* Copyright (C) 1999-2016, International Business Machines Corporation and
6* others. All Rights Reserved.
7******************************************************************************
8*
9* File unistr.cpp
10*
11* Modification History:
12*
13* Date Name Description
14* 09/25/98 stephen Creation.
15* 04/20/99 stephen Overhauled per 4/16 code review.
16* 07/09/99 stephen Renamed {hi,lo},{byte,word} to icu_X for HP/UX
17* 11/18/99 aliu Added handleReplaceBetween() to make inherit from
18* Replaceable.
19* 06/25/01 grhoten Removed the dependency on iostream
20******************************************************************************
21*/
22
23#include "unicode/utypes.h"
24#include "unicode/appendable.h"
25#include "unicode/putil.h"
26#include "cstring.h"
27#include "cmemory.h"
28#include "unicode/ustring.h"
29#include "unicode/unistr.h"
30#include "unicode/utf.h"
31#include "unicode/utf16.h"
32#include "uelement.h"
33#include "ustr_imp.h"
34#include "umutex.h"
35#include "uassert.h"
36
37#if 0
38
39#include <iostream>
40using namespace std;
41
42//DEBUGGING
43void
44print(const UnicodeString& s,
45 const char *name)
46{
47 UChar c;
48 cout << name << ":|";
49 for(int i = 0; i < s.length(); ++i) {
50 c = s[i];
51 if(c>= 0x007E || c < 0x0020)
52 cout << "[0x" << hex << s[i] << "]";
53 else
54 cout << (char) s[i];
55 }
56 cout << '|' << endl;
57}
58
59void
60print(const UChar *s,
61 int32_t len,
62 const char *name)
63{
64 UChar c;
65 cout << name << ":|";
66 for(int i = 0; i < len; ++i) {
67 c = s[i];
68 if(c>= 0x007E || c < 0x0020)
69 cout << "[0x" << hex << s[i] << "]";
70 else
71 cout << (char) s[i];
72 }
73 cout << '|' << endl;
74}
75// END DEBUGGING
76#endif
77
78// Local function definitions for now
79
80// need to copy areas that may overlap
81static
82inline void
83us_arrayCopy(const UChar *src, int32_t srcStart,
84 UChar *dst, int32_t dstStart, int32_t count)
85{
86 if(count>0) {
87 uprv_memmove(dst+dstStart, src+srcStart, (size_t)count*sizeof(*src))do { clang diagnostic push clang diagnostic ignored "-Waddress"
(void)0; (void)0; clang diagnostic pop :: memmove(dst+dstStart
, src+srcStart, (size_t)count*sizeof(*src)); } while (false)
;
88 }
89}
90
91// u_unescapeAt() callback to get a UChar from a UnicodeString
92U_CDECL_BEGINextern "C" {
93static UChar U_CALLCONV
94UnicodeString_charAt(int32_t offset, void *context) {
95 return ((icu::UnicodeString*) context)->charAt(offset);
96}
97U_CDECL_END}
98
99U_NAMESPACE_BEGINnamespace icu_71 {
100
101/* The Replaceable virtual destructor can't be defined in the header
102 due to how AIX works with multiple definitions of virtual functions.
103*/
104Replaceable::~Replaceable() {}
105
106UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UnicodeString)UClassID UnicodeString::getStaticClassID() { static char classID
= 0; return (UClassID)&classID; } UClassID UnicodeString
::getDynamicClassID() const { return UnicodeString::getStaticClassID
(); }
107
108UnicodeString U_EXPORT2
109operator+ (const UnicodeString &s1, const UnicodeString &s2) {
110 return
111 UnicodeString(s1.length()+s2.length()+1, (UChar32)0, 0).
112 append(s1).
113 append(s2);
114}
115
116//========================================
117// Reference Counting functions, put at top of file so that optimizing compilers
118// have a chance to automatically inline.
119//========================================
120
121void
122UnicodeString::addRef() {
123 umtx_atomic_inc((u_atomic_int32_t *)fUnion.fFields.fArray - 1);
124}
125
126int32_t
127UnicodeString::removeRef() {
128 return umtx_atomic_dec((u_atomic_int32_t *)fUnion.fFields.fArray - 1);
129}
130
131int32_t
132UnicodeString::refCount() const {
133 return umtx_loadAcquire(*((u_atomic_int32_t *)fUnion.fFields.fArray - 1));
134}
135
136void
137UnicodeString::releaseArray() {
138 if((fUnion.fFields.fLengthAndFlags & kRefCounted) && removeRef() == 0) {
139 uprv_freeuprv_free_71((int32_t *)fUnion.fFields.fArray - 1);
140 }
141}
142
143
144
145//========================================
146// Constructors
147//========================================
148
149// The default constructor is inline in unistr.h.
150
151UnicodeString::UnicodeString(int32_t capacity, UChar32 c, int32_t count) {
152 fUnion.fFields.fLengthAndFlags = 0;
153 if(count <= 0 || (uint32_t)c > 0x10ffff) {
154 // just allocate and do not do anything else
155 allocate(capacity);
156 } else if(c <= 0xffff) {
157 int32_t length = count;
158 if(capacity < length) {
159 capacity = length;
160 }
161 if(allocate(capacity)) {
162 UChar *array = getArrayStart();
163 UChar unit = (UChar)c;
164 for(int32_t i = 0; i < length; ++i) {
165 array[i] = unit;
166 }
167 setLength(length);
168 }
169 } else { // supplementary code point, write surrogate pairs
170 if(count > (INT32_MAX(2147483647) / 2)) {
171 // We would get more than 2G UChars.
172 allocate(capacity);
173 return;
174 }
175 int32_t length = count * 2;
176 if(capacity < length) {
177 capacity = length;
178 }
179 if(allocate(capacity)) {
180 UChar *array = getArrayStart();
181 UChar lead = U16_LEAD(c)(UChar)(((c)>>10)+0xd7c0);
182 UChar trail = U16_TRAIL(c)(UChar)(((c)&0x3ff)|0xdc00);
183 for(int32_t i = 0; i < length; i += 2) {
184 array[i] = lead;
185 array[i + 1] = trail;
186 }
187 setLength(length);
188 }
189 }
190}
191
192UnicodeString::UnicodeString(UChar ch) {
193 fUnion.fFields.fLengthAndFlags = kLength1 | kShortString;
194 fUnion.fStackFields.fBuffer[0] = ch;
195}
196
197UnicodeString::UnicodeString(UChar32 ch) {
198 fUnion.fFields.fLengthAndFlags = kShortString;
199 int32_t i = 0;
200 UBool isError = FALSE0;
201 U16_APPEND(fUnion.fStackFields.fBuffer, i, US_STACKBUF_SIZE, ch, isError)do { if((uint32_t)(ch)<=0xffff) { (fUnion.fStackFields.fBuffer
)[(i)++]=(uint16_t)(ch); } else if((uint32_t)(ch)<=0x10ffff
&& (i)+1<(US_STACKBUF_SIZE)) { (fUnion.fStackFields
.fBuffer)[(i)++]=(uint16_t)(((ch)>>10)+0xd7c0); (fUnion
.fStackFields.fBuffer)[(i)++]=(uint16_t)(((ch)&0x3ff)|0xdc00
); } else { (isError)=true; } } while (false)
;
202 // We test isError so that the compiler does not complain that we don't.
203 // If isError then i==0 which is what we want anyway.
204 if(!isError) {
205 setShortLength(i);
206 }
207}
208
209UnicodeString::UnicodeString(const UChar *text) {
210 fUnion.fFields.fLengthAndFlags = kShortString;
211 doAppend(text, 0, -1);
212}
213
214UnicodeString::UnicodeString(const UChar *text,
215 int32_t textLength) {
216 fUnion.fFields.fLengthAndFlags = kShortString;
217 doAppend(text, 0, textLength);
218}
219
220UnicodeString::UnicodeString(UBool isTerminated,
221 ConstChar16Ptr textPtr,
222 int32_t textLength) {
223 fUnion.fFields.fLengthAndFlags = kReadonlyAlias;
224 const UChar *text = textPtr;
225 if(text == NULL__null) {
226 // treat as an empty string, do not alias
227 setToEmpty();
228 } else if(textLength < -1 ||
229 (textLength == -1 && !isTerminated) ||
230 (textLength >= 0 && isTerminated && text[textLength] != 0)
231 ) {
232 setToBogus();
233 } else {
234 if(textLength == -1) {
235 // text is terminated, or else it would have failed the above test
236 textLength = u_strlenu_strlen_71(text);
237 }
238 setArray(const_cast<UChar *>(text), textLength,
239 isTerminated ? textLength + 1 : textLength);
240 }
241}
242
243UnicodeString::UnicodeString(UChar *buff,
244 int32_t buffLength,
245 int32_t buffCapacity) {
246 fUnion.fFields.fLengthAndFlags = kWritableAlias;
247 if(buff == NULL__null) {
248 // treat as an empty string, do not alias
249 setToEmpty();
250 } else if(buffLength < -1 || buffCapacity < 0 || buffLength > buffCapacity) {
251 setToBogus();
252 } else {
253 if(buffLength == -1) {
254 // fLength = u_strlen(buff); but do not look beyond buffCapacity
255 const UChar *p = buff, *limit = buff + buffCapacity;
256 while(p != limit && *p != 0) {
257 ++p;
258 }
259 buffLength = (int32_t)(p - buff);
260 }
261 setArray(buff, buffLength, buffCapacity);
262 }
263}
264
265UnicodeString::UnicodeString(const char *src, int32_t length, EInvariant) {
266 fUnion.fFields.fLengthAndFlags = kShortString;
267 if(src==NULL__null) {
268 // treat as an empty string
269 } else {
270 if(length<0) {
271 length=(int32_t)uprv_strlen(src):: strlen(src);
272 }
273 if(cloneArrayIfNeeded(length, length, FALSE0)) {
274 u_charsToUCharsu_charsToUChars_71(src, getArrayStart(), length);
275 setLength(length);
276 } else {
277 setToBogus();
278 }
279 }
280}
281
282#if U_CHARSET_IS_UTF81
283
284UnicodeString::UnicodeString(const char *codepageData) {
285 fUnion.fFields.fLengthAndFlags = kShortString;
286 if(codepageData != 0) {
287 setToUTF8(codepageData);
288 }
289}
290
291UnicodeString::UnicodeString(const char *codepageData, int32_t dataLength) {
292 fUnion.fFields.fLengthAndFlags = kShortString;
293 // if there's nothing to convert, do nothing
294 if(codepageData == 0 || dataLength == 0 || dataLength < -1) {
295 return;
296 }
297 if(dataLength == -1) {
298 dataLength = (int32_t)uprv_strlen(codepageData):: strlen(codepageData);
299 }
300 setToUTF8(StringPiece(codepageData, dataLength));
301}
302
303// else see unistr_cnv.cpp
304#endif
305
306UnicodeString::UnicodeString(const UnicodeString& that) {
307 fUnion.fFields.fLengthAndFlags = kShortString;
308 copyFrom(that);
309}
310
311UnicodeString::UnicodeString(UnicodeString &&src) U_NOEXCEPTnoexcept {
312 copyFieldsFrom(src, TRUE1);
313}
314
315UnicodeString::UnicodeString(const UnicodeString& that,
316 int32_t srcStart) {
317 fUnion.fFields.fLengthAndFlags = kShortString;
318 setTo(that, srcStart);
319}
320
321UnicodeString::UnicodeString(const UnicodeString& that,
322 int32_t srcStart,
323 int32_t srcLength) {
324 fUnion.fFields.fLengthAndFlags = kShortString;
325 setTo(that, srcStart, srcLength);
326}
327
328// Replaceable base class clone() default implementation, does not clone
329Replaceable *
330Replaceable::clone() const {
331 return NULL__null;
332}
333
334// UnicodeString overrides clone() with a real implementation
335UnicodeString *
336UnicodeString::clone() const {
337 LocalPointer<UnicodeString> clonedString(new UnicodeString(*this));
338 return clonedString.isValid() && !clonedString->isBogus() ? clonedString.orphan() : nullptr;
339}
340
341//========================================
342// array allocation
343//========================================
344
345namespace {
346
347const int32_t kGrowSize = 128;
348
349// The number of bytes for one int32_t reference counter and capacity UChars
350// must fit into a 32-bit size_t (at least when on a 32-bit platform).
351// We also add one for the NUL terminator, to avoid reallocation in getTerminatedBuffer(),
352// and round up to a multiple of 16 bytes.
353// This means that capacity must be at most (0xfffffff0 - 4) / 2 - 1 = 0x7ffffff5.
354// (With more complicated checks we could go up to 0x7ffffffd without rounding up,
355// but that does not seem worth it.)
356const int32_t kMaxCapacity = 0x7ffffff5;
357
358int32_t getGrowCapacity(int32_t newLength) {
359 int32_t growSize = (newLength >> 2) + kGrowSize;
360 if(growSize <= (kMaxCapacity - newLength)) {
361 return newLength + growSize;
362 } else {
363 return kMaxCapacity;
364 }
365}
366
367} // namespace
368
369UBool
370UnicodeString::allocate(int32_t capacity) {
371 if(capacity <= US_STACKBUF_SIZE) {
372 fUnion.fFields.fLengthAndFlags = kShortString;
373 return TRUE1;
374 }
375 if(capacity <= kMaxCapacity) {
376 ++capacity; // for the NUL
377 // Switch to size_t which is unsigned so that we can allocate up to 4GB.
378 // Reference counter + UChars.
379 size_t numBytes = sizeof(int32_t) + (size_t)capacity * U_SIZEOF_UCHAR2;
380 // Round up to a multiple of 16.
381 numBytes = (numBytes + 15) & ~15;
382 int32_t *array = (int32_t *) uprv_mallocuprv_malloc_71(numBytes);
383 if(array != NULL__null) {
384 // set initial refCount and point behind the refCount
385 *array++ = 1;
386 numBytes -= sizeof(int32_t);
387
388 // have fArray point to the first UChar
389 fUnion.fFields.fArray = (UChar *)array;
390 fUnion.fFields.fCapacity = (int32_t)(numBytes / U_SIZEOF_UCHAR2);
391 fUnion.fFields.fLengthAndFlags = kLongString;
392 return TRUE1;
393 }
394 }
395 fUnion.fFields.fLengthAndFlags = kIsBogus;
396 fUnion.fFields.fArray = 0;
397 fUnion.fFields.fCapacity = 0;
398 return FALSE0;
399}
400
401//========================================
402// Destructor
403//========================================
404
405#ifdef UNISTR_COUNT_FINAL_STRING_LENGTHS
406static u_atomic_int32_t finalLengthCounts[0x400]; // UnicodeString::kMaxShortLength+1
407static u_atomic_int32_t beyondCount(0);
408
409U_CAPIextern "C" void unistr_printLengths() {
410 int32_t i;
411 for(i = 0; i <= 59; ++i) {
412 printf("%2d, %9d\n", i, (int32_t)finalLengthCounts[i]);
413 }
414 int32_t beyond = beyondCount;
415 for(; i < UPRV_LENGTHOF(finalLengthCounts)(int32_t)(sizeof(finalLengthCounts)/sizeof((finalLengthCounts
)[0]))
; ++i) {
416 beyond += finalLengthCounts[i];
417 }
418 printf(">59, %9d\n", beyond);
419}
420#endif
421
422UnicodeString::~UnicodeString()
423{
424#ifdef UNISTR_COUNT_FINAL_STRING_LENGTHS
425 // Count lengths of strings at the end of their lifetime.
426 // Useful for discussion of a desirable stack buffer size.
427 // Count the contents length, not the optional NUL terminator nor further capacity.
428 // Ignore open-buffer strings and strings which alias external storage.
429 if((fUnion.fFields.fLengthAndFlags&(kOpenGetBuffer|kReadonlyAlias|kWritableAlias)) == 0) {
430 if(hasShortLength()) {
431 umtx_atomic_inc(finalLengthCounts + getShortLength());
432 } else {
433 umtx_atomic_inc(&beyondCount);
434 }
435 }
436#endif
437
438 releaseArray();
439}
440
441//========================================
442// Factory methods
443//========================================
444
445UnicodeString UnicodeString::fromUTF8(StringPiece utf8) {
446 UnicodeString result;
447 result.setToUTF8(utf8);
448 return result;
449}
450
451UnicodeString UnicodeString::fromUTF32(const UChar32 *utf32, int32_t length) {
452 UnicodeString result;
453 int32_t capacity;
454 // Most UTF-32 strings will be BMP-only and result in a same-length
455 // UTF-16 string. We overestimate the capacity just slightly,
456 // just in case there are a few supplementary characters.
457 if(length <= US_STACKBUF_SIZE) {
458 capacity = US_STACKBUF_SIZE;
459 } else {
460 capacity = length + (length >> 4) + 4;
461 }
462 do {
463 UChar *utf16 = result.getBuffer(capacity);
464 int32_t length16;
465 UErrorCode errorCode = U_ZERO_ERROR;
466 u_strFromUTF32WithSubu_strFromUTF32WithSub_71(utf16, result.getCapacity(), &length16,
467 utf32, length,
468 0xfffd, // Substitution character.
469 NULL__null, // Don't care about number of substitutions.
470 &errorCode);
471 result.releaseBuffer(length16);
472 if(errorCode == U_BUFFER_OVERFLOW_ERROR) {
473 capacity = length16 + 1; // +1 for the terminating NUL.
474 continue;
475 } else if(U_FAILURE(errorCode)) {
476 result.setToBogus();
477 }
478 break;
479 } while(TRUE1);
480 return result;
481}
482
483//========================================
484// Assignment
485//========================================
486
487UnicodeString &
488UnicodeString::operator=(const UnicodeString &src) {
489 return copyFrom(src);
490}
491
492UnicodeString &
493UnicodeString::fastCopyFrom(const UnicodeString &src) {
494 return copyFrom(src, TRUE1);
495}
496
497UnicodeString &
498UnicodeString::copyFrom(const UnicodeString &src, UBool fastCopy) {
499 // if assigning to ourselves, do nothing
500 if(this == &src) {
501 return *this;
502 }
503
504 // is the right side bogus?
505 if(src.isBogus()) {
506 setToBogus();
507 return *this;
508 }
509
510 // delete the current contents
511 releaseArray();
512
513 if(src.isEmpty()) {
514 // empty string - use the stack buffer
515 setToEmpty();
516 return *this;
517 }
518
519 // fLength>0 and not an "open" src.getBuffer(minCapacity)
520 fUnion.fFields.fLengthAndFlags = src.fUnion.fFields.fLengthAndFlags;
521 switch(src.fUnion.fFields.fLengthAndFlags & kAllStorageFlags) {
522 case kShortString:
523 // short string using the stack buffer, do the same
524 uprv_memcpy(fUnion.fStackFields.fBuffer, src.fUnion.fStackFields.fBuffer,do { clang diagnostic push clang diagnostic ignored "-Waddress"
(void)0; (void)0; clang diagnostic pop :: memcpy(fUnion.fStackFields
.fBuffer, src.fUnion.fStackFields.fBuffer, getShortLength() *
2); } while (false)
525 getShortLength() * U_SIZEOF_UCHAR)do { clang diagnostic push clang diagnostic ignored "-Waddress"
(void)0; (void)0; clang diagnostic pop :: memcpy(fUnion.fStackFields
.fBuffer, src.fUnion.fStackFields.fBuffer, getShortLength() *
2); } while (false)
;
526 break;
527 case kLongString:
528 // src uses a refCounted string buffer, use that buffer with refCount
529 // src is const, use a cast - we don't actually change it
530 ((UnicodeString &)src).addRef();
531 // copy all fields, share the reference-counted buffer
532 fUnion.fFields.fArray = src.fUnion.fFields.fArray;
533 fUnion.fFields.fCapacity = src.fUnion.fFields.fCapacity;
534 if(!hasShortLength()) {
535 fUnion.fFields.fLength = src.fUnion.fFields.fLength;
536 }
537 break;
538 case kReadonlyAlias:
539 if(fastCopy) {
540 // src is a readonly alias, do the same
541 // -> maintain the readonly alias as such
542 fUnion.fFields.fArray = src.fUnion.fFields.fArray;
543 fUnion.fFields.fCapacity = src.fUnion.fFields.fCapacity;
544 if(!hasShortLength()) {
545 fUnion.fFields.fLength = src.fUnion.fFields.fLength;
546 }
547 break;
548 }
549 // else if(!fastCopy) fall through to case kWritableAlias
550 // -> allocate a new buffer and copy the contents
551 U_FALLTHROUGH[[clang::fallthrough]];
552 case kWritableAlias: {
553 // src is a writable alias; we make a copy of that instead
554 int32_t srcLength = src.length();
555 if(allocate(srcLength)) {
556 u_memcpyu_memcpy_71(getArrayStart(), src.getArrayStart(), srcLength);
557 setLength(srcLength);
558 break;
559 }
560 // if there is not enough memory, then fall through to setting to bogus
561 U_FALLTHROUGH[[clang::fallthrough]];
562 }
563 default:
564 // if src is bogus, set ourselves to bogus
565 // do not call setToBogus() here because fArray and flags are not consistent here
566 fUnion.fFields.fLengthAndFlags = kIsBogus;
567 fUnion.fFields.fArray = 0;
568 fUnion.fFields.fCapacity = 0;
569 break;
570 }
571
572 return *this;
573}
574
575UnicodeString &UnicodeString::operator=(UnicodeString &&src) U_NOEXCEPTnoexcept {
576 // No explicit check for self move assignment, consistent with standard library.
577 // Self move assignment causes no crash nor leak but might make the object bogus.
578 releaseArray();
579 copyFieldsFrom(src, TRUE1);
580 return *this;
581}
582
583// Same as move assignment except without memory management.
584void UnicodeString::copyFieldsFrom(UnicodeString &src, UBool setSrcToBogus) U_NOEXCEPTnoexcept {
585 int16_t lengthAndFlags = fUnion.fFields.fLengthAndFlags = src.fUnion.fFields.fLengthAndFlags;
586 if(lengthAndFlags & kUsingStackBuffer) {
587 // Short string using the stack buffer, copy the contents.
588 // Check for self assignment to prevent "overlap in memcpy" warnings,
589 // although it should be harmless to copy a buffer to itself exactly.
590 if(this != &src) {
591 uprv_memcpy(fUnion.fStackFields.fBuffer, src.fUnion.fStackFields.fBuffer,do { clang diagnostic push clang diagnostic ignored "-Waddress"
(void)0; (void)0; clang diagnostic pop :: memcpy(fUnion.fStackFields
.fBuffer, src.fUnion.fStackFields.fBuffer, getShortLength() *
2); } while (false)
592 getShortLength() * U_SIZEOF_UCHAR)do { clang diagnostic push clang diagnostic ignored "-Waddress"
(void)0; (void)0; clang diagnostic pop :: memcpy(fUnion.fStackFields
.fBuffer, src.fUnion.fStackFields.fBuffer, getShortLength() *
2); } while (false)
;
593 }
594 } else {
595 // In all other cases, copy all fields.
596 fUnion.fFields.fArray = src.fUnion.fFields.fArray;
597 fUnion.fFields.fCapacity = src.fUnion.fFields.fCapacity;
598 if(!hasShortLength()) {
599 fUnion.fFields.fLength = src.fUnion.fFields.fLength;
600 }
601 if(setSrcToBogus) {
602 // Set src to bogus without releasing any memory.
603 src.fUnion.fFields.fLengthAndFlags = kIsBogus;
604 src.fUnion.fFields.fArray = NULL__null;
605 src.fUnion.fFields.fCapacity = 0;
606 }
607 }
608}
609
610void UnicodeString::swap(UnicodeString &other) U_NOEXCEPTnoexcept {
611 UnicodeString temp; // Empty short string: Known not to need releaseArray().
612 // Copy fields without resetting source values in between.
613 temp.copyFieldsFrom(*this, FALSE0);
614 this->copyFieldsFrom(other, FALSE0);
615 other.copyFieldsFrom(temp, FALSE0);
616 // Set temp to an empty string so that other's memory is not released twice.
617 temp.fUnion.fFields.fLengthAndFlags = kShortString;
618}
619
620//========================================
621// Miscellaneous operations
622//========================================
623
624UnicodeString UnicodeString::unescape() const {
625 UnicodeString result(length(), (UChar32)0, (int32_t)0); // construct with capacity
626 if (result.isBogus()) {
627 return result;
628 }
629 const UChar *array = getBuffer();
630 int32_t len = length();
631 int32_t prev = 0;
632 for (int32_t i=0;;) {
633 if (i == len) {
634 result.append(array, prev, len - prev);
635 break;
636 }
637 if (array[i++] == 0x5C /*'\\'*/) {
638 result.append(array, prev, (i - 1) - prev);
639 UChar32 c = unescapeAt(i); // advances i
640 if (c < 0) {
641 result.remove(); // return empty string
642 break; // invalid escape sequence
643 }
644 result.append(c);
645 prev = i;
646 }
647 }
648 return result;
649}
650
651UChar32 UnicodeString::unescapeAt(int32_t &offset) const {
652 return u_unescapeAtu_unescapeAt_71(UnicodeString_charAt, &offset, length(), (void*)this);
653}
654
655//========================================
656// Read-only implementation
657//========================================
658UBool
659UnicodeString::doEquals(const UnicodeString &text, int32_t len) const {
660 // Requires: this & text not bogus and have same lengths.
661 // Byte-wise comparison works for equality regardless of endianness.
662 return uprv_memcmp(getArrayStart(), text.getArrayStart(), len * U_SIZEOF_UCHAR):: memcmp(getArrayStart(), text.getArrayStart(),len * 2) == 0;
663}
664
665int8_t
666UnicodeString::doCompare( int32_t start,
667 int32_t length,
668 const UChar *srcChars,
669 int32_t srcStart,
670 int32_t srcLength) const
671{
672 // compare illegal string values
673 if(isBogus()) {
674 return -1;
675 }
676
677 // pin indices to legal values
678 pinIndices(start, length);
679
680 if(srcChars == NULL__null) {
681 // treat const UChar *srcChars==NULL as an empty string
682 return length == 0 ? 0 : 1;
683 }
684
685 // get the correct pointer
686 const UChar *chars = getArrayStart();
687
688 chars += start;
689 srcChars += srcStart;
690
691 int32_t minLength;
692 int8_t lengthResult;
693
694 // get the srcLength if necessary
695 if(srcLength < 0) {
696 srcLength = u_strlenu_strlen_71(srcChars + srcStart);
697 }
698
699 // are we comparing different lengths?
700 if(length != srcLength) {
701 if(length < srcLength) {
702 minLength = length;
703 lengthResult = -1;
704 } else {
705 minLength = srcLength;
706 lengthResult = 1;
707 }
708 } else {
709 minLength = length;
710 lengthResult = 0;
711 }
712
713 /*
714 * note that uprv_memcmp() returns an int but we return an int8_t;
715 * we need to take care not to truncate the result -
716 * one way to do this is to right-shift the value to
717 * move the sign bit into the lower 8 bits and making sure that this
718 * does not become 0 itself
719 */
720
721 if(minLength > 0 && chars != srcChars) {
722 int32_t result;
723
724# if U_IS_BIG_ENDIAN(1234 == 4321)
725 // big-endian: byte comparison works
726 result = uprv_memcmp(chars, srcChars, minLength * sizeof(UChar)):: memcmp(chars, srcChars,minLength * sizeof(UChar));
727 if(result != 0) {
728 return (int8_t)(result >> 15 | 1);
729 }
730# else
731 // little-endian: compare UChar units
732 do {
733 result = ((int32_t)*(chars++) - (int32_t)*(srcChars++));
734 if(result != 0) {
735 return (int8_t)(result >> 15 | 1);
736 }
737 } while(--minLength > 0);
738# endif
739 }
740 return lengthResult;
741}
742
743/* String compare in code point order - doCompare() compares in code unit order. */
744int8_t
745UnicodeString::doCompareCodePointOrder(int32_t start,
746 int32_t length,
747 const UChar *srcChars,
748 int32_t srcStart,
749 int32_t srcLength) const
750{
751 // compare illegal string values
752 // treat const UChar *srcChars==NULL as an empty string
753 if(isBogus()) {
754 return -1;
755 }
756
757 // pin indices to legal values
758 pinIndices(start, length);
759
760 if(srcChars == NULL__null) {
761 srcStart = srcLength = 0;
762 }
763
764 int32_t diff = uprv_strCompareuprv_strCompare_71(getArrayStart() + start, length, (srcChars!=NULL__null)?(srcChars + srcStart):NULL__null, srcLength, FALSE0, TRUE1);
765 /* translate the 32-bit result into an 8-bit one */
766 if(diff!=0) {
767 return (int8_t)(diff >> 15 | 1);
768 } else {
769 return 0;
770 }
771}
772
773int32_t
774UnicodeString::getLength() const {
775 return length();
776}
777
778UChar
779UnicodeString::getCharAt(int32_t offset) const {
780 return charAt(offset);
781}
782
783UChar32
784UnicodeString::getChar32At(int32_t offset) const {
785 return char32At(offset);
786}
787
788UChar32
789UnicodeString::char32At(int32_t offset) const
790{
791 int32_t len = length();
792 if((uint32_t)offset < (uint32_t)len) {
793 const UChar *array = getArrayStart();
794 UChar32 c;
795 U16_GET(array, 0, offset, len, c)do { (c)=(array)[offset]; if((((c)&0xfffff800)==0xd800)) {
uint16_t __c2; if((((c)&0x400)==0)) { if((offset)+1!=(len
) && (((__c2=(array)[(offset)+1])&0xfffffc00)==0xdc00
)) { (c)=(((UChar32)((c))<<10UL)+(UChar32)(__c2)-((0xd800
<<10UL)+0xdc00-0x10000)); } } else { if((offset)>(0)
&& (((__c2=(array)[(offset)-1])&0xfffffc00)==0xd800
)) { (c)=(((UChar32)(__c2)<<10UL)+(UChar32)((c))-((0xd800
<<10UL)+0xdc00-0x10000)); } } } } while (false)
;
796 return c;
797 } else {
798 return kInvalidUChar;
799 }
800}
801
802int32_t
803UnicodeString::getChar32Start(int32_t offset) const {
804 if((uint32_t)offset < (uint32_t)length()) {
805 const UChar *array = getArrayStart();
806 U16_SET_CP_START(array, 0, offset)do { if(((((array)[offset])&0xfffffc00)==0xdc00) &&
(offset)>(0) && ((((array)[(offset)-1])&0xfffffc00
)==0xd800)) { --(offset); } } while (false)
;
807 return offset;
808 } else {
809 return 0;
810 }
811}
812
813int32_t
814UnicodeString::getChar32Limit(int32_t offset) const {
815 int32_t len = length();
816 if((uint32_t)offset < (uint32_t)len) {
817 const UChar *array = getArrayStart();
818 U16_SET_CP_LIMIT(array, 0, offset, len)do { if((0)<(offset) && ((offset)<(len) || (len
)<0) && ((((array)[(offset)-1])&0xfffffc00)==0xd800
) && ((((array)[offset])&0xfffffc00)==0xdc00)) { ++
(offset); } } while (false)
;
819 return offset;
820 } else {
821 return len;
822 }
823}
824
825int32_t
826UnicodeString::countChar32(int32_t start, int32_t length) const {
827 pinIndices(start, length);
828 // if(isBogus()) then fArray==0 and start==0 - u_countChar32() checks for NULL
829 return u_countChar32u_countChar32_71(getArrayStart()+start, length);
830}
831
832UBool
833UnicodeString::hasMoreChar32Than(int32_t start, int32_t length, int32_t number) const {
834 pinIndices(start, length);
835 // if(isBogus()) then fArray==0 and start==0 - u_strHasMoreChar32Than() checks for NULL
836 return u_strHasMoreChar32Thanu_strHasMoreChar32Than_71(getArrayStart()+start, length, number);
837}
838
839int32_t
840UnicodeString::moveIndex32(int32_t index, int32_t delta) const {
841 // pin index
842 int32_t len = length();
843 if(index<0) {
844 index=0;
845 } else if(index>len) {
846 index=len;
847 }
848
849 const UChar *array = getArrayStart();
850 if(delta>0) {
851 U16_FWD_N(array, index, len, delta)do { int32_t __N=(delta); while(__N>0 && ((index)<
(len) || ((len)<0 && (array)[index]!=0))) { do { if
(((((array)[(index)++])&0xfffffc00)==0xd800) && (
index)!=(len) && ((((array)[index])&0xfffffc00)==
0xdc00)) { ++(index); } } while (false); --__N; } } while (false
)
;
852 } else {
853 U16_BACK_N(array, 0, index, -delta)do { int32_t __N=(-delta); while(__N>0 && (index)>
(0)) { do { if(((((array)[--(index)])&0xfffffc00)==0xdc00
) && (index)>(0) && ((((array)[(index)-1])
&0xfffffc00)==0xd800)) { --(index); } } while (false); --
__N; } } while (false)
;
854 }
855
856 return index;
857}
858
859void
860UnicodeString::doExtract(int32_t start,
861 int32_t length,
862 UChar *dst,
863 int32_t dstStart) const
864{
865 // pin indices to legal values
866 pinIndices(start, length);
867
868 // do not copy anything if we alias dst itself
869 const UChar *array = getArrayStart();
870 if(array + start != dst + dstStart) {
871 us_arrayCopy(array, start, dst, dstStart, length);
872 }
873}
874
875int32_t
876UnicodeString::extract(Char16Ptr dest, int32_t destCapacity,
877 UErrorCode &errorCode) const {
878 int32_t len = length();
879 if(U_SUCCESS(errorCode)) {
880 if(isBogus() || destCapacity<0 || (destCapacity>0 && dest==0)) {
881 errorCode=U_ILLEGAL_ARGUMENT_ERROR;
882 } else {
883 const UChar *array = getArrayStart();
884 if(len>0 && len<=destCapacity && array!=dest) {
885 u_memcpyu_memcpy_71(dest, array, len);
886 }
887 return u_terminateUCharsu_terminateUChars_71(dest, destCapacity, len, &errorCode);
888 }
889 }
890
891 return len;
892}
893
894int32_t
895UnicodeString::extract(int32_t start,
896 int32_t length,
897 char *target,
898 int32_t targetCapacity,
899 enum EInvariant) const
900{
901 // if the arguments are illegal, then do nothing
902 if(targetCapacity < 0 || (targetCapacity > 0 && target == NULL__null)) {
903 return 0;
904 }
905
906 // pin the indices to legal values
907 pinIndices(start, length);
908
909 if(length <= targetCapacity) {
910 u_UCharsToCharsu_UCharsToChars_71(getArrayStart() + start, target, length);
911 }
912 UErrorCode status = U_ZERO_ERROR;
913 return u_terminateCharsu_terminateChars_71(target, targetCapacity, length, &status);
914}
915
916UnicodeString
917UnicodeString::tempSubString(int32_t start, int32_t len) const {
918 pinIndices(start, len);
919 const UChar *array = getBuffer(); // not getArrayStart() to check kIsBogus & kOpenGetBuffer
920 if(array==NULL__null) {
921 array=fUnion.fStackFields.fBuffer; // anything not NULL because that would make an empty string
922 len=-2; // bogus result string
923 }
924 return UnicodeString(FALSE0, array + start, len);
925}
926
927int32_t
928UnicodeString::toUTF8(int32_t start, int32_t len,
929 char *target, int32_t capacity) const {
930 pinIndices(start, len);
931 int32_t length8;
932 UErrorCode errorCode = U_ZERO_ERROR;
933 u_strToUTF8WithSubu_strToUTF8WithSub_71(target, capacity, &length8,
934 getBuffer() + start, len,
935 0xFFFD, // Standard substitution character.
936 NULL__null, // Don't care about number of substitutions.
937 &errorCode);
938 return length8;
939}
940
941#if U_CHARSET_IS_UTF81
942
943int32_t
944UnicodeString::extract(int32_t start, int32_t len,
945 char *target, uint32_t dstSize) const {
946 // if the arguments are illegal, then do nothing
947 if(/*dstSize < 0 || */(dstSize > 0 && target == 0)) {
948 return 0;
949 }
950 return toUTF8(start, len, target, dstSize <= 0x7fffffff ? (int32_t)dstSize : 0x7fffffff);
951}
952
953// else see unistr_cnv.cpp
954#endif
955
956void
957UnicodeString::extractBetween(int32_t start,
958 int32_t limit,
959 UnicodeString& target) const {
960 pinIndex(start);
961 pinIndex(limit);
962 doExtract(start, limit - start, target);
963}
964
965// When converting from UTF-16 to UTF-8, the result will have at most 3 times
966// as many bytes as the source has UChars.
967// The "worst cases" are writing systems like Indic, Thai and CJK with
968// 3:1 bytes:UChars.
969void
970UnicodeString::toUTF8(ByteSink &sink) const {
971 int32_t length16 = length();
972 if(length16 != 0) {
973 char stackBuffer[1024];
974 int32_t capacity = (int32_t)sizeof(stackBuffer);
975 UBool utf8IsOwned = FALSE0;
976 char *utf8 = sink.GetAppendBuffer(length16 < capacity ? length16 : capacity,
977 3*length16,
978 stackBuffer, capacity,
979 &capacity);
980 int32_t length8 = 0;
981 UErrorCode errorCode = U_ZERO_ERROR;
982 u_strToUTF8WithSubu_strToUTF8WithSub_71(utf8, capacity, &length8,
983 getBuffer(), length16,
984 0xFFFD, // Standard substitution character.
985 NULL__null, // Don't care about number of substitutions.
986 &errorCode);
987 if(errorCode == U_BUFFER_OVERFLOW_ERROR) {
988 utf8 = (char *)uprv_mallocuprv_malloc_71(length8);
989 if(utf8 != NULL__null) {
990 utf8IsOwned = TRUE1;
991 errorCode = U_ZERO_ERROR;
992 u_strToUTF8WithSubu_strToUTF8WithSub_71(utf8, length8, &length8,
993 getBuffer(), length16,
994 0xFFFD, // Standard substitution character.
995 NULL__null, // Don't care about number of substitutions.
996 &errorCode);
997 } else {
998 errorCode = U_MEMORY_ALLOCATION_ERROR;
999 }
1000 }
1001 if(U_SUCCESS(errorCode)) {
1002 sink.Append(utf8, length8);
1003 sink.Flush();
1004 }
1005 if(utf8IsOwned) {
1006 uprv_freeuprv_free_71(utf8);
1007 }
1008 }
1009}
1010
1011int32_t
1012UnicodeString::toUTF32(UChar32 *utf32, int32_t capacity, UErrorCode &errorCode) const {
1013 int32_t length32=0;
1014 if(U_SUCCESS(errorCode)) {
1015 // getBuffer() and u_strToUTF32WithSub() check for illegal arguments.
1016 u_strToUTF32WithSubu_strToUTF32WithSub_71(utf32, capacity, &length32,
1017 getBuffer(), length(),
1018 0xfffd, // Substitution character.
1019 NULL__null, // Don't care about number of substitutions.
1020 &errorCode);
1021 }
1022 return length32;
1023}
1024
1025int32_t
1026UnicodeString::indexOf(const UChar *srcChars,
1027 int32_t srcStart,
1028 int32_t srcLength,
1029 int32_t start,
1030 int32_t length) const
1031{
1032 if(isBogus() || srcChars == 0 || srcStart < 0 || srcLength == 0) {
1033 return -1;
1034 }
1035
1036 // UnicodeString does not find empty substrings
1037 if(srcLength < 0 && srcChars[srcStart] == 0) {
1038 return -1;
1039 }
1040
1041 // get the indices within bounds
1042 pinIndices(start, length);
1043
1044 // find the first occurrence of the substring
1045 const UChar *array = getArrayStart();
1046 const UChar *match = u_strFindFirstu_strFindFirst_71(array + start, length, srcChars + srcStart, srcLength);
1047 if(match == NULL__null) {
1048 return -1;
1049 } else {
1050 return (int32_t)(match - array);
1051 }
1052}
1053
1054int32_t
1055UnicodeString::doIndexOf(UChar c,
1056 int32_t start,
1057 int32_t length) const
1058{
1059 // pin indices
1060 pinIndices(start, length);
1061
1062 // find the first occurrence of c
1063 const UChar *array = getArrayStart();
1064 const UChar *match = u_memchru_memchr_71(array + start, c, length);
1065 if(match == NULL__null) {
1066 return -1;
1067 } else {
1068 return (int32_t)(match - array);
1069 }
1070}
1071
1072int32_t
1073UnicodeString::doIndexOf(UChar32 c,
1074 int32_t start,
1075 int32_t length) const {
1076 // pin indices
1077 pinIndices(start, length);
1078
1079 // find the first occurrence of c
1080 const UChar *array = getArrayStart();
1081 const UChar *match = u_memchr32u_memchr32_71(array + start, c, length);
1082 if(match == NULL__null) {
1083 return -1;
1084 } else {
1085 return (int32_t)(match - array);
1086 }
1087}
1088
1089int32_t
1090UnicodeString::lastIndexOf(const UChar *srcChars,
1091 int32_t srcStart,
1092 int32_t srcLength,
1093 int32_t start,
1094 int32_t length) const
1095{
1096 if(isBogus() || srcChars == 0 || srcStart < 0 || srcLength == 0) {
1097 return -1;
1098 }
1099
1100 // UnicodeString does not find empty substrings
1101 if(srcLength < 0 && srcChars[srcStart] == 0) {
1102 return -1;
1103 }
1104
1105 // get the indices within bounds
1106 pinIndices(start, length);
1107
1108 // find the last occurrence of the substring
1109 const UChar *array = getArrayStart();
1110 const UChar *match = u_strFindLastu_strFindLast_71(array + start, length, srcChars + srcStart, srcLength);
1111 if(match == NULL__null) {
1112 return -1;
1113 } else {
1114 return (int32_t)(match - array);
1115 }
1116}
1117
1118int32_t
1119UnicodeString::doLastIndexOf(UChar c,
1120 int32_t start,
1121 int32_t length) const
1122{
1123 if(isBogus()) {
1124 return -1;
1125 }
1126
1127 // pin indices
1128 pinIndices(start, length);
1129
1130 // find the last occurrence of c
1131 const UChar *array = getArrayStart();
1132 const UChar *match = u_memrchru_memrchr_71(array + start, c, length);
1133 if(match == NULL__null) {
1134 return -1;
1135 } else {
1136 return (int32_t)(match - array);
1137 }
1138}
1139
1140int32_t
1141UnicodeString::doLastIndexOf(UChar32 c,
1142 int32_t start,
1143 int32_t length) const {
1144 // pin indices
1145 pinIndices(start, length);
1146
1147 // find the last occurrence of c
1148 const UChar *array = getArrayStart();
1149 const UChar *match = u_memrchr32u_memrchr32_71(array + start, c, length);
1150 if(match == NULL__null) {
1151 return -1;
1152 } else {
1153 return (int32_t)(match - array);
1154 }
1155}
1156
1157//========================================
1158// Write implementation
1159//========================================
1160
1161UnicodeString&
1162UnicodeString::findAndReplace(int32_t start,
1163 int32_t length,
1164 const UnicodeString& oldText,
1165 int32_t oldStart,
1166 int32_t oldLength,
1167 const UnicodeString& newText,
1168 int32_t newStart,
1169 int32_t newLength)
1170{
1171 if(isBogus() || oldText.isBogus() || newText.isBogus()) {
1172 return *this;
1173 }
1174
1175 pinIndices(start, length);
1176 oldText.pinIndices(oldStart, oldLength);
1177 newText.pinIndices(newStart, newLength);
1178
1179 if(oldLength == 0) {
1180 return *this;
1181 }
1182
1183 while(length > 0 && length >= oldLength) {
1184 int32_t pos = indexOf(oldText, oldStart, oldLength, start, length);
1185 if(pos < 0) {
1186 // no more oldText's here: done
1187 break;
1188 } else {
1189 // we found oldText, replace it by newText and go beyond it
1190 replace(pos, oldLength, newText, newStart, newLength);
1191 length -= pos + oldLength - start;
1192 start = pos + newLength;
1193 }
1194 }
1195
1196 return *this;
1197}
1198
1199
1200void
1201UnicodeString::setToBogus()
1202{
1203 releaseArray();
1204
1205 fUnion.fFields.fLengthAndFlags = kIsBogus;
1206 fUnion.fFields.fArray = 0;
1207 fUnion.fFields.fCapacity = 0;
1208}
1209
1210// turn a bogus string into an empty one
1211void
1212UnicodeString::unBogus() {
1213 if(fUnion.fFields.fLengthAndFlags & kIsBogus) {
1214 setToEmpty();
1215 }
1216}
1217
1218const char16_t *
1219UnicodeString::getTerminatedBuffer() {
1220 if(!isWritable()) {
1221 return nullptr;
1222 }
1223 UChar *array = getArrayStart();
1224 int32_t len = length();
1225 if(len < getCapacity()) {
1226 if(fUnion.fFields.fLengthAndFlags & kBufferIsReadonly) {
1227 // If len<capacity on a read-only alias, then array[len] is
1228 // either the original NUL (if constructed with (TRUE, s, length))
1229 // or one of the original string contents characters (if later truncated),
1230 // therefore we can assume that array[len] is initialized memory.
1231 if(array[len] == 0) {
1232 return array;
1233 }
1234 } else if(((fUnion.fFields.fLengthAndFlags & kRefCounted) == 0 || refCount() == 1)) {
1235 // kRefCounted: Do not write the NUL if the buffer is shared.
1236 // That is mostly safe, except when the length of one copy was modified
1237 // without copy-on-write, e.g., via truncate(newLength) or remove(void).
1238 // Then the NUL would be written into the middle of another copy's string.
1239
1240 // Otherwise, the buffer is fully writable and it is anyway safe to write the NUL.
1241 // Do not test if there is a NUL already because it might be uninitialized memory.
1242 // (That would be safe, but tools like valgrind & Purify would complain.)
1243 array[len] = 0;
1244 return array;
1245 }
1246 }
1247 if(len<INT32_MAX(2147483647) && cloneArrayIfNeeded(len+1)) {
1248 array = getArrayStart();
1249 array[len] = 0;
1250 return array;
1251 } else {
1252 return nullptr;
1253 }
1254}
1255
1256// setTo() analogous to the readonly-aliasing constructor with the same signature
1257UnicodeString &
1258UnicodeString::setTo(UBool isTerminated,
1259 ConstChar16Ptr textPtr,
1260 int32_t textLength)
1261{
1262 if(fUnion.fFields.fLengthAndFlags & kOpenGetBuffer) {
1263 // do not modify a string that has an "open" getBuffer(minCapacity)
1264 return *this;
1265 }
1266
1267 const UChar *text = textPtr;
1268 if(text == NULL__null) {
1269 // treat as an empty string, do not alias
1270 releaseArray();
1271 setToEmpty();
1272 return *this;
1273 }
1274
1275 if( textLength < -1 ||
1276 (textLength == -1 && !isTerminated) ||
1277 (textLength >= 0 && isTerminated && text[textLength] != 0)
1278 ) {
1279 setToBogus();
1280 return *this;
1281 }
1282
1283 releaseArray();
1284
1285 if(textLength == -1) {
1286 // text is terminated, or else it would have failed the above test
1287 textLength = u_strlenu_strlen_71(text);
1288 }
1289 fUnion.fFields.fLengthAndFlags = kReadonlyAlias;
1290 setArray((UChar *)text, textLength, isTerminated ? textLength + 1 : textLength);
1291 return *this;
1292}
1293
1294// setTo() analogous to the writable-aliasing constructor with the same signature
1295UnicodeString &
1296UnicodeString::setTo(UChar *buffer,
1297 int32_t buffLength,
1298 int32_t buffCapacity) {
1299 if(fUnion.fFields.fLengthAndFlags & kOpenGetBuffer) {
1300 // do not modify a string that has an "open" getBuffer(minCapacity)
1301 return *this;
1302 }
1303
1304 if(buffer == NULL__null) {
1305 // treat as an empty string, do not alias
1306 releaseArray();
1307 setToEmpty();
1308 return *this;
1309 }
1310
1311 if(buffLength < -1 || buffCapacity < 0 || buffLength > buffCapacity) {
1312 setToBogus();
1313 return *this;
1314 } else if(buffLength == -1) {
1315 // buffLength = u_strlen(buff); but do not look beyond buffCapacity
1316 const UChar *p = buffer, *limit = buffer + buffCapacity;
1317 while(p != limit && *p != 0) {
1318 ++p;
1319 }
1320 buffLength = (int32_t)(p - buffer);
1321 }
1322
1323 releaseArray();
1324
1325 fUnion.fFields.fLengthAndFlags = kWritableAlias;
1326 setArray(buffer, buffLength, buffCapacity);
1327 return *this;
1328}
1329
1330UnicodeString &UnicodeString::setToUTF8(StringPiece utf8) {
1331 unBogus();
1332 int32_t length = utf8.length();
1333 int32_t capacity;
1334 // The UTF-16 string will be at most as long as the UTF-8 string.
1335 if(length <= US_STACKBUF_SIZE) {
1336 capacity = US_STACKBUF_SIZE;
1337 } else {
1338 capacity = length + 1; // +1 for the terminating NUL.
1339 }
1340 UChar *utf16 = getBuffer(capacity);
1341 int32_t length16;
1342 UErrorCode errorCode = U_ZERO_ERROR;
1343 u_strFromUTF8WithSubu_strFromUTF8WithSub_71(utf16, getCapacity(), &length16,
1344 utf8.data(), length,
1345 0xfffd, // Substitution character.
1346 NULL__null, // Don't care about number of substitutions.
1347 &errorCode);
1348 releaseBuffer(length16);
1349 if(U_FAILURE(errorCode)) {
1350 setToBogus();
1351 }
1352 return *this;
1353}
1354
1355UnicodeString&
1356UnicodeString::setCharAt(int32_t offset,
1357 UChar c)
1358{
1359 int32_t len = length();
1360 if(cloneArrayIfNeeded() && len > 0) {
1361 if(offset < 0) {
1362 offset = 0;
1363 } else if(offset >= len) {
1364 offset = len - 1;
1365 }
1366
1367 getArrayStart()[offset] = c;
1368 }
1369 return *this;
1370}
1371
1372UnicodeString&
1373UnicodeString::replace(int32_t start,
1374 int32_t _length,
1375 UChar32 srcChar) {
1376 UChar buffer[U16_MAX_LENGTH2];
1377 int32_t count = 0;
1378 UBool isError = FALSE0;
1379 U16_APPEND(buffer, count, U16_MAX_LENGTH, srcChar, isError)do { if((uint32_t)(srcChar)<=0xffff) { (buffer)[(count)++]
=(uint16_t)(srcChar); } else if((uint32_t)(srcChar)<=0x10ffff
&& (count)+1<(2)) { (buffer)[(count)++]=(uint16_t
)(((srcChar)>>10)+0xd7c0); (buffer)[(count)++]=(uint16_t
)(((srcChar)&0x3ff)|0xdc00); } else { (isError)=true; } }
while (false)
;
1380 // We test isError so that the compiler does not complain that we don't.
1381 // If isError (srcChar is not a valid code point) then count==0 which means
1382 // we remove the source segment rather than replacing it with srcChar.
1383 return doReplace(start, _length, buffer, 0, isError ? 0 : count);
1384}
1385
1386UnicodeString&
1387UnicodeString::append(UChar32 srcChar) {
1388 UChar buffer[U16_MAX_LENGTH2];
1389 int32_t _length = 0;
1390 UBool isError = FALSE0;
1391 U16_APPEND(buffer, _length, U16_MAX_LENGTH, srcChar, isError)do { if((uint32_t)(srcChar)<=0xffff) { (buffer)[(_length)++
]=(uint16_t)(srcChar); } else if((uint32_t)(srcChar)<=0x10ffff
&& (_length)+1<(2)) { (buffer)[(_length)++]=(uint16_t
)(((srcChar)>>10)+0xd7c0); (buffer)[(_length)++]=(uint16_t
)(((srcChar)&0x3ff)|0xdc00); } else { (isError)=true; } }
while (false)
;
1392 // We test isError so that the compiler does not complain that we don't.
1393 // If isError then _length==0 which turns the doAppend() into a no-op anyway.
1394 return isError ? *this : doAppend(buffer, 0, _length);
1395}
1396
1397UnicodeString&
1398UnicodeString::doReplace( int32_t start,
1399 int32_t length,
1400 const UnicodeString& src,
1401 int32_t srcStart,
1402 int32_t srcLength)
1403{
1404 // pin the indices to legal values
1405 src.pinIndices(srcStart, srcLength);
1406
1407 // get the characters from src
1408 // and replace the range in ourselves with them
1409 return doReplace(start, length, src.getArrayStart(), srcStart, srcLength);
1410}
1411
1412UnicodeString&
1413UnicodeString::doReplace(int32_t start,
1414 int32_t length,
1415 const UChar *srcChars,
1416 int32_t srcStart,
1417 int32_t srcLength)
1418{
1419 if(!isWritable()) {
1420 return *this;
1421 }
1422
1423 int32_t oldLength = this->length();
1424
1425 // optimize (read-only alias).remove(0, start) and .remove(start, end)
1426 if((fUnion.fFields.fLengthAndFlags&kBufferIsReadonly) && srcLength == 0) {
1427 if(start == 0) {
1428 // remove prefix by adjusting the array pointer
1429 pinIndex(length);
1430 fUnion.fFields.fArray += length;
1431 fUnion.fFields.fCapacity -= length;
1432 setLength(oldLength - length);
1433 return *this;
1434 } else {
1435 pinIndex(start);
1436 if(length >= (oldLength - start)) {
1437 // remove suffix by reducing the length (like truncate())
1438 setLength(start);
1439 fUnion.fFields.fCapacity = start; // not NUL-terminated any more
1440 return *this;
1441 }
1442 }
1443 }
1444
1445 if(start == oldLength) {
1446 return doAppend(srcChars, srcStart, srcLength);
1447 }
1448
1449 if(srcChars == 0) {
1450 srcLength = 0;
1451 } else {
1452 // Perform all remaining operations relative to srcChars + srcStart.
1453 // From this point forward, do not use srcStart.
1454 srcChars += srcStart;
1455 if (srcLength < 0) {
1456 // get the srcLength if necessary
1457 srcLength = u_strlenu_strlen_71(srcChars);
1458 }
1459 }
1460
1461 // pin the indices to legal values
1462 pinIndices(start, length);
1463
1464 // Calculate the size of the string after the replace.
1465 // Avoid int32_t overflow.
1466 int32_t newLength = oldLength - length;
1467 if(srcLength > (INT32_MAX(2147483647) - newLength)) {
1468 setToBogus();
1469 return *this;
1470 }
1471 newLength += srcLength;
1472
1473 // Check for insertion into ourself
1474 const UChar *oldArray = getArrayStart();
1475 if (isBufferWritable() &&
1476 oldArray < srcChars + srcLength &&
1477 srcChars < oldArray + oldLength) {
1478 // Copy into a new UnicodeString and start over
1479 UnicodeString copy(srcChars, srcLength);
1480 if (copy.isBogus()) {
1481 setToBogus();
1482 return *this;
1483 }
1484 return doReplace(start, length, copy.getArrayStart(), 0, srcLength);
1485 }
1486
1487 // cloneArrayIfNeeded(doCopyArray=FALSE) may change fArray but will not copy the current contents;
1488 // therefore we need to keep the current fArray
1489 UChar oldStackBuffer[US_STACKBUF_SIZE];
1490 if((fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) && (newLength > US_STACKBUF_SIZE)) {
1491 // copy the stack buffer contents because it will be overwritten with
1492 // fUnion.fFields values
1493 u_memcpyu_memcpy_71(oldStackBuffer, oldArray, oldLength);
1494 oldArray = oldStackBuffer;
1495 }
1496
1497 // clone our array and allocate a bigger array if needed
1498 int32_t *bufferToDelete = 0;
1499 if(!cloneArrayIfNeeded(newLength, getGrowCapacity(newLength),
1500 FALSE0, &bufferToDelete)
1501 ) {
1502 return *this;
1503 }
1504
1505 // now do the replace
1506
1507 UChar *newArray = getArrayStart();
1508 if(newArray != oldArray) {
1509 // if fArray changed, then we need to copy everything except what will change
1510 us_arrayCopy(oldArray, 0, newArray, 0, start);
1511 us_arrayCopy(oldArray, start + length,
1512 newArray, start + srcLength,
1513 oldLength - (start + length));
1514 } else if(length != srcLength) {
1515 // fArray did not change; copy only the portion that isn't changing, leaving a hole
1516 us_arrayCopy(oldArray, start + length,
1517 newArray, start + srcLength,
1518 oldLength - (start + length));
1519 }
1520
1521 // now fill in the hole with the new string
1522 us_arrayCopy(srcChars, 0, newArray, start, srcLength);
1523
1524 setLength(newLength);
1525
1526 // delayed delete in case srcChars == fArray when we started, and
1527 // to keep oldArray alive for the above operations
1528 if (bufferToDelete) {
1529 uprv_freeuprv_free_71(bufferToDelete);
1530 }
1531
1532 return *this;
1533}
1534
1535// Versions of doReplace() only for append() variants.
1536// doReplace() and doAppend() optimize for different cases.
1537
1538UnicodeString&
1539UnicodeString::doAppend(const UnicodeString& src, int32_t srcStart, int32_t srcLength) {
1540 if(srcLength == 0) {
1541 return *this;
1542 }
1543
1544 // pin the indices to legal values
1545 src.pinIndices(srcStart, srcLength);
1546 return doAppend(src.getArrayStart(), srcStart, srcLength);
1547}
1548
1549UnicodeString&
1550UnicodeString::doAppend(const UChar *srcChars, int32_t srcStart, int32_t srcLength) {
1551 if(!isWritable() || srcLength == 0 || srcChars == NULL__null) {
2
Assuming 'srcLength' is not equal to 0
3
Assuming 'srcChars' is not equal to NULL
4
Taking false branch
1552 return *this;
1553 }
1554
1555 // Perform all remaining operations relative to srcChars + srcStart.
1556 // From this point forward, do not use srcStart.
1557 srcChars += srcStart;
1558
1559 if(srcLength < 0) {
5
Assuming 'srcLength' is >= 0
6
Taking false branch
1560 // get the srcLength if necessary
1561 if((srcLength = u_strlenu_strlen_71(srcChars)) == 0) {
1562 return *this;
1563 }
1564 }
1565
1566 int32_t oldLength = length();
1567 int32_t newLength;
1568 if (uprv_add32_overflowuprv_add32_overflow_71(oldLength, srcLength, &newLength)) {
7
Value assigned to 'newLength'
8
Assuming the condition is false
9
Taking false branch
1569 setToBogus();
1570 return *this;
1571 }
1572
1573 // Check for append onto ourself
1574 const UChar* oldArray = getArrayStart();
1575 if (isBufferWritable() &&
1576 oldArray < srcChars + srcLength &&
1577 srcChars < oldArray + oldLength) {
1578 // Copy into a new UnicodeString and start over
1579 UnicodeString copy(srcChars, srcLength);
1580 if (copy.isBogus()) {
1581 setToBogus();
1582 return *this;
1583 }
1584 return doAppend(copy.getArrayStart(), 0, srcLength);
1585 }
1586
1587 // optimize append() onto a large-enough, owned string
1588 if((newLength <= getCapacity() && isBufferWritable()) ||
10
Assuming the condition is false
11
Taking true branch
1589 cloneArrayIfNeeded(newLength, getGrowCapacity(newLength))) {
1590 UChar *newArray = getArrayStart();
1591 // Do not copy characters when
1592 // UChar *buffer=str.getAppendBuffer(...);
1593 // is followed by
1594 // str.append(buffer, length);
1595 // or
1596 // str.appendString(buffer, length)
1597 // or similar.
1598 if(srcChars != newArray + oldLength) {
12
Assuming the condition is false
13
Taking false branch
1599 us_arrayCopy(srcChars, 0, newArray, oldLength, srcLength);
1600 }
1601 setLength(newLength);
14
Passing the value -1 via 1st parameter 'len'
15
Calling 'UnicodeString::setLength'
1602 }
1603 return *this;
1604}
1605
1606/**
1607 * Replaceable API
1608 */
1609void
1610UnicodeString::handleReplaceBetween(int32_t start,
1611 int32_t limit,
1612 const UnicodeString& text) {
1613 replaceBetween(start, limit, text);
1614}
1615
1616/**
1617 * Replaceable API
1618 */
1619void
1620UnicodeString::copy(int32_t start, int32_t limit, int32_t dest) {
1621 if (limit <= start) {
1622 return; // Nothing to do; avoid bogus malloc call
1623 }
1624 UChar* text = (UChar*) uprv_mallocuprv_malloc_71( sizeof(UChar) * (limit - start) );
1625 // Check to make sure text is not null.
1626 if (text != NULL__null) {
1627 extractBetween(start, limit, text, 0);
1628 insert(dest, text, 0, limit - start);
1629 uprv_freeuprv_free_71(text);
1630 }
1631}
1632
1633/**
1634 * Replaceable API
1635 *
1636 * NOTE: This is for the Replaceable class. There is no rep.cpp,
1637 * so we implement this function here.
1638 */
1639UBool Replaceable::hasMetaData() const {
1640 return TRUE1;
1641}
1642
1643/**
1644 * Replaceable API
1645 */
1646UBool UnicodeString::hasMetaData() const {
1647 return FALSE0;
1648}
1649
1650UnicodeString&
1651UnicodeString::doReverse(int32_t start, int32_t length) {
1652 if(length <= 1 || !cloneArrayIfNeeded()) {
1653 return *this;
1654 }
1655
1656 // pin the indices to legal values
1657 pinIndices(start, length);
1658 if(length <= 1) { // pinIndices() might have shrunk the length
1659 return *this;
1660 }
1661
1662 UChar *left = getArrayStart() + start;
1663 UChar *right = left + length - 1; // -1 for inclusive boundary (length>=2)
1664 UChar swap;
1665 UBool hasSupplementary = FALSE0;
1666
1667 // Before the loop we know left<right because length>=2.
1668 do {
1669 hasSupplementary |= (UBool)U16_IS_LEAD(swap = *left)(((swap = *left)&0xfffffc00)==0xd800);
1670 hasSupplementary |= (UBool)U16_IS_LEAD(*left++ = *right)(((*left++ = *right)&0xfffffc00)==0xd800);
1671 *right-- = swap;
1672 } while(left < right);
1673 // Make sure to test the middle code unit of an odd-length string.
1674 // Redundant if the length is even.
1675 hasSupplementary |= (UBool)U16_IS_LEAD(*left)(((*left)&0xfffffc00)==0xd800);
1676
1677 /* if there are supplementary code points in the reversed range, then re-swap their surrogates */
1678 if(hasSupplementary) {
1679 UChar swap2;
1680
1681 left = getArrayStart() + start;
1682 right = left + length - 1; // -1 so that we can look at *(left+1) if left<right
1683 while(left < right) {
1684 if(U16_IS_TRAIL(swap = *left)(((swap = *left)&0xfffffc00)==0xdc00) && U16_IS_LEAD(swap2 = *(left + 1))(((swap2 = *(left + 1))&0xfffffc00)==0xd800)) {
1685 *left++ = swap2;
1686 *left++ = swap;
1687 } else {
1688 ++left;
1689 }
1690 }
1691 }
1692
1693 return *this;
1694}
1695
1696UBool
1697UnicodeString::padLeading(int32_t targetLength,
1698 UChar padChar)
1699{
1700 int32_t oldLength = length();
1701 if(oldLength >= targetLength || !cloneArrayIfNeeded(targetLength)) {
1702 return FALSE0;
1703 } else {
1704 // move contents up by padding width
1705 UChar *array = getArrayStart();
1706 int32_t start = targetLength - oldLength;
1707 us_arrayCopy(array, 0, array, start, oldLength);
1708
1709 // fill in padding character
1710 while(--start >= 0) {
1711 array[start] = padChar;
1712 }
1713 setLength(targetLength);
1714 return TRUE1;
1715 }
1716}
1717
1718UBool
1719UnicodeString::padTrailing(int32_t targetLength,
1720 UChar padChar)
1721{
1722 int32_t oldLength = length();
1723 if(oldLength >= targetLength || !cloneArrayIfNeeded(targetLength)) {
1724 return FALSE0;
1725 } else {
1726 // fill in padding character
1727 UChar *array = getArrayStart();
1728 int32_t length = targetLength;
1729 while(--length >= oldLength) {
1730 array[length] = padChar;
1731 }
1732 setLength(targetLength);
1733 return TRUE1;
1734 }
1735}
1736
1737//========================================
1738// Hashing
1739//========================================
1740int32_t
1741UnicodeString::doHashCode() const
1742{
1743 /* Delegate hash computation to uhash. This makes UnicodeString
1744 * hashing consistent with UChar* hashing. */
1745 int32_t hashCode = ustr_hashUCharsNustr_hashUCharsN_71(getArrayStart(), length());
1746 if (hashCode == kInvalidHashCode) {
1747 hashCode = kEmptyHashCode;
1748 }
1749 return hashCode;
1750}
1751
1752//========================================
1753// External Buffer
1754//========================================
1755
1756char16_t *
1757UnicodeString::getBuffer(int32_t minCapacity) {
1758 if(minCapacity>=-1 && cloneArrayIfNeeded(minCapacity)) {
1759 fUnion.fFields.fLengthAndFlags|=kOpenGetBuffer;
1760 setZeroLength();
1761 return getArrayStart();
1762 } else {
1763 return nullptr;
1764 }
1765}
1766
1767void
1768UnicodeString::releaseBuffer(int32_t newLength) {
1769 if(fUnion.fFields.fLengthAndFlags&kOpenGetBuffer && newLength>=-1) {
1770 // set the new fLength
1771 int32_t capacity=getCapacity();
1772 if(newLength==-1) {
1773 // the new length is the string length, capped by fCapacity
1774 const UChar *array=getArrayStart(), *p=array, *limit=array+capacity;
1775 while(p<limit && *p!=0) {
1776 ++p;
1777 }
1778 newLength=(int32_t)(p-array);
1779 } else if(newLength>capacity) {
1780 newLength=capacity;
1781 }
1782 setLength(newLength);
1783 fUnion.fFields.fLengthAndFlags&=~kOpenGetBuffer;
1784 }
1785}
1786
1787//========================================
1788// Miscellaneous
1789//========================================
1790UBool
1791UnicodeString::cloneArrayIfNeeded(int32_t newCapacity,
1792 int32_t growCapacity,
1793 UBool doCopyArray,
1794 int32_t **pBufferToDelete,
1795 UBool forceClone) {
1796 // default parameters need to be static, therefore
1797 // the defaults are -1 to have convenience defaults
1798 if(newCapacity == -1) {
1799 newCapacity = getCapacity();
1800 }
1801
1802 // while a getBuffer(minCapacity) is "open",
1803 // prevent any modifications of the string by returning FALSE here
1804 // if the string is bogus, then only an assignment or similar can revive it
1805 if(!isWritable()) {
1806 return FALSE0;
1807 }
1808
1809 /*
1810 * We need to make a copy of the array if
1811 * the buffer is read-only, or
1812 * the buffer is refCounted (shared), and refCount>1, or
1813 * the buffer is too small.
1814 * Return FALSE if memory could not be allocated.
1815 */
1816 if(forceClone ||
1817 fUnion.fFields.fLengthAndFlags & kBufferIsReadonly ||
1818 (fUnion.fFields.fLengthAndFlags & kRefCounted && refCount() > 1) ||
1819 newCapacity > getCapacity()
1820 ) {
1821 // check growCapacity for default value and use of the stack buffer
1822 if(growCapacity < 0) {
1823 growCapacity = newCapacity;
1824 } else if(newCapacity <= US_STACKBUF_SIZE && growCapacity > US_STACKBUF_SIZE) {
1825 growCapacity = US_STACKBUF_SIZE;
1826 }
1827
1828 // save old values
1829 UChar oldStackBuffer[US_STACKBUF_SIZE];
1830 UChar *oldArray;
1831 int32_t oldLength = length();
1832 int16_t flags = fUnion.fFields.fLengthAndFlags;
1833
1834 if(flags&kUsingStackBuffer) {
1835 U_ASSERT(!(flags&kRefCounted))(void)0; /* kRefCounted and kUsingStackBuffer are mutally exclusive */
1836 if(doCopyArray && growCapacity > US_STACKBUF_SIZE) {
1837 // copy the stack buffer contents because it will be overwritten with
1838 // fUnion.fFields values
1839 us_arrayCopy(fUnion.fStackFields.fBuffer, 0, oldStackBuffer, 0, oldLength);
1840 oldArray = oldStackBuffer;
1841 } else {
1842 oldArray = NULL__null; // no need to copy from the stack buffer to itself
1843 }
1844 } else {
1845 oldArray = fUnion.fFields.fArray;
1846 U_ASSERT(oldArray!=NULL)(void)0; /* when stack buffer is not used, oldArray must have a non-NULL reference */
1847 }
1848
1849 // allocate a new array
1850 if(allocate(growCapacity) ||
1851 (newCapacity < growCapacity && allocate(newCapacity))
1852 ) {
1853 if(doCopyArray) {
1854 // copy the contents
1855 // do not copy more than what fits - it may be smaller than before
1856 int32_t minLength = oldLength;
1857 newCapacity = getCapacity();
1858 if(newCapacity < minLength) {
1859 minLength = newCapacity;
1860 }
1861 if(oldArray != NULL__null) {
1862 us_arrayCopy(oldArray, 0, getArrayStart(), 0, minLength);
1863 }
1864 setLength(minLength);
1865 } else {
1866 setZeroLength();
1867 }
1868
1869 // release the old array
1870 if(flags & kRefCounted) {
1871 // the array is refCounted; decrement and release if 0
1872 u_atomic_int32_t *pRefCount = ((u_atomic_int32_t *)oldArray - 1);
1873 if(umtx_atomic_dec(pRefCount) == 0) {
1874 if(pBufferToDelete == 0) {
1875 // Note: cast to (void *) is needed with MSVC, where u_atomic_int32_t
1876 // is defined as volatile. (Volatile has useful non-standard behavior
1877 // with this compiler.)
1878 uprv_freeuprv_free_71((void *)pRefCount);
1879 } else {
1880 // the caller requested to delete it himself
1881 *pBufferToDelete = (int32_t *)pRefCount;
1882 }
1883 }
1884 }
1885 } else {
1886 // not enough memory for growCapacity and not even for the smaller newCapacity
1887 // reset the old values for setToBogus() to release the array
1888 if(!(flags&kUsingStackBuffer)) {
1889 fUnion.fFields.fArray = oldArray;
1890 }
1891 fUnion.fFields.fLengthAndFlags = flags;
1892 setToBogus();
1893 return FALSE0;
1894 }
1895 }
1896 return TRUE1;
1897}
1898
1899// UnicodeStringAppendable ------------------------------------------------- ***
1900
1901UnicodeStringAppendable::~UnicodeStringAppendable() {}
1902
1903UBool
1904UnicodeStringAppendable::appendCodeUnit(UChar c) {
1905 return str.doAppend(&c, 0, 1).isWritable();
1906}
1907
1908UBool
1909UnicodeStringAppendable::appendCodePoint(UChar32 c) {
1910 UChar buffer[U16_MAX_LENGTH2];
1911 int32_t cLength = 0;
1912 UBool isError = FALSE0;
1913 U16_APPEND(buffer, cLength, U16_MAX_LENGTH, c, isError)do { if((uint32_t)(c)<=0xffff) { (buffer)[(cLength)++]=(uint16_t
)(c); } else if((uint32_t)(c)<=0x10ffff && (cLength
)+1<(2)) { (buffer)[(cLength)++]=(uint16_t)(((c)>>10
)+0xd7c0); (buffer)[(cLength)++]=(uint16_t)(((c)&0x3ff)|0xdc00
); } else { (isError)=true; } } while (false)
;
1914 return !isError && str.doAppend(buffer, 0, cLength).isWritable();
1915}
1916
1917UBool
1918UnicodeStringAppendable::appendString(const UChar *s, int32_t length) {
1919 return str.doAppend(s, 0, length).isWritable();
1
Calling 'UnicodeString::doAppend'
1920}
1921
1922UBool
1923UnicodeStringAppendable::reserveAppendCapacity(int32_t appendCapacity) {
1924 return str.cloneArrayIfNeeded(str.length() + appendCapacity);
1925}
1926
1927UChar *
1928UnicodeStringAppendable::getAppendBuffer(int32_t minCapacity,
1929 int32_t desiredCapacityHint,
1930 UChar *scratch, int32_t scratchCapacity,
1931 int32_t *resultCapacity) {
1932 if(minCapacity < 1 || scratchCapacity < minCapacity) {
1933 *resultCapacity = 0;
1934 return NULL__null;
1935 }
1936 int32_t oldLength = str.length();
1937 if(minCapacity <= (kMaxCapacity - oldLength) &&
1938 desiredCapacityHint <= (kMaxCapacity - oldLength) &&
1939 str.cloneArrayIfNeeded(oldLength + minCapacity, oldLength + desiredCapacityHint)) {
1940 *resultCapacity = str.getCapacity() - oldLength;
1941 return str.getArrayStart() + oldLength;
1942 }
1943 *resultCapacity = scratchCapacity;
1944 return scratch;
1945}
1946
1947U_NAMESPACE_END}
1948
1949U_NAMESPACE_USEusing namespace icu_71;
1950
1951U_CAPIextern "C" int32_t U_EXPORT2
1952uhash_hashUnicodeStringuhash_hashUnicodeString_71(const UElement key) {
1953 const UnicodeString *str = (const UnicodeString*) key.pointer;
1954 return (str == NULL__null) ? 0 : str->hashCode();
1955}
1956
1957// Moved here from uhash_us.cpp so that using a UVector of UnicodeString*
1958// does not depend on hashtable code.
1959U_CAPIextern "C" UBool U_EXPORT2
1960uhash_compareUnicodeStringuhash_compareUnicodeString_71(const UElement key1, const UElement key2) {
1961 const UnicodeString *str1 = (const UnicodeString*) key1.pointer;
1962 const UnicodeString *str2 = (const UnicodeString*) key2.pointer;
1963 if (str1 == str2) {
1964 return TRUE1;
1965 }
1966 if (str1 == NULL__null || str2 == NULL__null) {
1967 return FALSE0;
1968 }
1969 return *str1 == *str2;
1970}
1971
1972#ifdef U_STATIC_IMPLEMENTATION1
1973/*
1974This should never be called. It is defined here to make sure that the
1975virtual vector deleting destructor is defined within unistr.cpp.
1976The vector deleting destructor is already a part of UObject,
1977but defining it here makes sure that it is included with this object file.
1978This makes sure that static library dependencies are kept to a minimum.
1979*/
1980#if defined(__clang__1) || U_GCC_MAJOR_MINOR(4 * 100 + 2) >= 1100
1981#pragma GCC diagnostic push
1982#pragma GCC diagnostic ignored "-Wunused-function"
1983static void uprv_UnicodeStringDummy(void) {
1984 delete [] (new UnicodeString[2]);
1985}
1986#pragma GCC diagnostic pop
1987#endif
1988#endif

../deps/icu-small/source/common/unicode/unistr.h

1// © 2016 and later: Unicode, Inc. and others.
2// License & terms of use: http://www.unicode.org/copyright.html
3/*
4**********************************************************************
5* Copyright (C) 1998-2016, International Business Machines
6* Corporation and others. All Rights Reserved.
7**********************************************************************
8*
9* File unistr.h
10*
11* Modification History:
12*
13* Date Name Description
14* 09/25/98 stephen Creation.
15* 11/11/98 stephen Changed per 11/9 code review.
16* 04/20/99 stephen Overhauled per 4/16 code review.
17* 11/18/99 aliu Made to inherit from Replaceable. Added method
18* handleReplaceBetween(); other methods unchanged.
19* 06/25/01 grhoten Remove dependency on iostream.
20******************************************************************************
21*/
22
23#ifndef UNISTR_H
24#define UNISTR_H
25
26/**
27 * \file
28 * \brief C++ API: Unicode String
29 */
30
31#include "unicode/utypes.h"
32
33#if U_SHOW_CPLUSPLUS_API1
34
35#include <cstddef>
36#include "unicode/char16ptr.h"
37#include "unicode/rep.h"
38#include "unicode/std_string.h"
39#include "unicode/stringpiece.h"
40#include "unicode/bytestream.h"
41
42struct UConverter; // unicode/ucnv.h
43
44#ifndef USTRING_H
45/**
46 * \ingroup ustring_ustrlen
47 * @param s Pointer to sequence of UChars.
48 * @return Length of sequence.
49 */
50U_CAPIextern "C" int32_t U_EXPORT2 u_strlenu_strlen_71(const UChar *s);
51#endif
52
53U_NAMESPACE_BEGINnamespace icu_71 {
54
55#if !UCONFIG_NO_BREAK_ITERATION0
56class BreakIterator; // unicode/brkiter.h
57#endif
58class Edits;
59
60U_NAMESPACE_END}
61
62// Not #ifndef U_HIDE_INTERNAL_API because UnicodeString needs the UStringCaseMapper.
63/**
64 * Internal string case mapping function type.
65 * All error checking must be done.
66 * src and dest must not overlap.
67 * @internal
68 */
69typedef int32_t U_CALLCONV
70UStringCaseMapper(int32_t caseLocale, uint32_t options,
71#if !UCONFIG_NO_BREAK_ITERATION0
72 icu::BreakIterator *iter,
73#endif
74 char16_t *dest, int32_t destCapacity,
75 const char16_t *src, int32_t srcLength,
76 icu::Edits *edits,
77 UErrorCode &errorCode);
78
79U_NAMESPACE_BEGINnamespace icu_71 {
80
81class Locale; // unicode/locid.h
82class StringCharacterIterator;
83class UnicodeStringAppendable; // unicode/appendable.h
84
85/* The <iostream> include has been moved to unicode/ustream.h */
86
87/**
88 * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
89 * which constructs a Unicode string from an invariant-character char * string.
90 * About invariant characters see utypes.h.
91 * This constructor has no runtime dependency on conversion code and is
92 * therefore recommended over ones taking a charset name string
93 * (where the empty string "" indicates invariant-character conversion).
94 *
95 * @stable ICU 3.2
96 */
97#define US_INVicu::UnicodeString::kInvariant icu::UnicodeString::kInvariant
98
99/**
100 * Unicode String literals in C++.
101 *
102 * Note: these macros are not recommended for new code.
103 * Prior to the availability of C++11 and u"unicode string literals",
104 * these macros were provided for portability and efficiency when
105 * initializing UnicodeStrings from literals.
106 *
107 * They work only for strings that contain "invariant characters", i.e.,
108 * only latin letters, digits, and some punctuation.
109 * See utypes.h for details.
110 *
111 * The string parameter must be a C string literal.
112 * The length of the string, not including the terminating
113 * `NUL`, must be specified as a constant.
114 * @stable ICU 2.0
115 */
116#if !U_CHAR16_IS_TYPEDEF0
117# define UNICODE_STRING(cs, _length)icu::UnicodeString(true, ucs, _length) icu::UnicodeString(true, u ## cs, _length)
118#else
119# define UNICODE_STRING(cs, _length)icu::UnicodeString(true, ucs, _length) icu::UnicodeString(true, (const char16_t*)u ## cs, _length)
120#endif
121
122/**
123 * Unicode String literals in C++.
124 * Dependent on the platform properties, different UnicodeString
125 * constructors should be used to create a UnicodeString object from
126 * a string literal.
127 * The macros are defined for improved performance.
128 * They work only for strings that contain "invariant characters", i.e.,
129 * only latin letters, digits, and some punctuation.
130 * See utypes.h for details.
131 *
132 * The string parameter must be a C string literal.
133 * @stable ICU 2.0
134 */
135#define UNICODE_STRING_SIMPLE(cs)icu::UnicodeString(true, ucs, -1) UNICODE_STRING(cs, -1)icu::UnicodeString(true, ucs, -1)
136
137/**
138 * \def UNISTR_FROM_CHAR_EXPLICIT
139 * This can be defined to be empty or "explicit".
140 * If explicit, then the UnicodeString(char16_t) and UnicodeString(UChar32)
141 * constructors are marked as explicit, preventing their inadvertent use.
142 * @stable ICU 49
143 */
144#ifndef UNISTR_FROM_CHAR_EXPLICITexplicit
145# if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION1) || defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION)
146 // Auto-"explicit" in ICU library code.
147# define UNISTR_FROM_CHAR_EXPLICITexplicit explicit
148# else
149 // Empty by default for source code compatibility.
150# define UNISTR_FROM_CHAR_EXPLICITexplicit
151# endif
152#endif
153
154/**
155 * \def UNISTR_FROM_STRING_EXPLICIT
156 * This can be defined to be empty or "explicit".
157 * If explicit, then the UnicodeString(const char *) and UnicodeString(const char16_t *)
158 * constructors are marked as explicit, preventing their inadvertent use.
159 *
160 * In particular, this helps prevent accidentally depending on ICU conversion code
161 * by passing a string literal into an API with a const UnicodeString & parameter.
162 * @stable ICU 49
163 */
164#ifndef UNISTR_FROM_STRING_EXPLICITexplicit
165# if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION1) || defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION)
166 // Auto-"explicit" in ICU library code.
167# define UNISTR_FROM_STRING_EXPLICITexplicit explicit
168# else
169 // Empty by default for source code compatibility.
170# define UNISTR_FROM_STRING_EXPLICITexplicit
171# endif
172#endif
173
174/**
175 * \def UNISTR_OBJECT_SIZE
176 * Desired sizeof(UnicodeString) in bytes.
177 * It should be a multiple of sizeof(pointer) to avoid unusable space for padding.
178 * The object size may want to be a multiple of 16 bytes,
179 * which is a common granularity for heap allocation.
180 *
181 * Any space inside the object beyond sizeof(vtable pointer) + 2
182 * is available for storing short strings inside the object.
183 * The bigger the object, the longer a string that can be stored inside the object,
184 * without additional heap allocation.
185 *
186 * Depending on a platform's pointer size, pointer alignment requirements,
187 * and struct padding, the compiler will usually round up sizeof(UnicodeString)
188 * to 4 * sizeof(pointer) (or 3 * sizeof(pointer) for P128 data models),
189 * to hold the fields for heap-allocated strings.
190 * Such a minimum size also ensures that the object is easily large enough
191 * to hold at least 2 char16_ts, for one supplementary code point (U16_MAX_LENGTH).
192 *
193 * sizeof(UnicodeString) >= 48 should work for all known platforms.
194 *
195 * For example, on a 64-bit machine where sizeof(vtable pointer) is 8,
196 * sizeof(UnicodeString) = 64 would leave space for
197 * (64 - sizeof(vtable pointer) - 2) / U_SIZEOF_UCHAR = (64 - 8 - 2) / 2 = 27
198 * char16_ts stored inside the object.
199 *
200 * The minimum object size on a 64-bit machine would be
201 * 4 * sizeof(pointer) = 4 * 8 = 32 bytes,
202 * and the internal buffer would hold up to 11 char16_ts in that case.
203 *
204 * @see U16_MAX_LENGTH
205 * @stable ICU 56
206 */
207#ifndef UNISTR_OBJECT_SIZE64
208# define UNISTR_OBJECT_SIZE64 64
209#endif
210
211/**
212 * UnicodeString is a string class that stores Unicode characters directly and provides
213 * similar functionality as the Java String and StringBuffer/StringBuilder classes.
214 * It is a concrete implementation of the abstract class Replaceable (for transliteration).
215 *
216 * The UnicodeString equivalent of std::string’s clear() is remove().
217 *
218 * A UnicodeString may "alias" an external array of characters
219 * (that is, point to it, rather than own the array)
220 * whose lifetime must then at least match the lifetime of the aliasing object.
221 * This aliasing may be preserved when returning a UnicodeString by value,
222 * depending on the compiler and the function implementation,
223 * via Return Value Optimization (RVO) or the move assignment operator.
224 * (However, the copy assignment operator does not preserve aliasing.)
225 * For details see the description of storage models at the end of the class API docs
226 * and in the User Guide chapter linked from there.
227 *
228 * The UnicodeString class is not suitable for subclassing.
229 *
230 * For an overview of Unicode strings in C and C++ see the
231 * [User Guide Strings chapter](https://unicode-org.github.io/icu/userguide/strings#strings-in-cc).
232 *
233 * In ICU, a Unicode string consists of 16-bit Unicode *code units*.
234 * A Unicode character may be stored with either one code unit
235 * (the most common case) or with a matched pair of special code units
236 * ("surrogates"). The data type for code units is char16_t.
237 * For single-character handling, a Unicode character code *point* is a value
238 * in the range 0..0x10ffff. ICU uses the UChar32 type for code points.
239 *
240 * Indexes and offsets into and lengths of strings always count code units, not code points.
241 * This is the same as with multi-byte char* strings in traditional string handling.
242 * Operations on partial strings typically do not test for code point boundaries.
243 * If necessary, the user needs to take care of such boundaries by testing for the code unit
244 * values or by using functions like
245 * UnicodeString::getChar32Start() and UnicodeString::getChar32Limit()
246 * (or, in C, the equivalent macros U16_SET_CP_START() and U16_SET_CP_LIMIT(), see utf.h).
247 *
248 * UnicodeString methods are more lenient with regard to input parameter values
249 * than other ICU APIs. In particular:
250 * - If indexes are out of bounds for a UnicodeString object
251 * (< 0 or > length()) then they are "pinned" to the nearest boundary.
252 * - If the buffer passed to an insert/append/replace operation is owned by the
253 * target object, e.g., calling str.append(str), an extra copy may take place
254 * to ensure safety.
255 * - If primitive string pointer values (e.g., const char16_t * or char *)
256 * for input strings are NULL, then those input string parameters are treated
257 * as if they pointed to an empty string.
258 * However, this is *not* the case for char * parameters for charset names
259 * or other IDs.
260 * - Most UnicodeString methods do not take a UErrorCode parameter because
261 * there are usually very few opportunities for failure other than a shortage
262 * of memory, error codes in low-level C++ string methods would be inconvenient,
263 * and the error code as the last parameter (ICU convention) would prevent
264 * the use of default parameter values.
265 * Instead, such methods set the UnicodeString into a "bogus" state
266 * (see isBogus()) if an error occurs.
267 *
268 * In string comparisons, two UnicodeString objects that are both "bogus"
269 * compare equal (to be transitive and prevent endless loops in sorting),
270 * and a "bogus" string compares less than any non-"bogus" one.
271 *
272 * Const UnicodeString methods are thread-safe. Multiple threads can use
273 * const methods on the same UnicodeString object simultaneously,
274 * but non-const methods must not be called concurrently (in multiple threads)
275 * with any other (const or non-const) methods.
276 *
277 * Similarly, const UnicodeString & parameters are thread-safe.
278 * One object may be passed in as such a parameter concurrently in multiple threads.
279 * This includes the const UnicodeString & parameters for
280 * copy construction, assignment, and cloning.
281 *
282 * UnicodeString uses several storage methods.
283 * String contents can be stored inside the UnicodeString object itself,
284 * in an allocated and shared buffer, or in an outside buffer that is "aliased".
285 * Most of this is done transparently, but careful aliasing in particular provides
286 * significant performance improvements.
287 * Also, the internal buffer is accessible via special functions.
288 * For details see the
289 * [User Guide Strings chapter](https://unicode-org.github.io/icu/userguide/strings#maximizing-performance-with-the-unicodestring-storage-model).
290 *
291 * @see utf.h
292 * @see CharacterIterator
293 * @stable ICU 2.0
294 */
295class U_COMMON_API UnicodeString : public Replaceable
296{
297public:
298
299 /**
300 * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
301 * which constructs a Unicode string from an invariant-character char * string.
302 * Use the macro US_INV instead of the full qualification for this value.
303 *
304 * @see US_INV
305 * @stable ICU 3.2
306 */
307 enum EInvariant {
308 /**
309 * @see EInvariant
310 * @stable ICU 3.2
311 */
312 kInvariant
313 };
314
315 //========================================
316 // Read-only operations
317 //========================================
318
319 /* Comparison - bitwise only - for international comparison use collation */
320
321 /**
322 * Equality operator. Performs only bitwise comparison.
323 * @param text The UnicodeString to compare to this one.
324 * @return true if `text` contains the same characters as this one,
325 * false otherwise.
326 * @stable ICU 2.0
327 */
328 inline bool operator== (const UnicodeString& text) const;
329
330 /**
331 * Inequality operator. Performs only bitwise comparison.
332 * @param text The UnicodeString to compare to this one.
333 * @return false if `text` contains the same characters as this one,
334 * true otherwise.
335 * @stable ICU 2.0
336 */
337 inline bool operator!= (const UnicodeString& text) const;
338
339 /**
340 * Greater than operator. Performs only bitwise comparison.
341 * @param text The UnicodeString to compare to this one.
342 * @return true if the characters in this are bitwise
343 * greater than the characters in `text`, false otherwise
344 * @stable ICU 2.0
345 */
346 inline UBool operator> (const UnicodeString& text) const;
347
348 /**
349 * Less than operator. Performs only bitwise comparison.
350 * @param text The UnicodeString to compare to this one.
351 * @return true if the characters in this are bitwise
352 * less than the characters in `text`, false otherwise
353 * @stable ICU 2.0
354 */
355 inline UBool operator< (const UnicodeString& text) const;
356
357 /**
358 * Greater than or equal operator. Performs only bitwise comparison.
359 * @param text The UnicodeString to compare to this one.
360 * @return true if the characters in this are bitwise
361 * greater than or equal to the characters in `text`, false otherwise
362 * @stable ICU 2.0
363 */
364 inline UBool operator>= (const UnicodeString& text) const;
365
366 /**
367 * Less than or equal operator. Performs only bitwise comparison.
368 * @param text The UnicodeString to compare to this one.
369 * @return true if the characters in this are bitwise
370 * less than or equal to the characters in `text`, false otherwise
371 * @stable ICU 2.0
372 */
373 inline UBool operator<= (const UnicodeString& text) const;
374
375 /**
376 * Compare the characters bitwise in this UnicodeString to
377 * the characters in `text`.
378 * @param text The UnicodeString to compare to this one.
379 * @return The result of bitwise character comparison: 0 if this
380 * contains the same characters as `text`, -1 if the characters in
381 * this are bitwise less than the characters in `text`, +1 if the
382 * characters in this are bitwise greater than the characters
383 * in `text`.
384 * @stable ICU 2.0
385 */
386 inline int8_t compare(const UnicodeString& text) const;
387
388 /**
389 * Compare the characters bitwise in the range
390 * [`start`, `start + length`) with the characters
391 * in the **entire string** `text`.
392 * (The parameters "start" and "length" are not applied to the other text "text".)
393 * @param start the offset at which the compare operation begins
394 * @param length the number of characters of text to compare.
395 * @param text the other text to be compared against this string.
396 * @return The result of bitwise character comparison: 0 if this
397 * contains the same characters as `text`, -1 if the characters in
398 * this are bitwise less than the characters in `text`, +1 if the
399 * characters in this are bitwise greater than the characters
400 * in `text`.
401 * @stable ICU 2.0
402 */
403 inline int8_t compare(int32_t start,
404 int32_t length,
405 const UnicodeString& text) const;
406
407 /**
408 * Compare the characters bitwise in the range
409 * [`start`, `start + length`) with the characters
410 * in `srcText` in the range
411 * [`srcStart`, `srcStart + srcLength`).
412 * @param start the offset at which the compare operation begins
413 * @param length the number of characters in this to compare.
414 * @param srcText the text to be compared
415 * @param srcStart the offset into `srcText` to start comparison
416 * @param srcLength the number of characters in `src` to compare
417 * @return The result of bitwise character comparison: 0 if this
418 * contains the same characters as `srcText`, -1 if the characters in
419 * this are bitwise less than the characters in `srcText`, +1 if the
420 * characters in this are bitwise greater than the characters
421 * in `srcText`.
422 * @stable ICU 2.0
423 */
424 inline int8_t compare(int32_t start,
425 int32_t length,
426 const UnicodeString& srcText,
427 int32_t srcStart,
428 int32_t srcLength) const;
429
430 /**
431 * Compare the characters bitwise in this UnicodeString with the first
432 * `srcLength` characters in `srcChars`.
433 * @param srcChars The characters to compare to this UnicodeString.
434 * @param srcLength the number of characters in `srcChars` to compare
435 * @return The result of bitwise character comparison: 0 if this
436 * contains the same characters as `srcChars`, -1 if the characters in
437 * this are bitwise less than the characters in `srcChars`, +1 if the
438 * characters in this are bitwise greater than the characters
439 * in `srcChars`.
440 * @stable ICU 2.0
441 */
442 inline int8_t compare(ConstChar16Ptr srcChars,
443 int32_t srcLength) const;
444
445 /**
446 * Compare the characters bitwise in the range
447 * [`start`, `start + length`) with the first
448 * `length` characters in `srcChars`
449 * @param start the offset at which the compare operation begins
450 * @param length the number of characters to compare.
451 * @param srcChars the characters to be compared
452 * @return The result of bitwise character comparison: 0 if this
453 * contains the same characters as `srcChars`, -1 if the characters in
454 * this are bitwise less than the characters in `srcChars`, +1 if the
455 * characters in this are bitwise greater than the characters
456 * in `srcChars`.
457 * @stable ICU 2.0
458 */
459 inline int8_t compare(int32_t start,
460 int32_t length,
461 const char16_t *srcChars) const;
462
463 /**
464 * Compare the characters bitwise in the range
465 * [`start`, `start + length`) with the characters
466 * in `srcChars` in the range
467 * [`srcStart`, `srcStart + srcLength`).
468 * @param start the offset at which the compare operation begins
469 * @param length the number of characters in this to compare
470 * @param srcChars the characters to be compared
471 * @param srcStart the offset into `srcChars` to start comparison
472 * @param srcLength the number of characters in `srcChars` to compare
473 * @return The result of bitwise character comparison: 0 if this
474 * contains the same characters as `srcChars`, -1 if the characters in
475 * this are bitwise less than the characters in `srcChars`, +1 if the
476 * characters in this are bitwise greater than the characters
477 * in `srcChars`.
478 * @stable ICU 2.0
479 */
480 inline int8_t compare(int32_t start,
481 int32_t length,
482 const char16_t *srcChars,
483 int32_t srcStart,
484 int32_t srcLength) const;
485
486 /**
487 * Compare the characters bitwise in the range
488 * [`start`, `limit`) with the characters
489 * in `srcText` in the range
490 * [`srcStart`, `srcLimit`).
491 * @param start the offset at which the compare operation begins
492 * @param limit the offset immediately following the compare operation
493 * @param srcText the text to be compared
494 * @param srcStart the offset into `srcText` to start comparison
495 * @param srcLimit the offset into `srcText` to limit comparison
496 * @return The result of bitwise character comparison: 0 if this
497 * contains the same characters as `srcText`, -1 if the characters in
498 * this are bitwise less than the characters in `srcText`, +1 if the
499 * characters in this are bitwise greater than the characters
500 * in `srcText`.
501 * @stable ICU 2.0
502 */
503 inline int8_t compareBetween(int32_t start,
504 int32_t limit,
505 const UnicodeString& srcText,
506 int32_t srcStart,
507 int32_t srcLimit) const;
508
509 /**
510 * Compare two Unicode strings in code point order.
511 * The result may be different from the results of compare(), operator<, etc.
512 * if supplementary characters are present:
513 *
514 * In UTF-16, supplementary characters (with code points U+10000 and above) are
515 * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
516 * which means that they compare as less than some other BMP characters like U+feff.
517 * This function compares Unicode strings in code point order.
518 * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
519 *
520 * @param text Another string to compare this one to.
521 * @return a negative/zero/positive integer corresponding to whether
522 * this string is less than/equal to/greater than the second one
523 * in code point order
524 * @stable ICU 2.0
525 */
526 inline int8_t compareCodePointOrder(const UnicodeString& text) const;
527
528 /**
529 * Compare two Unicode strings in code point order.
530 * The result may be different from the results of compare(), operator<, etc.
531 * if supplementary characters are present:
532 *
533 * In UTF-16, supplementary characters (with code points U+10000 and above) are
534 * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
535 * which means that they compare as less than some other BMP characters like U+feff.
536 * This function compares Unicode strings in code point order.
537 * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
538 *
539 * @param start The start offset in this string at which the compare operation begins.
540 * @param length The number of code units from this string to compare.
541 * @param srcText Another string to compare this one to.
542 * @return a negative/zero/positive integer corresponding to whether
543 * this string is less than/equal to/greater than the second one
544 * in code point order
545 * @stable ICU 2.0
546 */
547 inline int8_t compareCodePointOrder(int32_t start,
548 int32_t length,
549 const UnicodeString& srcText) const;
550
551 /**
552 * Compare two Unicode strings in code point order.
553 * The result may be different from the results of compare(), operator<, etc.
554 * if supplementary characters are present:
555 *
556 * In UTF-16, supplementary characters (with code points U+10000 and above) are
557 * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
558 * which means that they compare as less than some other BMP characters like U+feff.
559 * This function compares Unicode strings in code point order.
560 * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
561 *
562 * @param start The start offset in this string at which the compare operation begins.
563 * @param length The number of code units from this string to compare.
564 * @param srcText Another string to compare this one to.
565 * @param srcStart The start offset in that string at which the compare operation begins.
566 * @param srcLength The number of code units from that string to compare.
567 * @return a negative/zero/positive integer corresponding to whether
568 * this string is less than/equal to/greater than the second one
569 * in code point order
570 * @stable ICU 2.0
571 */
572 inline int8_t compareCodePointOrder(int32_t start,
573 int32_t length,
574 const UnicodeString& srcText,
575 int32_t srcStart,
576 int32_t srcLength) const;
577
578 /**
579 * Compare two Unicode strings in code point order.
580 * The result may be different from the results of compare(), operator<, etc.
581 * if supplementary characters are present:
582 *
583 * In UTF-16, supplementary characters (with code points U+10000 and above) are
584 * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
585 * which means that they compare as less than some other BMP characters like U+feff.
586 * This function compares Unicode strings in code point order.
587 * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
588 *
589 * @param srcChars A pointer to another string to compare this one to.
590 * @param srcLength The number of code units from that string to compare.
591 * @return a negative/zero/positive integer corresponding to whether
592 * this string is less than/equal to/greater than the second one
593 * in code point order
594 * @stable ICU 2.0
595 */
596 inline int8_t compareCodePointOrder(ConstChar16Ptr srcChars,
597 int32_t srcLength) const;
598
599 /**
600 * Compare two Unicode strings in code point order.
601 * The result may be different from the results of compare(), operator<, etc.
602 * if supplementary characters are present:
603 *
604 * In UTF-16, supplementary characters (with code points U+10000 and above) are
605 * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
606 * which means that they compare as less than some other BMP characters like U+feff.
607 * This function compares Unicode strings in code point order.
608 * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
609 *
610 * @param start The start offset in this string at which the compare operation begins.
611 * @param length The number of code units from this string to compare.
612 * @param srcChars A pointer to another string to compare this one to.
613 * @return a negative/zero/positive integer corresponding to whether
614 * this string is less than/equal to/greater than the second one
615 * in code point order
616 * @stable ICU 2.0
617 */
618 inline int8_t compareCodePointOrder(int32_t start,
619 int32_t length,
620 const char16_t *srcChars) const;
621
622 /**
623 * Compare two Unicode strings in code point order.
624 * The result may be different from the results of compare(), operator<, etc.
625 * if supplementary characters are present:
626 *
627 * In UTF-16, supplementary characters (with code points U+10000 and above) are
628 * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
629 * which means that they compare as less than some other BMP characters like U+feff.
630 * This function compares Unicode strings in code point order.
631 * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
632 *
633 * @param start The start offset in this string at which the compare operation begins.
634 * @param length The number of code units from this string to compare.
635 * @param srcChars A pointer to another string to compare this one to.
636 * @param srcStart The start offset in that string at which the compare operation begins.
637 * @param srcLength The number of code units from that string to compare.
638 * @return a negative/zero/positive integer corresponding to whether
639 * this string is less than/equal to/greater than the second one
640 * in code point order
641 * @stable ICU 2.0
642 */
643 inline int8_t compareCodePointOrder(int32_t start,
644 int32_t length,
645 const char16_t *srcChars,
646 int32_t srcStart,
647 int32_t srcLength) const;
648
649 /**
650 * Compare two Unicode strings in code point order.
651 * The result may be different from the results of compare(), operator<, etc.
652 * if supplementary characters are present:
653 *
654 * In UTF-16, supplementary characters (with code points U+10000 and above) are
655 * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
656 * which means that they compare as less than some other BMP characters like U+feff.
657 * This function compares Unicode strings in code point order.
658 * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
659 *
660 * @param start The start offset in this string at which the compare operation begins.
661 * @param limit The offset after the last code unit from this string to compare.
662 * @param srcText Another string to compare this one to.
663 * @param srcStart The start offset in that string at which the compare operation begins.
664 * @param srcLimit The offset after the last code unit from that string to compare.
665 * @return a negative/zero/positive integer corresponding to whether
666 * this string is less than/equal to/greater than the second one
667 * in code point order
668 * @stable ICU 2.0
669 */
670 inline int8_t compareCodePointOrderBetween(int32_t start,
671 int32_t limit,
672 const UnicodeString& srcText,
673 int32_t srcStart,
674 int32_t srcLimit) const;
675
676 /**
677 * Compare two strings case-insensitively using full case folding.
678 * This is equivalent to this->foldCase(options).compare(text.foldCase(options)).
679 *
680 * @param text Another string to compare this one to.
681 * @param options A bit set of options:
682 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
683 * Comparison in code unit order with default case folding.
684 *
685 * - U_COMPARE_CODE_POINT_ORDER
686 * Set to choose code point order instead of code unit order
687 * (see u_strCompare for details).
688 *
689 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
690 *
691 * @return A negative, zero, or positive integer indicating the comparison result.
692 * @stable ICU 2.0
693 */
694 inline int8_t caseCompare(const UnicodeString& text, uint32_t options) const;
695
696 /**
697 * Compare two strings case-insensitively using full case folding.
698 * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
699 *
700 * @param start The start offset in this string at which the compare operation begins.
701 * @param length The number of code units from this string to compare.
702 * @param srcText Another string to compare this one to.
703 * @param options A bit set of options:
704 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
705 * Comparison in code unit order with default case folding.
706 *
707 * - U_COMPARE_CODE_POINT_ORDER
708 * Set to choose code point order instead of code unit order
709 * (see u_strCompare for details).
710 *
711 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
712 *
713 * @return A negative, zero, or positive integer indicating the comparison result.
714 * @stable ICU 2.0
715 */
716 inline int8_t caseCompare(int32_t start,
717 int32_t length,
718 const UnicodeString& srcText,
719 uint32_t options) const;
720
721 /**
722 * Compare two strings case-insensitively using full case folding.
723 * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
724 *
725 * @param start The start offset in this string at which the compare operation begins.
726 * @param length The number of code units from this string to compare.
727 * @param srcText Another string to compare this one to.
728 * @param srcStart The start offset in that string at which the compare operation begins.
729 * @param srcLength The number of code units from that string to compare.
730 * @param options A bit set of options:
731 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
732 * Comparison in code unit order with default case folding.
733 *
734 * - U_COMPARE_CODE_POINT_ORDER
735 * Set to choose code point order instead of code unit order
736 * (see u_strCompare for details).
737 *
738 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
739 *
740 * @return A negative, zero, or positive integer indicating the comparison result.
741 * @stable ICU 2.0
742 */
743 inline int8_t caseCompare(int32_t start,
744 int32_t length,
745 const UnicodeString& srcText,
746 int32_t srcStart,
747 int32_t srcLength,
748 uint32_t options) const;
749
750 /**
751 * Compare two strings case-insensitively using full case folding.
752 * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
753 *
754 * @param srcChars A pointer to another string to compare this one to.
755 * @param srcLength The number of code units from that string to compare.
756 * @param options A bit set of options:
757 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
758 * Comparison in code unit order with default case folding.
759 *
760 * - U_COMPARE_CODE_POINT_ORDER
761 * Set to choose code point order instead of code unit order
762 * (see u_strCompare for details).
763 *
764 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
765 *
766 * @return A negative, zero, or positive integer indicating the comparison result.
767 * @stable ICU 2.0
768 */
769 inline int8_t caseCompare(ConstChar16Ptr srcChars,
770 int32_t srcLength,
771 uint32_t options) const;
772
773 /**
774 * Compare two strings case-insensitively using full case folding.
775 * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
776 *
777 * @param start The start offset in this string at which the compare operation begins.
778 * @param length The number of code units from this string to compare.
779 * @param srcChars A pointer to another string to compare this one to.
780 * @param options A bit set of options:
781 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
782 * Comparison in code unit order with default case folding.
783 *
784 * - U_COMPARE_CODE_POINT_ORDER
785 * Set to choose code point order instead of code unit order
786 * (see u_strCompare for details).
787 *
788 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
789 *
790 * @return A negative, zero, or positive integer indicating the comparison result.
791 * @stable ICU 2.0
792 */
793 inline int8_t caseCompare(int32_t start,
794 int32_t length,
795 const char16_t *srcChars,
796 uint32_t options) const;
797
798 /**
799 * Compare two strings case-insensitively using full case folding.
800 * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
801 *
802 * @param start The start offset in this string at which the compare operation begins.
803 * @param length The number of code units from this string to compare.
804 * @param srcChars A pointer to another string to compare this one to.
805 * @param srcStart The start offset in that string at which the compare operation begins.
806 * @param srcLength The number of code units from that string to compare.
807 * @param options A bit set of options:
808 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
809 * Comparison in code unit order with default case folding.
810 *
811 * - U_COMPARE_CODE_POINT_ORDER
812 * Set to choose code point order instead of code unit order
813 * (see u_strCompare for details).
814 *
815 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
816 *
817 * @return A negative, zero, or positive integer indicating the comparison result.
818 * @stable ICU 2.0
819 */
820 inline int8_t caseCompare(int32_t start,
821 int32_t length,
822 const char16_t *srcChars,
823 int32_t srcStart,
824 int32_t srcLength,
825 uint32_t options) const;
826
827 /**
828 * Compare two strings case-insensitively using full case folding.
829 * This is equivalent to this->foldCase(options).compareBetween(text.foldCase(options)).
830 *
831 * @param start The start offset in this string at which the compare operation begins.
832 * @param limit The offset after the last code unit from this string to compare.
833 * @param srcText Another string to compare this one to.
834 * @param srcStart The start offset in that string at which the compare operation begins.
835 * @param srcLimit The offset after the last code unit from that string to compare.
836 * @param options A bit set of options:
837 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
838 * Comparison in code unit order with default case folding.
839 *
840 * - U_COMPARE_CODE_POINT_ORDER
841 * Set to choose code point order instead of code unit order
842 * (see u_strCompare for details).
843 *
844 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
845 *
846 * @return A negative, zero, or positive integer indicating the comparison result.
847 * @stable ICU 2.0
848 */
849 inline int8_t caseCompareBetween(int32_t start,
850 int32_t limit,
851 const UnicodeString& srcText,
852 int32_t srcStart,
853 int32_t srcLimit,
854 uint32_t options) const;
855
856 /**
857 * Determine if this starts with the characters in `text`
858 * @param text The text to match.
859 * @return true if this starts with the characters in `text`,
860 * false otherwise
861 * @stable ICU 2.0
862 */
863 inline UBool startsWith(const UnicodeString& text) const;
864
865 /**
866 * Determine if this starts with the characters in `srcText`
867 * in the range [`srcStart`, `srcStart + srcLength`).
868 * @param srcText The text to match.
869 * @param srcStart the offset into `srcText` to start matching
870 * @param srcLength the number of characters in `srcText` to match
871 * @return true if this starts with the characters in `text`,
872 * false otherwise
873 * @stable ICU 2.0
874 */
875 inline UBool startsWith(const UnicodeString& srcText,
876 int32_t srcStart,
877 int32_t srcLength) const;
878
879 /**
880 * Determine if this starts with the characters in `srcChars`
881 * @param srcChars The characters to match.
882 * @param srcLength the number of characters in `srcChars`
883 * @return true if this starts with the characters in `srcChars`,
884 * false otherwise
885 * @stable ICU 2.0
886 */
887 inline UBool startsWith(ConstChar16Ptr srcChars,
888 int32_t srcLength) const;
889
890 /**
891 * Determine if this ends with the characters in `srcChars`
892 * in the range [`srcStart`, `srcStart + srcLength`).
893 * @param srcChars The characters to match.
894 * @param srcStart the offset into `srcText` to start matching
895 * @param srcLength the number of characters in `srcChars` to match
896 * @return true if this ends with the characters in `srcChars`, false otherwise
897 * @stable ICU 2.0
898 */
899 inline UBool startsWith(const char16_t *srcChars,
900 int32_t srcStart,
901 int32_t srcLength) const;
902
903 /**
904 * Determine if this ends with the characters in `text`
905 * @param text The text to match.
906 * @return true if this ends with the characters in `text`,
907 * false otherwise
908 * @stable ICU 2.0
909 */
910 inline UBool endsWith(const UnicodeString& text) const;
911
912 /**
913 * Determine if this ends with the characters in `srcText`
914 * in the range [`srcStart`, `srcStart + srcLength`).
915 * @param srcText The text to match.
916 * @param srcStart the offset into `srcText` to start matching
917 * @param srcLength the number of characters in `srcText` to match
918 * @return true if this ends with the characters in `text`,
919 * false otherwise
920 * @stable ICU 2.0
921 */
922 inline UBool endsWith(const UnicodeString& srcText,
923 int32_t srcStart,
924 int32_t srcLength) const;
925
926 /**
927 * Determine if this ends with the characters in `srcChars`
928 * @param srcChars The characters to match.
929 * @param srcLength the number of characters in `srcChars`
930 * @return true if this ends with the characters in `srcChars`,
931 * false otherwise
932 * @stable ICU 2.0
933 */
934 inline UBool endsWith(ConstChar16Ptr srcChars,
935 int32_t srcLength) const;
936
937 /**
938 * Determine if this ends with the characters in `srcChars`
939 * in the range [`srcStart`, `srcStart + srcLength`).
940 * @param srcChars The characters to match.
941 * @param srcStart the offset into `srcText` to start matching
942 * @param srcLength the number of characters in `srcChars` to match
943 * @return true if this ends with the characters in `srcChars`,
944 * false otherwise
945 * @stable ICU 2.0
946 */
947 inline UBool endsWith(const char16_t *srcChars,
948 int32_t srcStart,
949 int32_t srcLength) const;
950
951
952 /* Searching - bitwise only */
953
954 /**
955 * Locate in this the first occurrence of the characters in `text`,
956 * using bitwise comparison.
957 * @param text The text to search for.
958 * @return The offset into this of the start of `text`,
959 * or -1 if not found.
960 * @stable ICU 2.0
961 */
962 inline int32_t indexOf(const UnicodeString& text) const;
963
964 /**
965 * Locate in this the first occurrence of the characters in `text`
966 * starting at offset `start`, using bitwise comparison.
967 * @param text The text to search for.
968 * @param start The offset at which searching will start.
969 * @return The offset into this of the start of `text`,
970 * or -1 if not found.
971 * @stable ICU 2.0
972 */
973 inline int32_t indexOf(const UnicodeString& text,
974 int32_t start) const;
975
976 /**
977 * Locate in this the first occurrence in the range
978 * [`start`, `start + length`) of the characters
979 * in `text`, using bitwise comparison.
980 * @param text The text to search for.
981 * @param start The offset at which searching will start.
982 * @param length The number of characters to search
983 * @return The offset into this of the start of `text`,
984 * or -1 if not found.
985 * @stable ICU 2.0
986 */
987 inline int32_t indexOf(const UnicodeString& text,
988 int32_t start,
989 int32_t length) const;
990
991 /**
992 * Locate in this the first occurrence in the range
993 * [`start`, `start + length`) of the characters
994 * in `srcText` in the range
995 * [`srcStart`, `srcStart + srcLength`),
996 * using bitwise comparison.
997 * @param srcText The text to search for.
998 * @param srcStart the offset into `srcText` at which
999 * to start matching
1000 * @param srcLength the number of characters in `srcText` to match
1001 * @param start the offset into this at which to start matching
1002 * @param length the number of characters in this to search
1003 * @return The offset into this of the start of `text`,
1004 * or -1 if not found.
1005 * @stable ICU 2.0
1006 */
1007 inline int32_t indexOf(const UnicodeString& srcText,
1008 int32_t srcStart,
1009 int32_t srcLength,
1010 int32_t start,
1011 int32_t length) const;
1012
1013 /**
1014 * Locate in this the first occurrence of the characters in
1015 * `srcChars`
1016 * starting at offset `start`, using bitwise comparison.
1017 * @param srcChars The text to search for.
1018 * @param srcLength the number of characters in `srcChars` to match
1019 * @param start the offset into this at which to start matching
1020 * @return The offset into this of the start of `text`,
1021 * or -1 if not found.
1022 * @stable ICU 2.0
1023 */
1024 inline int32_t indexOf(const char16_t *srcChars,
1025 int32_t srcLength,
1026 int32_t start) const;
1027
1028 /**
1029 * Locate in this the first occurrence in the range
1030 * [`start`, `start + length`) of the characters
1031 * in `srcChars`, using bitwise comparison.
1032 * @param srcChars The text to search for.
1033 * @param srcLength the number of characters in `srcChars`
1034 * @param start The offset at which searching will start.
1035 * @param length The number of characters to search
1036 * @return The offset into this of the start of `srcChars`,
1037 * or -1 if not found.
1038 * @stable ICU 2.0
1039 */
1040 inline int32_t indexOf(ConstChar16Ptr srcChars,
1041 int32_t srcLength,
1042 int32_t start,
1043 int32_t length) const;
1044
1045 /**
1046 * Locate in this the first occurrence in the range
1047 * [`start`, `start + length`) of the characters
1048 * in `srcChars` in the range
1049 * [`srcStart`, `srcStart + srcLength`),
1050 * using bitwise comparison.
1051 * @param srcChars The text to search for.
1052 * @param srcStart the offset into `srcChars` at which
1053 * to start matching
1054 * @param srcLength the number of characters in `srcChars` to match
1055 * @param start the offset into this at which to start matching
1056 * @param length the number of characters in this to search
1057 * @return The offset into this of the start of `text`,
1058 * or -1 if not found.
1059 * @stable ICU 2.0
1060 */
1061 int32_t indexOf(const char16_t *srcChars,
1062 int32_t srcStart,
1063 int32_t srcLength,
1064 int32_t start,
1065 int32_t length) const;
1066
1067 /**
1068 * Locate in this the first occurrence of the BMP code point `c`,
1069 * using bitwise comparison.
1070 * @param c The code unit to search for.
1071 * @return The offset into this of `c`, or -1 if not found.
1072 * @stable ICU 2.0
1073 */
1074 inline int32_t indexOf(char16_t c) const;
1075
1076 /**
1077 * Locate in this the first occurrence of the code point `c`,
1078 * using bitwise comparison.
1079 *
1080 * @param c The code point to search for.
1081 * @return The offset into this of `c`, or -1 if not found.
1082 * @stable ICU 2.0
1083 */
1084 inline int32_t indexOf(UChar32 c) const;
1085
1086 /**
1087 * Locate in this the first occurrence of the BMP code point `c`,
1088 * starting at offset `start`, using bitwise comparison.
1089 * @param c The code unit to search for.
1090 * @param start The offset at which searching will start.
1091 * @return The offset into this of `c`, or -1 if not found.
1092 * @stable ICU 2.0
1093 */
1094 inline int32_t indexOf(char16_t c,
1095 int32_t start) const;
1096
1097 /**
1098 * Locate in this the first occurrence of the code point `c`
1099 * starting at offset `start`, using bitwise comparison.
1100 *
1101 * @param c The code point to search for.
1102 * @param start The offset at which searching will start.
1103 * @return The offset into this of `c`, or -1 if not found.
1104 * @stable ICU 2.0
1105 */
1106 inline int32_t indexOf(UChar32 c,
1107 int32_t start) const;
1108
1109 /**
1110 * Locate in this the first occurrence of the BMP code point `c`
1111 * in the range [`start`, `start + length`),
1112 * using bitwise comparison.
1113 * @param c The code unit to search for.
1114 * @param start the offset into this at which to start matching
1115 * @param length the number of characters in this to search
1116 * @return The offset into this of `c`, or -1 if not found.
1117 * @stable ICU 2.0
1118 */
1119 inline int32_t indexOf(char16_t c,
1120 int32_t start,
1121 int32_t length) const;
1122
1123 /**
1124 * Locate in this the first occurrence of the code point `c`
1125 * in the range [`start`, `start + length`),
1126 * using bitwise comparison.
1127 *
1128 * @param c The code point to search for.
1129 * @param start the offset into this at which to start matching
1130 * @param length the number of characters in this to search
1131 * @return The offset into this of `c`, or -1 if not found.
1132 * @stable ICU 2.0
1133 */
1134 inline int32_t indexOf(UChar32 c,
1135 int32_t start,
1136 int32_t length) const;
1137
1138 /**
1139 * Locate in this the last occurrence of the characters in `text`,
1140 * using bitwise comparison.
1141 * @param text The text to search for.
1142 * @return The offset into this of the start of `text`,
1143 * or -1 if not found.
1144 * @stable ICU 2.0
1145 */
1146 inline int32_t lastIndexOf(const UnicodeString& text) const;
1147
1148 /**
1149 * Locate in this the last occurrence of the characters in `text`
1150 * starting at offset `start`, using bitwise comparison.
1151 * @param text The text to search for.
1152 * @param start The offset at which searching will start.
1153 * @return The offset into this of the start of `text`,
1154 * or -1 if not found.
1155 * @stable ICU 2.0
1156 */
1157 inline int32_t lastIndexOf(const UnicodeString& text,
1158 int32_t start) const;
1159
1160 /**
1161 * Locate in this the last occurrence in the range
1162 * [`start`, `start + length`) of the characters
1163 * in `text`, using bitwise comparison.
1164 * @param text The text to search for.
1165 * @param start The offset at which searching will start.
1166 * @param length The number of characters to search
1167 * @return The offset into this of the start of `text`,
1168 * or -1 if not found.
1169 * @stable ICU 2.0
1170 */
1171 inline int32_t lastIndexOf(const UnicodeString& text,
1172 int32_t start,
1173 int32_t length) const;
1174
1175 /**
1176 * Locate in this the last occurrence in the range
1177 * [`start`, `start + length`) of the characters
1178 * in `srcText` in the range
1179 * [`srcStart`, `srcStart + srcLength`),
1180 * using bitwise comparison.
1181 * @param srcText The text to search for.
1182 * @param srcStart the offset into `srcText` at which
1183 * to start matching
1184 * @param srcLength the number of characters in `srcText` to match
1185 * @param start the offset into this at which to start matching
1186 * @param length the number of characters in this to search
1187 * @return The offset into this of the start of `text`,
1188 * or -1 if not found.
1189 * @stable ICU 2.0
1190 */
1191 inline int32_t lastIndexOf(const UnicodeString& srcText,
1192 int32_t srcStart,
1193 int32_t srcLength,
1194 int32_t start,
1195 int32_t length) const;
1196
1197 /**
1198 * Locate in this the last occurrence of the characters in `srcChars`
1199 * starting at offset `start`, using bitwise comparison.
1200 * @param srcChars The text to search for.
1201 * @param srcLength the number of characters in `srcChars` to match
1202 * @param start the offset into this at which to start matching
1203 * @return The offset into this of the start of `text`,
1204 * or -1 if not found.
1205 * @stable ICU 2.0
1206 */
1207 inline int32_t lastIndexOf(const char16_t *srcChars,
1208 int32_t srcLength,
1209 int32_t start) const;
1210
1211 /**
1212 * Locate in this the last occurrence in the range
1213 * [`start`, `start + length`) of the characters
1214 * in `srcChars`, using bitwise comparison.
1215 * @param srcChars The text to search for.
1216 * @param srcLength the number of characters in `srcChars`
1217 * @param start The offset at which searching will start.
1218 * @param length The number of characters to search
1219 * @return The offset into this of the start of `srcChars`,
1220 * or -1 if not found.
1221 * @stable ICU 2.0
1222 */
1223 inline int32_t lastIndexOf(ConstChar16Ptr srcChars,
1224 int32_t srcLength,
1225 int32_t start,
1226 int32_t length) const;
1227
1228 /**
1229 * Locate in this the last occurrence in the range
1230 * [`start`, `start + length`) of the characters
1231 * in `srcChars` in the range
1232 * [`srcStart`, `srcStart + srcLength`),
1233 * using bitwise comparison.
1234 * @param srcChars The text to search for.
1235 * @param srcStart the offset into `srcChars` at which
1236 * to start matching
1237 * @param srcLength the number of characters in `srcChars` to match
1238 * @param start the offset into this at which to start matching
1239 * @param length the number of characters in this to search
1240 * @return The offset into this of the start of `text`,
1241 * or -1 if not found.
1242 * @stable ICU 2.0
1243 */
1244 int32_t lastIndexOf(const char16_t *srcChars,
1245 int32_t srcStart,
1246 int32_t srcLength,
1247 int32_t start,
1248 int32_t length) const;
1249
1250 /**
1251 * Locate in this the last occurrence of the BMP code point `c`,
1252 * using bitwise comparison.
1253 * @param c The code unit to search for.
1254 * @return The offset into this of `c`, or -1 if not found.
1255 * @stable ICU 2.0
1256 */
1257 inline int32_t lastIndexOf(char16_t c) const;
1258
1259 /**
1260 * Locate in this the last occurrence of the code point `c`,
1261 * using bitwise comparison.
1262 *
1263 * @param c The code point to search for.
1264 * @return The offset into this of `c`, or -1 if not found.
1265 * @stable ICU 2.0
1266 */
1267 inline int32_t lastIndexOf(UChar32 c) const;
1268
1269 /**
1270 * Locate in this the last occurrence of the BMP code point `c`
1271 * starting at offset `start`, using bitwise comparison.
1272 * @param c The code unit to search for.
1273 * @param start The offset at which searching will start.
1274 * @return The offset into this of `c`, or -1 if not found.
1275 * @stable ICU 2.0
1276 */
1277 inline int32_t lastIndexOf(char16_t c,
1278 int32_t start) const;
1279
1280 /**
1281 * Locate in this the last occurrence of the code point `c`
1282 * starting at offset `start`, using bitwise comparison.
1283 *
1284 * @param c The code point to search for.
1285 * @param start The offset at which searching will start.
1286 * @return The offset into this of `c`, or -1 if not found.
1287 * @stable ICU 2.0
1288 */
1289 inline int32_t lastIndexOf(UChar32 c,
1290 int32_t start) const;
1291
1292 /**
1293 * Locate in this the last occurrence of the BMP code point `c`
1294 * in the range [`start`, `start + length`),
1295 * using bitwise comparison.
1296 * @param c The code unit to search for.
1297 * @param start the offset into this at which to start matching
1298 * @param length the number of characters in this to search
1299 * @return The offset into this of `c`, or -1 if not found.
1300 * @stable ICU 2.0
1301 */
1302 inline int32_t lastIndexOf(char16_t c,
1303 int32_t start,
1304 int32_t length) const;
1305
1306 /**
1307 * Locate in this the last occurrence of the code point `c`
1308 * in the range [`start`, `start + length`),
1309 * using bitwise comparison.
1310 *
1311 * @param c The code point to search for.
1312 * @param start the offset into this at which to start matching
1313 * @param length the number of characters in this to search
1314 * @return The offset into this of `c`, or -1 if not found.
1315 * @stable ICU 2.0
1316 */
1317 inline int32_t lastIndexOf(UChar32 c,
1318 int32_t start,
1319 int32_t length) const;
1320
1321
1322 /* Character access */
1323
1324 /**
1325 * Return the code unit at offset `offset`.
1326 * If the offset is not valid (0..length()-1) then U+ffff is returned.
1327 * @param offset a valid offset into the text
1328 * @return the code unit at offset `offset`
1329 * or 0xffff if the offset is not valid for this string
1330 * @stable ICU 2.0
1331 */
1332 inline char16_t charAt(int32_t offset) const;
1333
1334 /**
1335 * Return the code unit at offset `offset`.
1336 * If the offset is not valid (0..length()-1) then U+ffff is returned.
1337 * @param offset a valid offset into the text
1338 * @return the code unit at offset `offset`
1339 * @stable ICU 2.0
1340 */
1341 inline char16_t operator[] (int32_t offset) const;
1342
1343 /**
1344 * Return the code point that contains the code unit
1345 * at offset `offset`.
1346 * If the offset is not valid (0..length()-1) then U+ffff is returned.
1347 * @param offset a valid offset into the text
1348 * that indicates the text offset of any of the code units
1349 * that will be assembled into a code point (21-bit value) and returned
1350 * @return the code point of text at `offset`
1351 * or 0xffff if the offset is not valid for this string
1352 * @stable ICU 2.0
1353 */
1354 UChar32 char32At(int32_t offset) const;
1355
1356 /**
1357 * Adjust a random-access offset so that
1358 * it points to the beginning of a Unicode character.
1359 * The offset that is passed in points to
1360 * any code unit of a code point,
1361 * while the returned offset will point to the first code unit
1362 * of the same code point.
1363 * In UTF-16, if the input offset points to a second surrogate
1364 * of a surrogate pair, then the returned offset will point
1365 * to the first surrogate.
1366 * @param offset a valid offset into one code point of the text
1367 * @return offset of the first code unit of the same code point
1368 * @see U16_SET_CP_START
1369 * @stable ICU 2.0
1370 */
1371 int32_t getChar32Start(int32_t offset) const;
1372
1373 /**
1374 * Adjust a random-access offset so that
1375 * it points behind a Unicode character.
1376 * The offset that is passed in points behind
1377 * any code unit of a code point,
1378 * while the returned offset will point behind the last code unit
1379 * of the same code point.
1380 * In UTF-16, if the input offset points behind the first surrogate
1381 * (i.e., to the second surrogate)
1382 * of a surrogate pair, then the returned offset will point
1383 * behind the second surrogate (i.e., to the first surrogate).
1384 * @param offset a valid offset after any code unit of a code point of the text
1385 * @return offset of the first code unit after the same code point
1386 * @see U16_SET_CP_LIMIT
1387 * @stable ICU 2.0
1388 */
1389 int32_t getChar32Limit(int32_t offset) const;
1390
1391 /**
1392 * Move the code unit index along the string by delta code points.
1393 * Interpret the input index as a code unit-based offset into the string,
1394 * move the index forward or backward by delta code points, and
1395 * return the resulting index.
1396 * The input index should point to the first code unit of a code point,
1397 * if there is more than one.
1398 *
1399 * Both input and output indexes are code unit-based as for all
1400 * string indexes/offsets in ICU (and other libraries, like MBCS char*).
1401 * If delta<0 then the index is moved backward (toward the start of the string).
1402 * If delta>0 then the index is moved forward (toward the end of the string).
1403 *
1404 * This behaves like CharacterIterator::move32(delta, kCurrent).
1405 *
1406 * Behavior for out-of-bounds indexes:
1407 * `moveIndex32` pins the input index to 0..length(), i.e.,
1408 * if the input index<0 then it is pinned to 0;
1409 * if it is index>length() then it is pinned to length().
1410 * Afterwards, the index is moved by `delta` code points
1411 * forward or backward,
1412 * but no further backward than to 0 and no further forward than to length().
1413 * The resulting index return value will be in between 0 and length(), inclusively.
1414 *
1415 * Examples:
1416 * \code
1417 * // s has code points 'a' U+10000 'b' U+10ffff U+2029
1418 * UnicodeString s(u"a\U00010000b\U0010ffff\u2029");
1419 *
1420 * // initial index: position of U+10000
1421 * int32_t index=1;
1422 *
1423 * // the following examples will all result in index==4, position of U+10ffff
1424 *
1425 * // skip 2 code points from some position in the string
1426 * index=s.moveIndex32(index, 2); // skips U+10000 and 'b'
1427 *
1428 * // go to the 3rd code point from the start of s (0-based)
1429 * index=s.moveIndex32(0, 3); // skips 'a', U+10000, and 'b'
1430 *
1431 * // go to the next-to-last code point of s
1432 * index=s.moveIndex32(s.length(), -2); // backward-skips U+2029 and U+10ffff
1433 * \endcode
1434 *
1435 * @param index input code unit index
1436 * @param delta (signed) code point count to move the index forward or backward
1437 * in the string
1438 * @return the resulting code unit index
1439 * @stable ICU 2.0
1440 */
1441 int32_t moveIndex32(int32_t index, int32_t delta) const;
1442
1443 /* Substring extraction */
1444
1445 /**
1446 * Copy the characters in the range
1447 * [`start`, `start + length`) into the array `dst`,
1448 * beginning at `dstStart`.
1449 * If the string aliases to `dst` itself as an external buffer,
1450 * then extract() will not copy the contents.
1451 *
1452 * @param start offset of first character which will be copied into the array
1453 * @param length the number of characters to extract
1454 * @param dst array in which to copy characters. The length of `dst`
1455 * must be at least (`dstStart + length`).
1456 * @param dstStart the offset in `dst` where the first character
1457 * will be extracted
1458 * @stable ICU 2.0
1459 */
1460 inline void extract(int32_t start,
1461 int32_t length,
1462 Char16Ptr dst,
1463 int32_t dstStart = 0) const;
1464
1465 /**
1466 * Copy the contents of the string into dest.
1467 * This is a convenience function that
1468 * checks if there is enough space in dest,
1469 * extracts the entire string if possible,
1470 * and NUL-terminates dest if possible.
1471 *
1472 * If the string fits into dest but cannot be NUL-terminated
1473 * (length()==destCapacity) then the error code is set to U_STRING_NOT_TERMINATED_WARNING.
1474 * If the string itself does not fit into dest
1475 * (length()>destCapacity) then the error code is set to U_BUFFER_OVERFLOW_ERROR.
1476 *
1477 * If the string aliases to `dest` itself as an external buffer,
1478 * then extract() will not copy the contents.
1479 *
1480 * @param dest Destination string buffer.
1481 * @param destCapacity Number of char16_ts available at dest.
1482 * @param errorCode ICU error code.
1483 * @return length()
1484 * @stable ICU 2.0
1485 */
1486 int32_t
1487 extract(Char16Ptr dest, int32_t destCapacity,
1488 UErrorCode &errorCode) const;
1489
1490 /**
1491 * Copy the characters in the range
1492 * [`start`, `start + length`) into the UnicodeString
1493 * `target`.
1494 * @param start offset of first character which will be copied
1495 * @param length the number of characters to extract
1496 * @param target UnicodeString into which to copy characters.
1497 * @stable ICU 2.0
1498 */
1499 inline void extract(int32_t start,
1500 int32_t length,
1501 UnicodeString& target) const;
1502
1503 /**
1504 * Copy the characters in the range [`start`, `limit`)
1505 * into the array `dst`, beginning at `dstStart`.
1506 * @param start offset of first character which will be copied into the array
1507 * @param limit offset immediately following the last character to be copied
1508 * @param dst array in which to copy characters. The length of `dst`
1509 * must be at least (`dstStart + (limit - start)`).
1510 * @param dstStart the offset in `dst` where the first character
1511 * will be extracted
1512 * @stable ICU 2.0
1513 */
1514 inline void extractBetween(int32_t start,
1515 int32_t limit,
1516 char16_t *dst,
1517 int32_t dstStart = 0) const;
1518
1519 /**
1520 * Copy the characters in the range [`start`, `limit`)
1521 * into the UnicodeString `target`. Replaceable API.
1522 * @param start offset of first character which will be copied
1523 * @param limit offset immediately following the last character to be copied
1524 * @param target UnicodeString into which to copy characters.
1525 * @stable ICU 2.0
1526 */
1527 virtual void extractBetween(int32_t start,
1528 int32_t limit,
1529 UnicodeString& target) const override;
1530
1531 /**
1532 * Copy the characters in the range
1533 * [`start`, `start + startLength`) into an array of characters.
1534 * All characters must be invariant (see utypes.h).
1535 * Use US_INV as the last, signature-distinguishing parameter.
1536 *
1537 * This function does not write any more than `targetCapacity`
1538 * characters but returns the length of the entire output string
1539 * so that one can allocate a larger buffer and call the function again
1540 * if necessary.
1541 * The output string is NUL-terminated if possible.
1542 *
1543 * @param start offset of first character which will be copied
1544 * @param startLength the number of characters to extract
1545 * @param target the target buffer for extraction, can be NULL
1546 * if targetLength is 0
1547 * @param targetCapacity the length of the target buffer
1548 * @param inv Signature-distinguishing parameter, use US_INV.
1549 * @return the output string length, not including the terminating NUL
1550 * @stable ICU 3.2
1551 */
1552 int32_t extract(int32_t start,
1553 int32_t startLength,
1554 char *target,
1555 int32_t targetCapacity,
1556 enum EInvariant inv) const;
1557
1558#if U_CHARSET_IS_UTF81 || !UCONFIG_NO_CONVERSION0
1559
1560 /**
1561 * Copy the characters in the range
1562 * [`start`, `start + length`) into an array of characters
1563 * in the platform's default codepage.
1564 * This function does not write any more than `targetLength`
1565 * characters but returns the length of the entire output string
1566 * so that one can allocate a larger buffer and call the function again
1567 * if necessary.
1568 * The output string is NUL-terminated if possible.
1569 *
1570 * @param start offset of first character which will be copied
1571 * @param startLength the number of characters to extract
1572 * @param target the target buffer for extraction
1573 * @param targetLength the length of the target buffer
1574 * If `target` is NULL, then the number of bytes required for
1575 * `target` is returned.
1576 * @return the output string length, not including the terminating NUL
1577 * @stable ICU 2.0
1578 */
1579 int32_t extract(int32_t start,
1580 int32_t startLength,
1581 char *target,
1582 uint32_t targetLength) const;
1583
1584#endif
1585
1586#if !UCONFIG_NO_CONVERSION0
1587
1588 /**
1589 * Copy the characters in the range
1590 * [`start`, `start + length`) into an array of characters
1591 * in a specified codepage.
1592 * The output string is NUL-terminated.
1593 *
1594 * Recommendation: For invariant-character strings use
1595 * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
1596 * because it avoids object code dependencies of UnicodeString on
1597 * the conversion code.
1598 *
1599 * @param start offset of first character which will be copied
1600 * @param startLength the number of characters to extract
1601 * @param target the target buffer for extraction
1602 * @param codepage the desired codepage for the characters. 0 has
1603 * the special meaning of the default codepage
1604 * If `codepage` is an empty string (`""`),
1605 * then a simple conversion is performed on the codepage-invariant
1606 * subset ("invariant characters") of the platform encoding. See utypes.h.
1607 * If `target` is NULL, then the number of bytes required for
1608 * `target` is returned. It is assumed that the target is big enough
1609 * to fit all of the characters.
1610 * @return the output string length, not including the terminating NUL
1611 * @stable ICU 2.0
1612 */
1613 inline int32_t extract(int32_t start,
1614 int32_t startLength,
1615 char *target,
1616 const char *codepage = 0) const;
1617
1618 /**
1619 * Copy the characters in the range
1620 * [`start`, `start + length`) into an array of characters
1621 * in a specified codepage.
1622 * This function does not write any more than `targetLength`
1623 * characters but returns the length of the entire output string
1624 * so that one can allocate a larger buffer and call the function again
1625 * if necessary.
1626 * The output string is NUL-terminated if possible.
1627 *
1628 * Recommendation: For invariant-character strings use
1629 * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
1630 * because it avoids object code dependencies of UnicodeString on
1631 * the conversion code.
1632 *
1633 * @param start offset of first character which will be copied
1634 * @param startLength the number of characters to extract
1635 * @param target the target buffer for extraction
1636 * @param targetLength the length of the target buffer
1637 * @param codepage the desired codepage for the characters. 0 has
1638 * the special meaning of the default codepage
1639 * If `codepage` is an empty string (`""`),
1640 * then a simple conversion is performed on the codepage-invariant
1641 * subset ("invariant characters") of the platform encoding. See utypes.h.
1642 * If `target` is NULL, then the number of bytes required for
1643 * `target` is returned.
1644 * @return the output string length, not including the terminating NUL
1645 * @stable ICU 2.0
1646 */
1647 int32_t extract(int32_t start,
1648 int32_t startLength,
1649 char *target,
1650 uint32_t targetLength,
1651 const char *codepage) const;
1652
1653 /**
1654 * Convert the UnicodeString into a codepage string using an existing UConverter.
1655 * The output string is NUL-terminated if possible.
1656 *
1657 * This function avoids the overhead of opening and closing a converter if
1658 * multiple strings are extracted.
1659 *
1660 * @param dest destination string buffer, can be NULL if destCapacity==0
1661 * @param destCapacity the number of chars available at dest
1662 * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called),
1663 * or NULL for the default converter
1664 * @param errorCode normal ICU error code
1665 * @return the length of the output string, not counting the terminating NUL;
1666 * if the length is greater than destCapacity, then the string will not fit
1667 * and a buffer of the indicated length would need to be passed in
1668 * @stable ICU 2.0
1669 */
1670 int32_t extract(char *dest, int32_t destCapacity,
1671 UConverter *cnv,
1672 UErrorCode &errorCode) const;
1673
1674#endif
1675
1676 /**
1677 * Create a temporary substring for the specified range.
1678 * Unlike the substring constructor and setTo() functions,
1679 * the object returned here will be a read-only alias (using getBuffer())
1680 * rather than copying the text.
1681 * As a result, this substring operation is much faster but requires
1682 * that the original string not be modified or deleted during the lifetime
1683 * of the returned substring object.
1684 * @param start offset of the first character visible in the substring
1685 * @param length length of the substring
1686 * @return a read-only alias UnicodeString object for the substring
1687 * @stable ICU 4.4
1688 */
1689 UnicodeString tempSubString(int32_t start=0, int32_t length=INT32_MAX(2147483647)) const;
1690
1691 /**
1692 * Create a temporary substring for the specified range.
1693 * Same as tempSubString(start, length) except that the substring range
1694 * is specified as a (start, limit) pair (with an exclusive limit index)
1695 * rather than a (start, length) pair.
1696 * @param start offset of the first character visible in the substring
1697 * @param limit offset immediately following the last character visible in the substring
1698 * @return a read-only alias UnicodeString object for the substring
1699 * @stable ICU 4.4
1700 */
1701 inline UnicodeString tempSubStringBetween(int32_t start, int32_t limit=INT32_MAX(2147483647)) const;
1702
1703 /**
1704 * Convert the UnicodeString to UTF-8 and write the result
1705 * to a ByteSink. This is called by toUTF8String().
1706 * Unpaired surrogates are replaced with U+FFFD.
1707 * Calls u_strToUTF8WithSub().
1708 *
1709 * @param sink A ByteSink to which the UTF-8 version of the string is written.
1710 * sink.Flush() is called at the end.
1711 * @stable ICU 4.2
1712 * @see toUTF8String
1713 */
1714 void toUTF8(ByteSink &sink) const;
1715
1716 /**
1717 * Convert the UnicodeString to UTF-8 and append the result
1718 * to a standard string.
1719 * Unpaired surrogates are replaced with U+FFFD.
1720 * Calls toUTF8().
1721 *
1722 * @param result A standard string (or a compatible object)
1723 * to which the UTF-8 version of the string is appended.
1724 * @return The string object.
1725 * @stable ICU 4.2
1726 * @see toUTF8
1727 */
1728 template<typename StringClass>
1729 StringClass &toUTF8String(StringClass &result) const {
1730 StringByteSink<StringClass> sbs(&result, length());
1731 toUTF8(sbs);
1732 return result;
1733 }
1734
1735 /**
1736 * Convert the UnicodeString to UTF-32.
1737 * Unpaired surrogates are replaced with U+FFFD.
1738 * Calls u_strToUTF32WithSub().
1739 *
1740 * @param utf32 destination string buffer, can be NULL if capacity==0
1741 * @param capacity the number of UChar32s available at utf32
1742 * @param errorCode Standard ICU error code. Its input value must
1743 * pass the U_SUCCESS() test, or else the function returns
1744 * immediately. Check for U_FAILURE() on output or use with
1745 * function chaining. (See User Guide for details.)
1746 * @return The length of the UTF-32 string.
1747 * @see fromUTF32
1748 * @stable ICU 4.2
1749 */
1750 int32_t toUTF32(UChar32 *utf32, int32_t capacity, UErrorCode &errorCode) const;
1751
1752 /* Length operations */
1753
1754 /**
1755 * Return the length of the UnicodeString object.
1756 * The length is the number of char16_t code units are in the UnicodeString.
1757 * If you want the number of code points, please use countChar32().
1758 * @return the length of the UnicodeString object
1759 * @see countChar32
1760 * @stable ICU 2.0
1761 */
1762 inline int32_t length(void) const;
1763
1764 /**
1765 * Count Unicode code points in the length char16_t code units of the string.
1766 * A code point may occupy either one or two char16_t code units.
1767 * Counting code points involves reading all code units.
1768 *
1769 * This functions is basically the inverse of moveIndex32().
1770 *
1771 * @param start the index of the first code unit to check
1772 * @param length the number of char16_t code units to check
1773 * @return the number of code points in the specified code units
1774 * @see length
1775 * @stable ICU 2.0
1776 */
1777 int32_t
1778 countChar32(int32_t start=0, int32_t length=INT32_MAX(2147483647)) const;
1779
1780 /**
1781 * Check if the length char16_t code units of the string
1782 * contain more Unicode code points than a certain number.
1783 * This is more efficient than counting all code points in this part of the string
1784 * and comparing that number with a threshold.
1785 * This function may not need to scan the string at all if the length
1786 * falls within a certain range, and
1787 * never needs to count more than 'number+1' code points.
1788 * Logically equivalent to (countChar32(start, length)>number).
1789 * A Unicode code point may occupy either one or two char16_t code units.
1790 *
1791 * @param start the index of the first code unit to check (0 for the entire string)
1792 * @param length the number of char16_t code units to check
1793 * (use INT32_MAX for the entire string; remember that start/length
1794 * values are pinned)
1795 * @param number The number of code points in the (sub)string is compared against
1796 * the 'number' parameter.
1797 * @return Boolean value for whether the string contains more Unicode code points
1798 * than 'number'. Same as (u_countChar32(s, length)>number).
1799 * @see countChar32
1800 * @see u_strHasMoreChar32Than
1801 * @stable ICU 2.4
1802 */
1803 UBool
1804 hasMoreChar32Than(int32_t start, int32_t length, int32_t number) const;
1805
1806 /**
1807 * Determine if this string is empty.
1808 * @return true if this string contains 0 characters, false otherwise.
1809 * @stable ICU 2.0
1810 */
1811 inline UBool isEmpty(void) const;
1812
1813 /**
1814 * Return the capacity of the internal buffer of the UnicodeString object.
1815 * This is useful together with the getBuffer functions.
1816 * See there for details.
1817 *
1818 * @return the number of char16_ts available in the internal buffer
1819 * @see getBuffer
1820 * @stable ICU 2.0
1821 */
1822 inline int32_t getCapacity(void) const;
1823
1824 /* Other operations */
1825
1826 /**
1827 * Generate a hash code for this object.
1828 * @return The hash code of this UnicodeString.
1829 * @stable ICU 2.0
1830 */
1831 inline int32_t hashCode(void) const;
1832
1833 /**
1834 * Determine if this object contains a valid string.
1835 * A bogus string has no value. It is different from an empty string,
1836 * although in both cases isEmpty() returns true and length() returns 0.
1837 * setToBogus() and isBogus() can be used to indicate that no string value is available.
1838 * For a bogus string, getBuffer() and getTerminatedBuffer() return NULL, and
1839 * length() returns 0.
1840 *
1841 * @return true if the string is bogus/invalid, false otherwise
1842 * @see setToBogus()
1843 * @stable ICU 2.0
1844 */
1845 inline UBool isBogus(void) const;
1846
1847
1848 //========================================
1849 // Write operations
1850 //========================================
1851
1852 /* Assignment operations */
1853
1854 /**
1855 * Assignment operator. Replace the characters in this UnicodeString
1856 * with the characters from `srcText`.
1857 *
1858 * Starting with ICU 2.4, the assignment operator and the copy constructor
1859 * allocate a new buffer and copy the buffer contents even for readonly aliases.
1860 * By contrast, the fastCopyFrom() function implements the old,
1861 * more efficient but less safe behavior
1862 * of making this string also a readonly alias to the same buffer.
1863 *
1864 * If the source object has an "open" buffer from getBuffer(minCapacity),
1865 * then the copy is an empty string.
1866 *
1867 * @param srcText The text containing the characters to replace
1868 * @return a reference to this
1869 * @stable ICU 2.0
1870 * @see fastCopyFrom
1871 */
1872 UnicodeString &operator=(const UnicodeString &srcText);
1873
1874 /**
1875 * Almost the same as the assignment operator.
1876 * Replace the characters in this UnicodeString
1877 * with the characters from `srcText`.
1878 *
1879 * This function works the same as the assignment operator
1880 * for all strings except for ones that are readonly aliases.
1881 *
1882 * Starting with ICU 2.4, the assignment operator and the copy constructor
1883 * allocate a new buffer and copy the buffer contents even for readonly aliases.
1884 * This function implements the old, more efficient but less safe behavior
1885 * of making this string also a readonly alias to the same buffer.
1886 *
1887 * The fastCopyFrom function must be used only if it is known that the lifetime of
1888 * this UnicodeString does not exceed the lifetime of the aliased buffer
1889 * including its contents, for example for strings from resource bundles
1890 * or aliases to string constants.
1891 *
1892 * If the source object has an "open" buffer from getBuffer(minCapacity),
1893 * then the copy is an empty string.
1894 *
1895 * @param src The text containing the characters to replace.
1896 * @return a reference to this
1897 * @stable ICU 2.4
1898 */
1899 UnicodeString &fastCopyFrom(const UnicodeString &src);
1900
1901 /**
1902 * Move assignment operator; might leave src in bogus state.
1903 * This string will have the same contents and state that the source string had.
1904 * The behavior is undefined if *this and src are the same object.
1905 * @param src source string
1906 * @return *this
1907 * @stable ICU 56
1908 */
1909 UnicodeString &operator=(UnicodeString &&src) U_NOEXCEPTnoexcept;
1910
1911 /**
1912 * Swap strings.
1913 * @param other other string
1914 * @stable ICU 56
1915 */
1916 void swap(UnicodeString &other) U_NOEXCEPTnoexcept;
1917
1918 /**
1919 * Non-member UnicodeString swap function.
1920 * @param s1 will get s2's contents and state
1921 * @param s2 will get s1's contents and state
1922 * @stable ICU 56
1923 */
1924 friend inline void U_EXPORT2
1925 swap(UnicodeString &s1, UnicodeString &s2) U_NOEXCEPTnoexcept {
1926 s1.swap(s2);
1927 }
1928
1929 /**
1930 * Assignment operator. Replace the characters in this UnicodeString
1931 * with the code unit `ch`.
1932 * @param ch the code unit to replace
1933 * @return a reference to this
1934 * @stable ICU 2.0
1935 */
1936 inline UnicodeString& operator= (char16_t ch);
1937
1938 /**
1939 * Assignment operator. Replace the characters in this UnicodeString
1940 * with the code point `ch`.
1941 * @param ch the code point to replace
1942 * @return a reference to this
1943 * @stable ICU 2.0
1944 */
1945 inline UnicodeString& operator= (UChar32 ch);
1946
1947 /**
1948 * Set the text in the UnicodeString object to the characters
1949 * in `srcText` in the range
1950 * [`srcStart`, `srcText.length()`).
1951 * `srcText` is not modified.
1952 * @param srcText the source for the new characters
1953 * @param srcStart the offset into `srcText` where new characters
1954 * will be obtained
1955 * @return a reference to this
1956 * @stable ICU 2.2
1957 */
1958 inline UnicodeString& setTo(const UnicodeString& srcText,
1959 int32_t srcStart);
1960
1961 /**
1962 * Set the text in the UnicodeString object to the characters
1963 * in `srcText` in the range
1964 * [`srcStart`, `srcStart + srcLength`).
1965 * `srcText` is not modified.
1966 * @param srcText the source for the new characters
1967 * @param srcStart the offset into `srcText` where new characters
1968 * will be obtained
1969 * @param srcLength the number of characters in `srcText` in the
1970 * replace string.
1971 * @return a reference to this
1972 * @stable ICU 2.0
1973 */
1974 inline UnicodeString& setTo(const UnicodeString& srcText,
1975 int32_t srcStart,
1976 int32_t srcLength);
1977
1978 /**
1979 * Set the text in the UnicodeString object to the characters in
1980 * `srcText`.
1981 * `srcText` is not modified.
1982 * @param srcText the source for the new characters
1983 * @return a reference to this
1984 * @stable ICU 2.0
1985 */
1986 inline UnicodeString& setTo(const UnicodeString& srcText);
1987
1988 /**
1989 * Set the characters in the UnicodeString object to the characters
1990 * in `srcChars`. `srcChars` is not modified.
1991 * @param srcChars the source for the new characters
1992 * @param srcLength the number of Unicode characters in srcChars.
1993 * @return a reference to this
1994 * @stable ICU 2.0
1995 */
1996 inline UnicodeString& setTo(const char16_t *srcChars,
1997 int32_t srcLength);
1998
1999 /**
2000 * Set the characters in the UnicodeString object to the code unit
2001 * `srcChar`.
2002 * @param srcChar the code unit which becomes the UnicodeString's character
2003 * content
2004 * @return a reference to this
2005 * @stable ICU 2.0
2006 */
2007 inline UnicodeString& setTo(char16_t srcChar);
2008
2009 /**
2010 * Set the characters in the UnicodeString object to the code point
2011 * `srcChar`.
2012 * @param srcChar the code point which becomes the UnicodeString's character
2013 * content
2014 * @return a reference to this
2015 * @stable ICU 2.0
2016 */
2017 inline UnicodeString& setTo(UChar32 srcChar);
2018
2019 /**
2020 * Aliasing setTo() function, analogous to the readonly-aliasing char16_t* constructor.
2021 * The text will be used for the UnicodeString object, but
2022 * it will not be released when the UnicodeString is destroyed.
2023 * This has copy-on-write semantics:
2024 * When the string is modified, then the buffer is first copied into
2025 * newly allocated memory.
2026 * The aliased buffer is never modified.
2027 *
2028 * In an assignment to another UnicodeString, when using the copy constructor
2029 * or the assignment operator, the text will be copied.
2030 * When using fastCopyFrom(), the text will be aliased again,
2031 * so that both strings then alias the same readonly-text.
2032 *
2033 * @param isTerminated specifies if `text` is `NUL`-terminated.
2034 * This must be true if `textLength==-1`.
2035 * @param text The characters to alias for the UnicodeString.
2036 * @param textLength The number of Unicode characters in `text` to alias.
2037 * If -1, then this constructor will determine the length
2038 * by calling `u_strlen()`.
2039 * @return a reference to this
2040 * @stable ICU 2.0
2041 */
2042 UnicodeString &setTo(UBool isTerminated,
2043 ConstChar16Ptr text,
2044 int32_t textLength);
2045
2046 /**
2047 * Aliasing setTo() function, analogous to the writable-aliasing char16_t* constructor.
2048 * The text will be used for the UnicodeString object, but
2049 * it will not be released when the UnicodeString is destroyed.
2050 * This has write-through semantics:
2051 * For as long as the capacity of the buffer is sufficient, write operations
2052 * will directly affect the buffer. When more capacity is necessary, then
2053 * a new buffer will be allocated and the contents copied as with regularly
2054 * constructed strings.
2055 * In an assignment to another UnicodeString, the buffer will be copied.
2056 * The extract(Char16Ptr dst) function detects whether the dst pointer is the same
2057 * as the string buffer itself and will in this case not copy the contents.
2058 *
2059 * @param buffer The characters to alias for the UnicodeString.
2060 * @param buffLength The number of Unicode characters in `buffer` to alias.
2061 * @param buffCapacity The size of `buffer` in char16_ts.
2062 * @return a reference to this
2063 * @stable ICU 2.0
2064 */
2065 UnicodeString &setTo(char16_t *buffer,
2066 int32_t buffLength,
2067 int32_t buffCapacity);
2068
2069 /**
2070 * Make this UnicodeString object invalid.
2071 * The string will test true with isBogus().
2072 *
2073 * A bogus string has no value. It is different from an empty string.
2074 * It can be used to indicate that no string value is available.
2075 * getBuffer() and getTerminatedBuffer() return NULL, and
2076 * length() returns 0.
2077 *
2078 * This utility function is used throughout the UnicodeString
2079 * implementation to indicate that a UnicodeString operation failed,
2080 * and may be used in other functions,
2081 * especially but not exclusively when such functions do not
2082 * take a UErrorCode for simplicity.
2083 *
2084 * The following methods, and no others, will clear a string object's bogus flag:
2085 * - remove()
2086 * - remove(0, INT32_MAX)
2087 * - truncate(0)
2088 * - operator=() (assignment operator)
2089 * - setTo(...)
2090 *
2091 * The simplest ways to turn a bogus string into an empty one
2092 * is to use the remove() function.
2093 * Examples for other functions that are equivalent to "set to empty string":
2094 * \code
2095 * if(s.isBogus()) {
2096 * s.remove(); // set to an empty string (remove all), or
2097 * s.remove(0, INT32_MAX); // set to an empty string (remove all), or
2098 * s.truncate(0); // set to an empty string (complete truncation), or
2099 * s=UnicodeString(); // assign an empty string, or
2100 * s.setTo((UChar32)-1); // set to a pseudo code point that is out of range, or
2101 * s.setTo(u"", 0); // set to an empty C Unicode string
2102 * }
2103 * \endcode
2104 *
2105 * @see isBogus()
2106 * @stable ICU 2.0
2107 */
2108 void setToBogus();
2109
2110 /**
2111 * Set the character at the specified offset to the specified character.
2112 * @param offset A valid offset into the text of the character to set
2113 * @param ch The new character
2114 * @return A reference to this
2115 * @stable ICU 2.0
2116 */
2117 UnicodeString& setCharAt(int32_t offset,
2118 char16_t ch);
2119
2120
2121 /* Append operations */
2122
2123 /**
2124 * Append operator. Append the code unit `ch` to the UnicodeString
2125 * object.
2126 * @param ch the code unit to be appended
2127 * @return a reference to this
2128 * @stable ICU 2.0
2129 */
2130 inline UnicodeString& operator+= (char16_t ch);
2131
2132 /**
2133 * Append operator. Append the code point `ch` to the UnicodeString
2134 * object.
2135 * @param ch the code point to be appended
2136 * @return a reference to this
2137 * @stable ICU 2.0
2138 */
2139 inline UnicodeString& operator+= (UChar32 ch);
2140
2141 /**
2142 * Append operator. Append the characters in `srcText` to the
2143 * UnicodeString object. `srcText` is not modified.
2144 * @param srcText the source for the new characters
2145 * @return a reference to this
2146 * @stable ICU 2.0
2147 */
2148 inline UnicodeString& operator+= (const UnicodeString& srcText);
2149
2150 /**
2151 * Append the characters
2152 * in `srcText` in the range
2153 * [`srcStart`, `srcStart + srcLength`) to the
2154 * UnicodeString object at offset `start`. `srcText`
2155 * is not modified.
2156 * @param srcText the source for the new characters
2157 * @param srcStart the offset into `srcText` where new characters
2158 * will be obtained
2159 * @param srcLength the number of characters in `srcText` in
2160 * the append string
2161 * @return a reference to this
2162 * @stable ICU 2.0
2163 */
2164 inline UnicodeString& append(const UnicodeString& srcText,
2165 int32_t srcStart,
2166 int32_t srcLength);
2167
2168 /**
2169 * Append the characters in `srcText` to the UnicodeString object.
2170 * `srcText` is not modified.
2171 * @param srcText the source for the new characters
2172 * @return a reference to this
2173 * @stable ICU 2.0
2174 */
2175 inline UnicodeString& append(const UnicodeString& srcText);
2176
2177 /**
2178 * Append the characters in `srcChars` in the range
2179 * [`srcStart`, `srcStart + srcLength`) to the UnicodeString
2180 * object at offset
2181 * `start`. `srcChars` is not modified.
2182 * @param srcChars the source for the new characters
2183 * @param srcStart the offset into `srcChars` where new characters
2184 * will be obtained
2185 * @param srcLength the number of characters in `srcChars` in
2186 * the append string; can be -1 if `srcChars` is NUL-terminated
2187 * @return a reference to this
2188 * @stable ICU 2.0
2189 */
2190 inline UnicodeString& append(const char16_t *srcChars,
2191 int32_t srcStart,
2192 int32_t srcLength);
2193
2194 /**
2195 * Append the characters in `srcChars` to the UnicodeString object
2196 * at offset `start`. `srcChars` is not modified.
2197 * @param srcChars the source for the new characters
2198 * @param srcLength the number of Unicode characters in `srcChars`;
2199 * can be -1 if `srcChars` is NUL-terminated
2200 * @return a reference to this
2201 * @stable ICU 2.0
2202 */
2203 inline UnicodeString& append(ConstChar16Ptr srcChars,
2204 int32_t srcLength);
2205
2206 /**
2207 * Append the code unit `srcChar` to the UnicodeString object.
2208 * @param srcChar the code unit to append
2209 * @return a reference to this
2210 * @stable ICU 2.0
2211 */
2212 inline UnicodeString& append(char16_t srcChar);
2213
2214 /**
2215 * Append the code point `srcChar` to the UnicodeString object.
2216 * @param srcChar the code point to append
2217 * @return a reference to this
2218 * @stable ICU 2.0
2219 */
2220 UnicodeString& append(UChar32 srcChar);
2221
2222
2223 /* Insert operations */
2224
2225 /**
2226 * Insert the characters in `srcText` in the range
2227 * [`srcStart`, `srcStart + srcLength`) into the UnicodeString
2228 * object at offset `start`. `srcText` is not modified.
2229 * @param start the offset where the insertion begins
2230 * @param srcText the source for the new characters
2231 * @param srcStart the offset into `srcText` where new characters
2232 * will be obtained
2233 * @param srcLength the number of characters in `srcText` in
2234 * the insert string
2235 * @return a reference to this
2236 * @stable ICU 2.0
2237 */
2238 inline UnicodeString& insert(int32_t start,
2239 const UnicodeString& srcText,
2240 int32_t srcStart,
2241 int32_t srcLength);
2242
2243 /**
2244 * Insert the characters in `srcText` into the UnicodeString object
2245 * at offset `start`. `srcText` is not modified.
2246 * @param start the offset where the insertion begins
2247 * @param srcText the source for the new characters
2248 * @return a reference to this
2249 * @stable ICU 2.0
2250 */
2251 inline UnicodeString& insert(int32_t start,
2252 const UnicodeString& srcText);
2253
2254 /**
2255 * Insert the characters in `srcChars` in the range
2256 * [`srcStart`, `srcStart + srcLength`) into the UnicodeString
2257 * object at offset `start`. `srcChars` is not modified.
2258 * @param start the offset at which the insertion begins
2259 * @param srcChars the source for the new characters
2260 * @param srcStart the offset into `srcChars` where new characters
2261 * will be obtained
2262 * @param srcLength the number of characters in `srcChars`
2263 * in the insert string
2264 * @return a reference to this
2265 * @stable ICU 2.0
2266 */
2267 inline UnicodeString& insert(int32_t start,
2268 const char16_t *srcChars,
2269 int32_t srcStart,
2270 int32_t srcLength);
2271
2272 /**
2273 * Insert the characters in `srcChars` into the UnicodeString object
2274 * at offset `start`. `srcChars` is not modified.
2275 * @param start the offset where the insertion begins
2276 * @param srcChars the source for the new characters
2277 * @param srcLength the number of Unicode characters in srcChars.
2278 * @return a reference to this
2279 * @stable ICU 2.0
2280 */
2281 inline UnicodeString& insert(int32_t start,
2282 ConstChar16Ptr srcChars,
2283 int32_t srcLength);
2284
2285 /**
2286 * Insert the code unit `srcChar` into the UnicodeString object at
2287 * offset `start`.
2288 * @param start the offset at which the insertion occurs
2289 * @param srcChar the code unit to insert
2290 * @return a reference to this
2291 * @stable ICU 2.0
2292 */
2293 inline UnicodeString& insert(int32_t start,
2294 char16_t srcChar);
2295
2296 /**
2297 * Insert the code point `srcChar` into the UnicodeString object at
2298 * offset `start`.
2299 * @param start the offset at which the insertion occurs
2300 * @param srcChar the code point to insert
2301 * @return a reference to this
2302 * @stable ICU 2.0
2303 */
2304 inline UnicodeString& insert(int32_t start,
2305 UChar32 srcChar);
2306
2307
2308 /* Replace operations */
2309
2310 /**
2311 * Replace the characters in the range
2312 * [`start`, `start + length`) with the characters in
2313 * `srcText` in the range
2314 * [`srcStart`, `srcStart + srcLength`).
2315 * `srcText` is not modified.
2316 * @param start the offset at which the replace operation begins
2317 * @param length the number of characters to replace. The character at
2318 * `start + length` is not modified.
2319 * @param srcText the source for the new characters
2320 * @param srcStart the offset into `srcText` where new characters
2321 * will be obtained
2322 * @param srcLength the number of characters in `srcText` in
2323 * the replace string
2324 * @return a reference to this
2325 * @stable ICU 2.0
2326 */
2327 inline UnicodeString& replace(int32_t start,
2328 int32_t length,
2329 const UnicodeString& srcText,
2330 int32_t srcStart,
2331 int32_t srcLength);
2332
2333 /**
2334 * Replace the characters in the range
2335 * [`start`, `start + length`)
2336 * with the characters in `srcText`. `srcText` is
2337 * not modified.
2338 * @param start the offset at which the replace operation begins
2339 * @param length the number of characters to replace. The character at
2340 * `start + length` is not modified.
2341 * @param srcText the source for the new characters
2342 * @return a reference to this
2343 * @stable ICU 2.0
2344 */
2345 inline UnicodeString& replace(int32_t start,
2346 int32_t length,
2347 const UnicodeString& srcText);
2348
2349 /**
2350 * Replace the characters in the range
2351 * [`start`, `start + length`) with the characters in
2352 * `srcChars` in the range
2353 * [`srcStart`, `srcStart + srcLength`). `srcChars`
2354 * is not modified.
2355 * @param start the offset at which the replace operation begins
2356 * @param length the number of characters to replace. The character at
2357 * `start + length` is not modified.
2358 * @param srcChars the source for the new characters
2359 * @param srcStart the offset into `srcChars` where new characters
2360 * will be obtained
2361 * @param srcLength the number of characters in `srcChars`
2362 * in the replace string
2363 * @return a reference to this
2364 * @stable ICU 2.0
2365 */
2366 inline UnicodeString& replace(int32_t start,
2367 int32_t length,
2368 const char16_t *srcChars,
2369 int32_t srcStart,
2370 int32_t srcLength);
2371
2372 /**
2373 * Replace the characters in the range
2374 * [`start`, `start + length`) with the characters in
2375 * `srcChars`. `srcChars` is not modified.
2376 * @param start the offset at which the replace operation begins
2377 * @param length number of characters to replace. The character at
2378 * `start + length` is not modified.
2379 * @param srcChars the source for the new characters
2380 * @param srcLength the number of Unicode characters in srcChars
2381 * @return a reference to this
2382 * @stable ICU 2.0
2383 */
2384 inline UnicodeString& replace(int32_t start,
2385 int32_t length,
2386 ConstChar16Ptr srcChars,
2387 int32_t srcLength);
2388
2389 /**
2390 * Replace the characters in the range
2391 * [`start`, `start + length`) with the code unit
2392 * `srcChar`.
2393 * @param start the offset at which the replace operation begins
2394 * @param length the number of characters to replace. The character at
2395 * `start + length` is not modified.
2396 * @param srcChar the new code unit
2397 * @return a reference to this
2398 * @stable ICU 2.0
2399 */
2400 inline UnicodeString& replace(int32_t start,
2401 int32_t length,
2402 char16_t srcChar);
2403
2404 /**
2405 * Replace the characters in the range
2406 * [`start`, `start + length`) with the code point
2407 * `srcChar`.
2408 * @param start the offset at which the replace operation begins
2409 * @param length the number of characters to replace. The character at
2410 * `start + length` is not modified.
2411 * @param srcChar the new code point
2412 * @return a reference to this
2413 * @stable ICU 2.0
2414 */
2415 UnicodeString& replace(int32_t start, int32_t length, UChar32 srcChar);
2416
2417 /**
2418 * Replace the characters in the range [`start`, `limit`)
2419 * with the characters in `srcText`. `srcText` is not modified.
2420 * @param start the offset at which the replace operation begins
2421 * @param limit the offset immediately following the replace range
2422 * @param srcText the source for the new characters
2423 * @return a reference to this
2424 * @stable ICU 2.0
2425 */
2426 inline UnicodeString& replaceBetween(int32_t start,
2427 int32_t limit,
2428 const UnicodeString& srcText);
2429
2430 /**
2431 * Replace the characters in the range [`start`, `limit`)
2432 * with the characters in `srcText` in the range
2433 * [`srcStart`, `srcLimit`). `srcText` is not modified.
2434 * @param start the offset at which the replace operation begins
2435 * @param limit the offset immediately following the replace range
2436 * @param srcText the source for the new characters
2437 * @param srcStart the offset into `srcChars` where new characters
2438 * will be obtained
2439 * @param srcLimit the offset immediately following the range to copy
2440 * in `srcText`
2441 * @return a reference to this
2442 * @stable ICU 2.0
2443 */
2444 inline UnicodeString& replaceBetween(int32_t start,
2445 int32_t limit,
2446 const UnicodeString& srcText,
2447 int32_t srcStart,
2448 int32_t srcLimit);
2449
2450 /**
2451 * Replace a substring of this object with the given text.
2452 * @param start the beginning index, inclusive; `0 <= start <= limit`.
2453 * @param limit the ending index, exclusive; `start <= limit <= length()`.
2454 * @param text the text to replace characters `start` to `limit - 1`
2455 * @stable ICU 2.0
2456 */
2457 virtual void handleReplaceBetween(int32_t start,
2458 int32_t limit,
2459 const UnicodeString& text) override;
2460
2461 /**
2462 * Replaceable API
2463 * @return true if it has MetaData
2464 * @stable ICU 2.4
2465 */
2466 virtual UBool hasMetaData() const override;
2467
2468 /**
2469 * Copy a substring of this object, retaining attribute (out-of-band)
2470 * information. This method is used to duplicate or reorder substrings.
2471 * The destination index must not overlap the source range.
2472 *
2473 * @param start the beginning index, inclusive; `0 <= start <= limit`.
2474 * @param limit the ending index, exclusive; `start <= limit <= length()`.
2475 * @param dest the destination index. The characters from
2476 * `start..limit-1` will be copied to `dest`.
2477 * Implementations of this method may assume that `dest <= start ||
2478 * dest >= limit`.
2479 * @stable ICU 2.0
2480 */
2481 virtual void copy(int32_t start, int32_t limit, int32_t dest) override;
2482
2483 /* Search and replace operations */
2484
2485 /**
2486 * Replace all occurrences of characters in oldText with the characters
2487 * in newText
2488 * @param oldText the text containing the search text
2489 * @param newText the text containing the replacement text
2490 * @return a reference to this
2491 * @stable ICU 2.0
2492 */
2493 inline UnicodeString& findAndReplace(const UnicodeString& oldText,
2494 const UnicodeString& newText);
2495
2496 /**
2497 * Replace all occurrences of characters in oldText with characters
2498 * in newText
2499 * in the range [`start`, `start + length`).
2500 * @param start the start of the range in which replace will performed
2501 * @param length the length of the range in which replace will be performed
2502 * @param oldText the text containing the search text
2503 * @param newText the text containing the replacement text
2504 * @return a reference to this
2505 * @stable ICU 2.0
2506 */
2507 inline UnicodeString& findAndReplace(int32_t start,
2508 int32_t length,
2509 const UnicodeString& oldText,
2510 const UnicodeString& newText);
2511
2512 /**
2513 * Replace all occurrences of characters in oldText in the range
2514 * [`oldStart`, `oldStart + oldLength`) with the characters
2515 * in newText in the range
2516 * [`newStart`, `newStart + newLength`)
2517 * in the range [`start`, `start + length`).
2518 * @param start the start of the range in which replace will performed
2519 * @param length the length of the range in which replace will be performed
2520 * @param oldText the text containing the search text
2521 * @param oldStart the start of the search range in `oldText`
2522 * @param oldLength the length of the search range in `oldText`
2523 * @param newText the text containing the replacement text
2524 * @param newStart the start of the replacement range in `newText`
2525 * @param newLength the length of the replacement range in `newText`
2526 * @return a reference to this
2527 * @stable ICU 2.0
2528 */
2529 UnicodeString& findAndReplace(int32_t start,
2530 int32_t length,
2531 const UnicodeString& oldText,
2532 int32_t oldStart,
2533 int32_t oldLength,
2534 const UnicodeString& newText,
2535 int32_t newStart,
2536 int32_t newLength);
2537
2538
2539 /* Remove operations */
2540
2541 /**
2542 * Removes all characters from the UnicodeString object and clears the bogus flag.
2543 * This is the UnicodeString equivalent of std::string’s clear().
2544 *
2545 * @return a reference to this
2546 * @see setToBogus
2547 * @stable ICU 2.0
2548 */
2549 inline UnicodeString& remove();
2550
2551 /**
2552 * Remove the characters in the range
2553 * [`start`, `start + length`) from the UnicodeString object.
2554 * @param start the offset of the first character to remove
2555 * @param length the number of characters to remove
2556 * @return a reference to this
2557 * @stable ICU 2.0
2558 */
2559 inline UnicodeString& remove(int32_t start,
2560 int32_t length = (int32_t)INT32_MAX(2147483647));
2561
2562 /**
2563 * Remove the characters in the range
2564 * [`start`, `limit`) from the UnicodeString object.
2565 * @param start the offset of the first character to remove
2566 * @param limit the offset immediately following the range to remove
2567 * @return a reference to this
2568 * @stable ICU 2.0
2569 */
2570 inline UnicodeString& removeBetween(int32_t start,
2571 int32_t limit = (int32_t)INT32_MAX(2147483647));
2572
2573 /**
2574 * Retain only the characters in the range
2575 * [`start`, `limit`) from the UnicodeString object.
2576 * Removes characters before `start` and at and after `limit`.
2577 * @param start the offset of the first character to retain
2578 * @param limit the offset immediately following the range to retain
2579 * @return a reference to this
2580 * @stable ICU 4.4
2581 */
2582 inline UnicodeString &retainBetween(int32_t start, int32_t limit = INT32_MAX(2147483647));
2583
2584 /* Length operations */
2585
2586 /**
2587 * Pad the start of this UnicodeString with the character `padChar`.
2588 * If the length of this UnicodeString is less than targetLength,
2589 * length() - targetLength copies of padChar will be added to the
2590 * beginning of this UnicodeString.
2591 * @param targetLength the desired length of the string
2592 * @param padChar the character to use for padding. Defaults to
2593 * space (U+0020)
2594 * @return true if the text was padded, false otherwise.
2595 * @stable ICU 2.0
2596 */
2597 UBool padLeading(int32_t targetLength,
2598 char16_t padChar = 0x0020);
2599
2600 /**
2601 * Pad the end of this UnicodeString with the character `padChar`.
2602 * If the length of this UnicodeString is less than targetLength,
2603 * length() - targetLength copies of padChar will be added to the
2604 * end of this UnicodeString.
2605 * @param targetLength the desired length of the string
2606 * @param padChar the character to use for padding. Defaults to
2607 * space (U+0020)
2608 * @return true if the text was padded, false otherwise.
2609 * @stable ICU 2.0
2610 */
2611 UBool padTrailing(int32_t targetLength,
2612 char16_t padChar = 0x0020);
2613
2614 /**
2615 * Truncate this UnicodeString to the `targetLength`.
2616 * @param targetLength the desired length of this UnicodeString.
2617 * @return true if the text was truncated, false otherwise
2618 * @stable ICU 2.0
2619 */
2620 inline UBool truncate(int32_t targetLength);
2621
2622 /**
2623 * Trims leading and trailing whitespace from this UnicodeString.
2624 * @return a reference to this
2625 * @stable ICU 2.0
2626 */
2627 UnicodeString& trim(void);
2628
2629
2630 /* Miscellaneous operations */
2631
2632 /**
2633 * Reverse this UnicodeString in place.
2634 * @return a reference to this
2635 * @stable ICU 2.0
2636 */
2637 inline UnicodeString& reverse(void);
2638
2639 /**
2640 * Reverse the range [`start`, `start + length`) in
2641 * this UnicodeString.
2642 * @param start the start of the range to reverse
2643 * @param length the number of characters to to reverse
2644 * @return a reference to this
2645 * @stable ICU 2.0
2646 */
2647 inline UnicodeString& reverse(int32_t start,
2648 int32_t length);
2649
2650 /**
2651 * Convert the characters in this to UPPER CASE following the conventions of
2652 * the default locale.
2653 * @return A reference to this.
2654 * @stable ICU 2.0
2655 */
2656 UnicodeString& toUpper(void);
2657
2658 /**
2659 * Convert the characters in this to UPPER CASE following the conventions of
2660 * a specific locale.
2661 * @param locale The locale containing the conventions to use.
2662 * @return A reference to this.
2663 * @stable ICU 2.0
2664 */
2665 UnicodeString& toUpper(const Locale& locale);
2666
2667 /**
2668 * Convert the characters in this to lower case following the conventions of
2669 * the default locale.
2670 * @return A reference to this.
2671 * @stable ICU 2.0
2672 */
2673 UnicodeString& toLower(void);
2674
2675 /**
2676 * Convert the characters in this to lower case following the conventions of
2677 * a specific locale.
2678 * @param locale The locale containing the conventions to use.
2679 * @return A reference to this.
2680 * @stable ICU 2.0
2681 */
2682 UnicodeString& toLower(const Locale& locale);
2683
2684#if !UCONFIG_NO_BREAK_ITERATION0
2685
2686 /**
2687 * Titlecase this string, convenience function using the default locale.
2688 *
2689 * Casing is locale-dependent and context-sensitive.
2690 * Titlecasing uses a break iterator to find the first characters of words
2691 * that are to be titlecased. It titlecases those characters and lowercases
2692 * all others.
2693 *
2694 * The titlecase break iterator can be provided to customize for arbitrary
2695 * styles, using rules and dictionaries beyond the standard iterators.
2696 * It may be more efficient to always provide an iterator to avoid
2697 * opening and closing one for each string.
2698 * The standard titlecase iterator for the root locale implements the
2699 * algorithm of Unicode TR 21.
2700 *
2701 * This function uses only the setText(), first() and next() methods of the
2702 * provided break iterator.
2703 *
2704 * @param titleIter A break iterator to find the first characters of words
2705 * that are to be titlecased.
2706 * If none is provided (0), then a standard titlecase
2707 * break iterator is opened.
2708 * Otherwise the provided iterator is set to the string's text.
2709 * @return A reference to this.
2710 * @stable ICU 2.1
2711 */
2712 UnicodeString &toTitle(BreakIterator *titleIter);
2713
2714 /**
2715 * Titlecase this string.
2716 *
2717 * Casing is locale-dependent and context-sensitive.
2718 * Titlecasing uses a break iterator to find the first characters of words
2719 * that are to be titlecased. It titlecases those characters and lowercases
2720 * all others.
2721 *
2722 * The titlecase break iterator can be provided to customize for arbitrary
2723 * styles, using rules and dictionaries beyond the standard iterators.
2724 * It may be more efficient to always provide an iterator to avoid
2725 * opening and closing one for each string.
2726 * The standard titlecase iterator for the root locale implements the
2727 * algorithm of Unicode TR 21.
2728 *
2729 * This function uses only the setText(), first() and next() methods of the
2730 * provided break iterator.
2731 *
2732 * @param titleIter A break iterator to find the first characters of words
2733 * that are to be titlecased.
2734 * If none is provided (0), then a standard titlecase
2735 * break iterator is opened.
2736 * Otherwise the provided iterator is set to the string's text.
2737 * @param locale The locale to consider.
2738 * @return A reference to this.
2739 * @stable ICU 2.1
2740 */
2741 UnicodeString &toTitle(BreakIterator *titleIter, const Locale &locale);
2742
2743 /**
2744 * Titlecase this string, with options.
2745 *
2746 * Casing is locale-dependent and context-sensitive.
2747 * Titlecasing uses a break iterator to find the first characters of words
2748 * that are to be titlecased. It titlecases those characters and lowercases
2749 * all others. (This can be modified with options.)
2750 *
2751 * The titlecase break iterator can be provided to customize for arbitrary
2752 * styles, using rules and dictionaries beyond the standard iterators.
2753 * It may be more efficient to always provide an iterator to avoid
2754 * opening and closing one for each string.
2755 * The standard titlecase iterator for the root locale implements the
2756 * algorithm of Unicode TR 21.
2757 *
2758 * This function uses only the setText(), first() and next() methods of the
2759 * provided break iterator.
2760 *
2761 * @param titleIter A break iterator to find the first characters of words
2762 * that are to be titlecased.
2763 * If none is provided (0), then a standard titlecase
2764 * break iterator is opened.
2765 * Otherwise the provided iterator is set to the string's text.
2766 * @param locale The locale to consider.
2767 * @param options Options bit set, usually 0. See U_TITLECASE_NO_LOWERCASE,
2768 * U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED,
2769 * U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES.
2770 * @return A reference to this.
2771 * @stable ICU 3.8
2772 */
2773 UnicodeString &toTitle(BreakIterator *titleIter, const Locale &locale, uint32_t options);
2774
2775#endif
2776
2777 /**
2778 * Case-folds the characters in this string.
2779 *
2780 * Case-folding is locale-independent and not context-sensitive,
2781 * but there is an option for whether to include or exclude mappings for dotted I
2782 * and dotless i that are marked with 'T' in CaseFolding.txt.
2783 *
2784 * The result may be longer or shorter than the original.
2785 *
2786 * @param options Either U_FOLD_CASE_DEFAULT or U_FOLD_CASE_EXCLUDE_SPECIAL_I
2787 * @return A reference to this.
2788 * @stable ICU 2.0
2789 */
2790 UnicodeString &foldCase(uint32_t options=0 /*U_FOLD_CASE_DEFAULT*/);
2791
2792 //========================================
2793 // Access to the internal buffer
2794 //========================================
2795
2796 /**
2797 * Get a read/write pointer to the internal buffer.
2798 * The buffer is guaranteed to be large enough for at least minCapacity char16_ts,
2799 * writable, and is still owned by the UnicodeString object.
2800 * Calls to getBuffer(minCapacity) must not be nested, and
2801 * must be matched with calls to releaseBuffer(newLength).
2802 * If the string buffer was read-only or shared,
2803 * then it will be reallocated and copied.
2804 *
2805 * An attempted nested call will return 0, and will not further modify the
2806 * state of the UnicodeString object.
2807 * It also returns 0 if the string is bogus.
2808 *
2809 * The actual capacity of the string buffer may be larger than minCapacity.
2810 * getCapacity() returns the actual capacity.
2811 * For many operations, the full capacity should be used to avoid reallocations.
2812 *
2813 * While the buffer is "open" between getBuffer(minCapacity)
2814 * and releaseBuffer(newLength), the following applies:
2815 * - The string length is set to 0.
2816 * - Any read API call on the UnicodeString object will behave like on a 0-length string.
2817 * - Any write API call on the UnicodeString object is disallowed and will have no effect.
2818 * - You can read from and write to the returned buffer.
2819 * - The previous string contents will still be in the buffer;
2820 * if you want to use it, then you need to call length() before getBuffer(minCapacity).
2821 * If the length() was greater than minCapacity, then any contents after minCapacity
2822 * may be lost.
2823 * The buffer contents is not NUL-terminated by getBuffer().
2824 * If length() < getCapacity() then you can terminate it by writing a NUL
2825 * at index length().
2826 * - You must call releaseBuffer(newLength) before and in order to
2827 * return to normal UnicodeString operation.
2828 *
2829 * @param minCapacity the minimum number of char16_ts that are to be available
2830 * in the buffer, starting at the returned pointer;
2831 * default to the current string capacity if minCapacity==-1
2832 * @return a writable pointer to the internal string buffer,
2833 * or nullptr if an error occurs (nested calls, out of memory)
2834 *
2835 * @see releaseBuffer
2836 * @see getTerminatedBuffer()
2837 * @stable ICU 2.0
2838 */
2839 char16_t *getBuffer(int32_t minCapacity);
2840
2841 /**
2842 * Release a read/write buffer on a UnicodeString object with an
2843 * "open" getBuffer(minCapacity).
2844 * This function must be called in a matched pair with getBuffer(minCapacity).
2845 * releaseBuffer(newLength) must be called if and only if a getBuffer(minCapacity) is "open".
2846 *
2847 * It will set the string length to newLength, at most to the current capacity.
2848 * If newLength==-1 then it will set the length according to the
2849 * first NUL in the buffer, or to the capacity if there is no NUL.
2850 *
2851 * After calling releaseBuffer(newLength) the UnicodeString is back to normal operation.
2852 *
2853 * @param newLength the new length of the UnicodeString object;
2854 * defaults to the current capacity if newLength is greater than that;
2855 * if newLength==-1, it defaults to u_strlen(buffer) but not more than
2856 * the current capacity of the string
2857 *
2858 * @see getBuffer(int32_t minCapacity)
2859 * @stable ICU 2.0
2860 */
2861 void releaseBuffer(int32_t newLength=-1);
2862
2863 /**
2864 * Get a read-only pointer to the internal buffer.
2865 * This can be called at any time on a valid UnicodeString.
2866 *
2867 * It returns 0 if the string is bogus, or
2868 * during an "open" getBuffer(minCapacity).
2869 *
2870 * It can be called as many times as desired.
2871 * The pointer that it returns will remain valid until the UnicodeString object is modified,
2872 * at which time the pointer is semantically invalidated and must not be used any more.
2873 *
2874 * The capacity of the buffer can be determined with getCapacity().
2875 * The part after length() may or may not be initialized and valid,
2876 * depending on the history of the UnicodeString object.
2877 *
2878 * The buffer contents is (probably) not NUL-terminated.
2879 * You can check if it is with
2880 * `(s.length() < s.getCapacity() && buffer[s.length()]==0)`.
2881 * (See getTerminatedBuffer().)
2882 *
2883 * The buffer may reside in read-only memory. Its contents must not
2884 * be modified.
2885 *
2886 * @return a read-only pointer to the internal string buffer,
2887 * or nullptr if the string is empty or bogus
2888 *
2889 * @see getBuffer(int32_t minCapacity)
2890 * @see getTerminatedBuffer()
2891 * @stable ICU 2.0
2892 */
2893 inline const char16_t *getBuffer() const;
2894
2895 /**
2896 * Get a read-only pointer to the internal buffer,
2897 * making sure that it is NUL-terminated.
2898 * This can be called at any time on a valid UnicodeString.
2899 *
2900 * It returns 0 if the string is bogus, or
2901 * during an "open" getBuffer(minCapacity), or if the buffer cannot
2902 * be NUL-terminated (because memory allocation failed).
2903 *
2904 * It can be called as many times as desired.
2905 * The pointer that it returns will remain valid until the UnicodeString object is modified,
2906 * at which time the pointer is semantically invalidated and must not be used any more.
2907 *
2908 * The capacity of the buffer can be determined with getCapacity().
2909 * The part after length()+1 may or may not be initialized and valid,
2910 * depending on the history of the UnicodeString object.
2911 *
2912 * The buffer contents is guaranteed to be NUL-terminated.
2913 * getTerminatedBuffer() may reallocate the buffer if a terminating NUL
2914 * is written.
2915 * For this reason, this function is not const, unlike getBuffer().
2916 * Note that a UnicodeString may also contain NUL characters as part of its contents.
2917 *
2918 * The buffer may reside in read-only memory. Its contents must not
2919 * be modified.
2920 *
2921 * @return a read-only pointer to the internal string buffer,
2922 * or 0 if the string is empty or bogus
2923 *
2924 * @see getBuffer(int32_t minCapacity)
2925 * @see getBuffer()
2926 * @stable ICU 2.2
2927 */
2928 const char16_t *getTerminatedBuffer();
2929
2930 //========================================
2931 // Constructors
2932 //========================================
2933
2934 /** Construct an empty UnicodeString.
2935 * @stable ICU 2.0
2936 */
2937 inline UnicodeString();
2938
2939 /**
2940 * Construct a UnicodeString with capacity to hold `capacity` char16_ts
2941 * @param capacity the number of char16_ts this UnicodeString should hold
2942 * before a resize is necessary; if count is greater than 0 and count
2943 * code points c take up more space than capacity, then capacity is adjusted
2944 * accordingly.
2945 * @param c is used to initially fill the string
2946 * @param count specifies how many code points c are to be written in the
2947 * string
2948 * @stable ICU 2.0
2949 */
2950 UnicodeString(int32_t capacity, UChar32 c, int32_t count);
2951
2952 /**
2953 * Single char16_t (code unit) constructor.
2954 *
2955 * It is recommended to mark this constructor "explicit" by
2956 * `-DUNISTR_FROM_CHAR_EXPLICIT=explicit`
2957 * on the compiler command line or similar.
2958 * @param ch the character to place in the UnicodeString
2959 * @stable ICU 2.0
2960 */
2961 UNISTR_FROM_CHAR_EXPLICITexplicit UnicodeString(char16_t ch);
2962
2963 /**
2964 * Single UChar32 (code point) constructor.
2965 *
2966 * It is recommended to mark this constructor "explicit" by
2967 * `-DUNISTR_FROM_CHAR_EXPLICIT=explicit`
2968 * on the compiler command line or similar.
2969 * @param ch the character to place in the UnicodeString
2970 * @stable ICU 2.0
2971 */
2972 UNISTR_FROM_CHAR_EXPLICITexplicit UnicodeString(UChar32 ch);
2973
2974 /**
2975 * char16_t* constructor.
2976 *
2977 * It is recommended to mark this constructor "explicit" by
2978 * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
2979 * on the compiler command line or similar.
2980 * @param text The characters to place in the UnicodeString. `text`
2981 * must be NULL (U+0000) terminated.
2982 * @stable ICU 2.0
2983 */
2984 UNISTR_FROM_STRING_EXPLICITexplicit UnicodeString(const char16_t *text);
2985
2986#if !U_CHAR16_IS_TYPEDEF0
2987 /**
2988 * uint16_t * constructor.
2989 * Delegates to UnicodeString(const char16_t *).
2990 *
2991 * It is recommended to mark this constructor "explicit" by
2992 * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
2993 * on the compiler command line or similar.
2994 * @param text NUL-terminated UTF-16 string
2995 * @stable ICU 59
2996 */
2997 UNISTR_FROM_STRING_EXPLICITexplicit UnicodeString(const uint16_t *text) :
2998 UnicodeString(ConstChar16Ptr(text)) {}
2999#endif
3000
3001#if U_SIZEOF_WCHAR_T4==2 || defined(U_IN_DOXYGEN)
3002 /**
3003 * wchar_t * constructor.
3004 * (Only defined if U_SIZEOF_WCHAR_T==2.)
3005 * Delegates to UnicodeString(const char16_t *).
3006 *
3007 * It is recommended to mark this constructor "explicit" by
3008 * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
3009 * on the compiler command line or similar.
3010 * @param text NUL-terminated UTF-16 string
3011 * @stable ICU 59
3012 */
3013 UNISTR_FROM_STRING_EXPLICITexplicit UnicodeString(const wchar_t *text) :
3014 UnicodeString(ConstChar16Ptr(text)) {}
3015#endif
3016
3017 /**
3018 * nullptr_t constructor.
3019 * Effectively the same as the default constructor, makes an empty string object.
3020 *
3021 * It is recommended to mark this constructor "explicit" by
3022 * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
3023 * on the compiler command line or similar.
3024 * @param text nullptr
3025 * @stable ICU 59
3026 */
3027 UNISTR_FROM_STRING_EXPLICITexplicit inline UnicodeString(const std::nullptr_t text);
3028
3029 /**
3030 * char16_t* constructor.
3031 * @param text The characters to place in the UnicodeString.
3032 * @param textLength The number of Unicode characters in `text`
3033 * to copy.
3034 * @stable ICU 2.0
3035 */
3036 UnicodeString(const char16_t *text,
3037 int32_t textLength);
3038
3039#if !U_CHAR16_IS_TYPEDEF0
3040 /**
3041 * uint16_t * constructor.
3042 * Delegates to UnicodeString(const char16_t *, int32_t).
3043 * @param text UTF-16 string
3044 * @param textLength string length
3045 * @stable ICU 59
3046 */
3047 UnicodeString(const uint16_t *text, int32_t textLength) :
3048 UnicodeString(ConstChar16Ptr(text), textLength) {}
3049#endif
3050
3051#if U_SIZEOF_WCHAR_T4==2 || defined(U_IN_DOXYGEN)
3052 /**
3053 * wchar_t * constructor.
3054 * (Only defined if U_SIZEOF_WCHAR_T==2.)
3055 * Delegates to UnicodeString(const char16_t *, int32_t).
3056 * @param text NUL-terminated UTF-16 string
3057 * @param textLength string length
3058 * @stable ICU 59
3059 */
3060 UnicodeString(const wchar_t *text, int32_t textLength) :
3061 UnicodeString(ConstChar16Ptr(text), textLength) {}
3062#endif
3063
3064 /**
3065 * nullptr_t constructor.
3066 * Effectively the same as the default constructor, makes an empty string object.
3067 * @param text nullptr
3068 * @param textLength ignored
3069 * @stable ICU 59
3070 */
3071 inline UnicodeString(const std::nullptr_t text, int32_t textLength);
3072
3073 /**
3074 * Readonly-aliasing char16_t* constructor.
3075 * The text will be used for the UnicodeString object, but
3076 * it will not be released when the UnicodeString is destroyed.
3077 * This has copy-on-write semantics:
3078 * When the string is modified, then the buffer is first copied into
3079 * newly allocated memory.
3080 * The aliased buffer is never modified.
3081 *
3082 * In an assignment to another UnicodeString, when using the copy constructor
3083 * or the assignment operator, the text will be copied.
3084 * When using fastCopyFrom(), the text will be aliased again,
3085 * so that both strings then alias the same readonly-text.
3086 *
3087 * @param isTerminated specifies if `text` is `NUL`-terminated.
3088 * This must be true if `textLength==-1`.
3089 * @param text The characters to alias for the UnicodeString.
3090 * @param textLength The number of Unicode characters in `text` to alias.
3091 * If -1, then this constructor will determine the length
3092 * by calling `u_strlen()`.
3093 * @stable ICU 2.0
3094 */
3095 UnicodeString(UBool isTerminated,
3096 ConstChar16Ptr text,
3097 int32_t textLength);
3098
3099 /**
3100 * Writable-aliasing char16_t* constructor.
3101 * The text will be used for the UnicodeString object, but
3102 * it will not be released when the UnicodeString is destroyed.
3103 * This has write-through semantics:
3104 * For as long as the capacity of the buffer is sufficient, write operations
3105 * will directly affect the buffer. When more capacity is necessary, then
3106 * a new buffer will be allocated and the contents copied as with regularly
3107 * constructed strings.
3108 * In an assignment to another UnicodeString, the buffer will be copied.
3109 * The extract(Char16Ptr dst) function detects whether the dst pointer is the same
3110 * as the string buffer itself and will in this case not copy the contents.
3111 *
3112 * @param buffer The characters to alias for the UnicodeString.
3113 * @param buffLength The number of Unicode characters in `buffer` to alias.
3114 * @param buffCapacity The size of `buffer` in char16_ts.
3115 * @stable ICU 2.0
3116 */
3117 UnicodeString(char16_t *buffer, int32_t buffLength, int32_t buffCapacity);
3118
3119#if !U_CHAR16_IS_TYPEDEF0
3120 /**
3121 * Writable-aliasing uint16_t * constructor.
3122 * Delegates to UnicodeString(const char16_t *, int32_t, int32_t).
3123 * @param buffer writable buffer of/for UTF-16 text
3124 * @param buffLength length of the current buffer contents
3125 * @param buffCapacity buffer capacity
3126 * @stable ICU 59
3127 */
3128 UnicodeString(uint16_t *buffer, int32_t buffLength, int32_t buffCapacity) :
3129 UnicodeString(Char16Ptr(buffer), buffLength, buffCapacity) {}
3130#endif
3131
3132#if U_SIZEOF_WCHAR_T4==2 || defined(U_IN_DOXYGEN)
3133 /**
3134 * Writable-aliasing wchar_t * constructor.
3135 * (Only defined if U_SIZEOF_WCHAR_T==2.)
3136 * Delegates to UnicodeString(const char16_t *, int32_t, int32_t).
3137 * @param buffer writable buffer of/for UTF-16 text
3138 * @param buffLength length of the current buffer contents
3139 * @param buffCapacity buffer capacity
3140 * @stable ICU 59
3141 */
3142 UnicodeString(wchar_t *buffer, int32_t buffLength, int32_t buffCapacity) :
3143 UnicodeString(Char16Ptr(buffer), buffLength, buffCapacity) {}
3144#endif
3145
3146 /**
3147 * Writable-aliasing nullptr_t constructor.
3148 * Effectively the same as the default constructor, makes an empty string object.
3149 * @param buffer nullptr
3150 * @param buffLength ignored
3151 * @param buffCapacity ignored
3152 * @stable ICU 59
3153 */
3154 inline UnicodeString(std::nullptr_t buffer, int32_t buffLength, int32_t buffCapacity);
3155
3156#if U_CHARSET_IS_UTF81 || !UCONFIG_NO_CONVERSION0
3157
3158 /**
3159 * char* constructor.
3160 * Uses the default converter (and thus depends on the ICU conversion code)
3161 * unless U_CHARSET_IS_UTF8 is set to 1.
3162 *
3163 * For ASCII (really "invariant character") strings it is more efficient to use
3164 * the constructor that takes a US_INV (for its enum EInvariant).
3165 * For ASCII (invariant-character) string literals, see UNICODE_STRING and
3166 * UNICODE_STRING_SIMPLE.
3167 *
3168 * It is recommended to mark this constructor "explicit" by
3169 * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
3170 * on the compiler command line or similar.
3171 * @param codepageData an array of bytes, null-terminated,
3172 * in the platform's default codepage.
3173 * @stable ICU 2.0
3174 * @see UNICODE_STRING
3175 * @see UNICODE_STRING_SIMPLE
3176 */
3177 UNISTR_FROM_STRING_EXPLICITexplicit UnicodeString(const char *codepageData);
3178
3179 /**
3180 * char* constructor.
3181 * Uses the default converter (and thus depends on the ICU conversion code)
3182 * unless U_CHARSET_IS_UTF8 is set to 1.
3183 * @param codepageData an array of bytes in the platform's default codepage.
3184 * @param dataLength The number of bytes in `codepageData`.
3185 * @stable ICU 2.0
3186 */
3187 UnicodeString(const char *codepageData, int32_t dataLength);
3188
3189#endif
3190
3191#if !UCONFIG_NO_CONVERSION0
3192
3193 /**
3194 * char* constructor.
3195 * @param codepageData an array of bytes, null-terminated
3196 * @param codepage the encoding of `codepageData`. The special
3197 * value 0 for `codepage` indicates that the text is in the
3198 * platform's default codepage.
3199 *
3200 * If `codepage` is an empty string (`""`),
3201 * then a simple conversion is performed on the codepage-invariant
3202 * subset ("invariant characters") of the platform encoding. See utypes.h.
3203 * Recommendation: For invariant-character strings use the constructor
3204 * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
3205 * because it avoids object code dependencies of UnicodeString on
3206 * the conversion code.
3207 *
3208 * @stable ICU 2.0
3209 */
3210 UnicodeString(const char *codepageData, const char *codepage);
3211
3212 /**
3213 * char* constructor.
3214 * @param codepageData an array of bytes.
3215 * @param dataLength The number of bytes in `codepageData`.
3216 * @param codepage the encoding of `codepageData`. The special
3217 * value 0 for `codepage` indicates that the text is in the
3218 * platform's default codepage.
3219 * If `codepage` is an empty string (`""`),
3220 * then a simple conversion is performed on the codepage-invariant
3221 * subset ("invariant characters") of the platform encoding. See utypes.h.
3222 * Recommendation: For invariant-character strings use the constructor
3223 * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
3224 * because it avoids object code dependencies of UnicodeString on
3225 * the conversion code.
3226 *
3227 * @stable ICU 2.0
3228 */
3229 UnicodeString(const char *codepageData, int32_t dataLength, const char *codepage);
3230
3231 /**
3232 * char * / UConverter constructor.
3233 * This constructor uses an existing UConverter object to
3234 * convert the codepage string to Unicode and construct a UnicodeString
3235 * from that.
3236 *
3237 * The converter is reset at first.
3238 * If the error code indicates a failure before this constructor is called,
3239 * or if an error occurs during conversion or construction,
3240 * then the string will be bogus.
3241 *
3242 * This function avoids the overhead of opening and closing a converter if
3243 * multiple strings are constructed.
3244 *
3245 * @param src input codepage string
3246 * @param srcLength length of the input string, can be -1 for NUL-terminated strings
3247 * @param cnv converter object (ucnv_resetToUnicode() will be called),
3248 * can be NULL for the default converter
3249 * @param errorCode normal ICU error code
3250 * @stable ICU 2.0
3251 */
3252 UnicodeString(
3253 const char *src, int32_t srcLength,
3254 UConverter *cnv,
3255 UErrorCode &errorCode);
3256
3257#endif
3258
3259 /**
3260 * Constructs a Unicode string from an invariant-character char * string.
3261 * About invariant characters see utypes.h.
3262 * This constructor has no runtime dependency on conversion code and is
3263 * therefore recommended over ones taking a charset name string
3264 * (where the empty string "" indicates invariant-character conversion).
3265 *
3266 * Use the macro US_INV as the third, signature-distinguishing parameter.
3267 *
3268 * For example:
3269 * \code
3270 * void fn(const char *s) {
3271 * UnicodeString ustr(s, -1, US_INV);
3272 * // use ustr ...
3273 * }
3274 * \endcode
3275 * @param src String using only invariant characters.
3276 * @param textLength Length of src, or -1 if NUL-terminated.
3277 * @param inv Signature-distinguishing parameter, use US_INV.
3278 *
3279 * @see US_INV
3280 * @stable ICU 3.2
3281 */
3282 UnicodeString(const char *src, int32_t textLength, enum EInvariant inv);
3283
3284
3285 /**
3286 * Copy constructor.
3287 *
3288 * Starting with ICU 2.4, the assignment operator and the copy constructor
3289 * allocate a new buffer and copy the buffer contents even for readonly aliases.
3290 * By contrast, the fastCopyFrom() function implements the old,
3291 * more efficient but less safe behavior
3292 * of making this string also a readonly alias to the same buffer.
3293 *
3294 * If the source object has an "open" buffer from getBuffer(minCapacity),
3295 * then the copy is an empty string.
3296 *
3297 * @param that The UnicodeString object to copy.
3298 * @stable ICU 2.0
3299 * @see fastCopyFrom
3300 */
3301 UnicodeString(const UnicodeString& that);
3302
3303 /**
3304 * Move constructor; might leave src in bogus state.
3305 * This string will have the same contents and state that the source string had.
3306 * @param src source string
3307 * @stable ICU 56
3308 */
3309 UnicodeString(UnicodeString &&src) U_NOEXCEPTnoexcept;
3310
3311 /**
3312 * 'Substring' constructor from tail of source string.
3313 * @param src The UnicodeString object to copy.
3314 * @param srcStart The offset into `src` at which to start copying.
3315 * @stable ICU 2.2
3316 */
3317 UnicodeString(const UnicodeString& src, int32_t srcStart);
3318
3319 /**
3320 * 'Substring' constructor from subrange of source string.
3321 * @param src The UnicodeString object to copy.
3322 * @param srcStart The offset into `src` at which to start copying.
3323 * @param srcLength The number of characters from `src` to copy.
3324 * @stable ICU 2.2
3325 */
3326 UnicodeString(const UnicodeString& src, int32_t srcStart, int32_t srcLength);
3327
3328 /**
3329 * Clone this object, an instance of a subclass of Replaceable.
3330 * Clones can be used concurrently in multiple threads.
3331 * If a subclass does not implement clone(), or if an error occurs,
3332 * then NULL is returned.
3333 * The caller must delete the clone.
3334 *
3335 * @return a clone of this object
3336 *
3337 * @see Replaceable::clone
3338 * @see getDynamicClassID
3339 * @stable ICU 2.6
3340 */
3341 virtual UnicodeString *clone() const override;
3342
3343 /** Destructor.
3344 * @stable ICU 2.0
3345 */
3346 virtual ~UnicodeString();
3347
3348 /**
3349 * Create a UnicodeString from a UTF-8 string.
3350 * Illegal input is replaced with U+FFFD. Otherwise, errors result in a bogus string.
3351 * Calls u_strFromUTF8WithSub().
3352 *
3353 * @param utf8 UTF-8 input string.
3354 * Note that a StringPiece can be implicitly constructed
3355 * from a std::string or a NUL-terminated const char * string.
3356 * @return A UnicodeString with equivalent UTF-16 contents.
3357 * @see toUTF8
3358 * @see toUTF8String
3359 * @stable ICU 4.2
3360 */
3361 static UnicodeString fromUTF8(StringPiece utf8);
3362
3363 /**
3364 * Create a UnicodeString from a UTF-32 string.
3365 * Illegal input is replaced with U+FFFD. Otherwise, errors result in a bogus string.
3366 * Calls u_strFromUTF32WithSub().
3367 *
3368 * @param utf32 UTF-32 input string. Must not be NULL.
3369 * @param length Length of the input string, or -1 if NUL-terminated.
3370 * @return A UnicodeString with equivalent UTF-16 contents.
3371 * @see toUTF32
3372 * @stable ICU 4.2
3373 */
3374 static UnicodeString fromUTF32(const UChar32 *utf32, int32_t length);
3375
3376 /* Miscellaneous operations */
3377
3378 /**
3379 * Unescape a string of characters and return a string containing
3380 * the result. The following escape sequences are recognized:
3381 *
3382 * \\uhhhh 4 hex digits; h in [0-9A-Fa-f]
3383 * \\Uhhhhhhhh 8 hex digits
3384 * \\xhh 1-2 hex digits
3385 * \\ooo 1-3 octal digits; o in [0-7]
3386 * \\cX control-X; X is masked with 0x1F
3387 *
3388 * as well as the standard ANSI C escapes:
3389 *
3390 * \\a => U+0007, \\b => U+0008, \\t => U+0009, \\n => U+000A,
3391 * \\v => U+000B, \\f => U+000C, \\r => U+000D, \\e => U+001B,
3392 * \\" => U+0022, \\' => U+0027, \\? => U+003F, \\\\ => U+005C
3393 *
3394 * Anything else following a backslash is generically escaped. For
3395 * example, "[a\\-z]" returns "[a-z]".
3396 *
3397 * If an escape sequence is ill-formed, this method returns an empty
3398 * string. An example of an ill-formed sequence is "\\u" followed by
3399 * fewer than 4 hex digits.
3400 *
3401 * This function is similar to u_unescape() but not identical to it.
3402 * The latter takes a source char*, so it does escape recognition
3403 * and also invariant conversion.
3404 *
3405 * @return a string with backslash escapes interpreted, or an
3406 * empty string on error.
3407 * @see UnicodeString#unescapeAt()
3408 * @see u_unescape()
3409 * @see u_unescapeAt()
3410 * @stable ICU 2.0
3411 */
3412 UnicodeString unescape() const;
3413
3414 /**
3415 * Unescape a single escape sequence and return the represented
3416 * character. See unescape() for a listing of the recognized escape
3417 * sequences. The character at offset-1 is assumed (without
3418 * checking) to be a backslash. If the escape sequence is
3419 * ill-formed, or the offset is out of range, U_SENTINEL=-1 is
3420 * returned.
3421 *
3422 * @param offset an input output parameter. On input, it is the
3423 * offset into this string where the escape sequence is located,
3424 * after the initial backslash. On output, it is advanced after the
3425 * last character parsed. On error, it is not advanced at all.
3426 * @return the character represented by the escape sequence at
3427 * offset, or U_SENTINEL=-1 on error.
3428 * @see UnicodeString#unescape()
3429 * @see u_unescape()
3430 * @see u_unescapeAt()
3431 * @stable ICU 2.0
3432 */
3433 UChar32 unescapeAt(int32_t &offset) const;
3434
3435 /**
3436 * ICU "poor man's RTTI", returns a UClassID for this class.
3437 *
3438 * @stable ICU 2.2
3439 */
3440 static UClassID U_EXPORT2 getStaticClassID();
3441
3442 /**
3443 * ICU "poor man's RTTI", returns a UClassID for the actual class.
3444 *
3445 * @stable ICU 2.2
3446 */
3447 virtual UClassID getDynamicClassID() const override;
3448
3449 //========================================
3450 // Implementation methods
3451 //========================================
3452
3453protected:
3454 /**
3455 * Implement Replaceable::getLength() (see jitterbug 1027).
3456 * @stable ICU 2.4
3457 */
3458 virtual int32_t getLength() const override;
3459
3460 /**
3461 * The change in Replaceable to use virtual getCharAt() allows
3462 * UnicodeString::charAt() to be inline again (see jitterbug 709).
3463 * @stable ICU 2.4
3464 */
3465 virtual char16_t getCharAt(int32_t offset) const override;
3466
3467 /**
3468 * The change in Replaceable to use virtual getChar32At() allows
3469 * UnicodeString::char32At() to be inline again (see jitterbug 709).
3470 * @stable ICU 2.4
3471 */
3472 virtual UChar32 getChar32At(int32_t offset) const override;
3473
3474private:
3475 // For char* constructors. Could be made public.
3476 UnicodeString &setToUTF8(StringPiece utf8);
3477 // For extract(char*).
3478 // We could make a toUTF8(target, capacity, errorCode) public but not
3479 // this version: New API will be cleaner if we make callers create substrings
3480 // rather than having start+length on every method,
3481 // and it should take a UErrorCode&.
3482 int32_t
3483 toUTF8(int32_t start, int32_t len,
3484 char *target, int32_t capacity) const;
3485
3486 /**
3487 * Internal string contents comparison, called by operator==.
3488 * Requires: this & text not bogus and have same lengths.
3489 */
3490 UBool doEquals(const UnicodeString &text, int32_t len) const;
3491
3492 inline int8_t
3493 doCompare(int32_t start,
3494 int32_t length,
3495 const UnicodeString& srcText,
3496 int32_t srcStart,
3497 int32_t srcLength) const;
3498
3499 int8_t doCompare(int32_t start,
3500 int32_t length,
3501 const char16_t *srcChars,
3502 int32_t srcStart,
3503 int32_t srcLength) const;
3504
3505 inline int8_t
3506 doCompareCodePointOrder(int32_t start,
3507 int32_t length,
3508 const UnicodeString& srcText,
3509 int32_t srcStart,
3510 int32_t srcLength) const;
3511
3512 int8_t doCompareCodePointOrder(int32_t start,
3513 int32_t length,
3514 const char16_t *srcChars,
3515 int32_t srcStart,
3516 int32_t srcLength) const;
3517
3518 inline int8_t
3519 doCaseCompare(int32_t start,
3520 int32_t length,
3521 const UnicodeString &srcText,
3522 int32_t srcStart,
3523 int32_t srcLength,
3524 uint32_t options) const;
3525
3526 int8_t
3527 doCaseCompare(int32_t start,
3528 int32_t length,
3529 const char16_t *srcChars,
3530 int32_t srcStart,
3531 int32_t srcLength,
3532 uint32_t options) const;
3533
3534 int32_t doIndexOf(char16_t c,
3535 int32_t start,
3536 int32_t length) const;
3537
3538 int32_t doIndexOf(UChar32 c,
3539 int32_t start,
3540 int32_t length) const;
3541
3542 int32_t doLastIndexOf(char16_t c,
3543 int32_t start,
3544 int32_t length) const;
3545
3546 int32_t doLastIndexOf(UChar32 c,
3547 int32_t start,
3548 int32_t length) const;
3549
3550 void doExtract(int32_t start,
3551 int32_t length,
3552 char16_t *dst,
3553 int32_t dstStart) const;
3554
3555 inline void doExtract(int32_t start,
3556 int32_t length,
3557 UnicodeString& target) const;
3558
3559 inline char16_t doCharAt(int32_t offset) const;
3560
3561 UnicodeString& doReplace(int32_t start,
3562 int32_t length,
3563 const UnicodeString& srcText,
3564 int32_t srcStart,
3565 int32_t srcLength);
3566
3567 UnicodeString& doReplace(int32_t start,
3568 int32_t length,
3569 const char16_t *srcChars,
3570 int32_t srcStart,
3571 int32_t srcLength);
3572
3573 UnicodeString& doAppend(const UnicodeString& src, int32_t srcStart, int32_t srcLength);
3574 UnicodeString& doAppend(const char16_t *srcChars, int32_t srcStart, int32_t srcLength);
3575
3576 UnicodeString& doReverse(int32_t start,
3577 int32_t length);
3578
3579 // calculate hash code
3580 int32_t doHashCode(void) const;
3581
3582 // get pointer to start of array
3583 // these do not check for kOpenGetBuffer, unlike the public getBuffer() function
3584 inline char16_t* getArrayStart(void);
3585 inline const char16_t* getArrayStart(void) const;
3586
3587 inline UBool hasShortLength() const;
3588 inline int32_t getShortLength() const;
3589
3590 // A UnicodeString object (not necessarily its current buffer)
3591 // is writable unless it isBogus() or it has an "open" getBuffer(minCapacity).
3592 inline UBool isWritable() const;
3593
3594 // Is the current buffer writable?
3595 inline UBool isBufferWritable() const;
3596
3597 // None of the following does releaseArray().
3598 inline void setZeroLength();
3599 inline void setShortLength(int32_t len);
3600 inline void setLength(int32_t len);
3601 inline void setToEmpty();
3602 inline void setArray(char16_t *array, int32_t len, int32_t capacity); // sets length but not flags
3603
3604 // allocate the array; result may be the stack buffer
3605 // sets refCount to 1 if appropriate
3606 // sets fArray, fCapacity, and flags
3607 // sets length to 0
3608 // returns boolean for success or failure
3609 UBool allocate(int32_t capacity);
3610
3611 // release the array if owned
3612 void releaseArray(void);
3613
3614 // turn a bogus string into an empty one
3615 void unBogus();
3616
3617 // implements assignment operator, copy constructor, and fastCopyFrom()
3618 UnicodeString &copyFrom(const UnicodeString &src, UBool fastCopy=false);
3619
3620 // Copies just the fields without memory management.
3621 void copyFieldsFrom(UnicodeString &src, UBool setSrcToBogus) U_NOEXCEPTnoexcept;
3622
3623 // Pin start and limit to acceptable values.
3624 inline void pinIndex(int32_t& start) const;
3625 inline void pinIndices(int32_t& start,
3626 int32_t& length) const;
3627
3628#if !UCONFIG_NO_CONVERSION0
3629
3630 /* Internal extract() using UConverter. */
3631 int32_t doExtract(int32_t start, int32_t length,
3632 char *dest, int32_t destCapacity,
3633 UConverter *cnv,
3634 UErrorCode &errorCode) const;
3635
3636 /*
3637 * Real constructor for converting from codepage data.
3638 * It assumes that it is called with !fRefCounted.
3639 *
3640 * If `codepage==0`, then the default converter
3641 * is used for the platform encoding.
3642 * If `codepage` is an empty string (`""`),
3643 * then a simple conversion is performed on the codepage-invariant
3644 * subset ("invariant characters") of the platform encoding. See utypes.h.
3645 */
3646 void doCodepageCreate(const char *codepageData,
3647 int32_t dataLength,
3648 const char *codepage);
3649
3650 /*
3651 * Worker function for creating a UnicodeString from
3652 * a codepage string using a UConverter.
3653 */
3654 void
3655 doCodepageCreate(const char *codepageData,
3656 int32_t dataLength,
3657 UConverter *converter,
3658 UErrorCode &status);
3659
3660#endif
3661
3662 /*
3663 * This function is called when write access to the array
3664 * is necessary.
3665 *
3666 * We need to make a copy of the array if
3667 * the buffer is read-only, or
3668 * the buffer is refCounted (shared), and refCount>1, or
3669 * the buffer is too small.
3670 *
3671 * Return false if memory could not be allocated.
3672 */
3673 UBool cloneArrayIfNeeded(int32_t newCapacity = -1,
3674 int32_t growCapacity = -1,
3675 UBool doCopyArray = true,
3676 int32_t **pBufferToDelete = 0,
3677 UBool forceClone = false);
3678
3679 /**
3680 * Common function for UnicodeString case mappings.
3681 * The stringCaseMapper has the same type UStringCaseMapper
3682 * as in ustr_imp.h for ustrcase_map().
3683 */
3684 UnicodeString &
3685 caseMap(int32_t caseLocale, uint32_t options,
3686#if !UCONFIG_NO_BREAK_ITERATION0
3687 BreakIterator *iter,
3688#endif
3689 UStringCaseMapper *stringCaseMapper);
3690
3691 // ref counting
3692 void addRef(void);
3693 int32_t removeRef(void);
3694 int32_t refCount(void) const;
3695
3696 // constants
3697 enum {
3698 /**
3699 * Size of stack buffer for short strings.
3700 * Must be at least U16_MAX_LENGTH for the single-code point constructor to work.
3701 * @see UNISTR_OBJECT_SIZE
3702 */
3703 US_STACKBUF_SIZE=(int32_t)(UNISTR_OBJECT_SIZE64-sizeof(void *)-2)/U_SIZEOF_UCHAR2,
3704 kInvalidUChar=0xffff, // U+FFFF returned by charAt(invalid index)
3705 kInvalidHashCode=0, // invalid hash code
3706 kEmptyHashCode=1, // hash code for empty string
3707
3708 // bit flag values for fLengthAndFlags
3709 kIsBogus=1, // this string is bogus, i.e., not valid or NULL
3710 kUsingStackBuffer=2,// using fUnion.fStackFields instead of fUnion.fFields
3711 kRefCounted=4, // there is a refCount field before the characters in fArray
3712 kBufferIsReadonly=8,// do not write to this buffer
3713 kOpenGetBuffer=16, // getBuffer(minCapacity) was called (is "open"),
3714 // and releaseBuffer(newLength) must be called
3715 kAllStorageFlags=0x1f,
3716
3717 kLengthShift=5, // remaining 11 bits for non-negative short length, or negative if long
3718 kLength1=1<<kLengthShift,
3719 kMaxShortLength=0x3ff, // max non-negative short length (leaves top bit 0)
3720 kLengthIsLarge=0xffe0, // short length < 0, real length is in fUnion.fFields.fLength
3721
3722 // combined values for convenience
3723 kShortString=kUsingStackBuffer,
3724 kLongString=kRefCounted,
3725 kReadonlyAlias=kBufferIsReadonly,
3726 kWritableAlias=0
3727 };
3728
3729 friend class UnicodeStringAppendable;
3730
3731 union StackBufferOrFields; // forward declaration necessary before friend declaration
3732 friend union StackBufferOrFields; // make US_STACKBUF_SIZE visible inside fUnion
3733
3734 /*
3735 * The following are all the class fields that are stored
3736 * in each UnicodeString object.
3737 * Note that UnicodeString has virtual functions,
3738 * therefore there is an implicit vtable pointer
3739 * as the first real field.
3740 * The fields should be aligned such that no padding is necessary.
3741 * On 32-bit machines, the size should be 32 bytes,
3742 * on 64-bit machines (8-byte pointers), it should be 40 bytes.
3743 *
3744 * We use a hack to achieve this.
3745 *
3746 * With at least some compilers, each of the following is forced to
3747 * a multiple of sizeof(pointer) [the largest field base unit here is a data pointer],
3748 * rounded up with additional padding if the fields do not already fit that requirement:
3749 * - sizeof(class UnicodeString)
3750 * - offsetof(UnicodeString, fUnion)
3751 * - sizeof(fUnion)
3752 * - sizeof(fStackFields)
3753 *
3754 * We optimize for the longest possible internal buffer for short strings.
3755 * fUnion.fStackFields begins with 2 bytes for storage flags
3756 * and the length of relatively short strings,
3757 * followed by the buffer for short string contents.
3758 * There is no padding inside fStackFields.
3759 *
3760 * Heap-allocated and aliased strings use fUnion.fFields.
3761 * Both fStackFields and fFields must begin with the same fields for flags and short length,
3762 * that is, those must have the same memory offsets inside the object,
3763 * because the flags must be inspected in order to decide which half of fUnion is being used.
3764 * We assume that the compiler does not reorder the fields.
3765 *
3766 * (Padding at the end of fFields is ok:
3767 * As long as it is no larger than fStackFields, it is not wasted space.)
3768 *
3769 * For some of the history of the UnicodeString class fields layout, see
3770 * - ICU ticket #11551 "longer UnicodeString contents in stack buffer"
3771 * - ICU ticket #11336 "UnicodeString: recombine stack buffer arrays"
3772 * - ICU ticket #8322 "why is sizeof(UnicodeString)==48?"
3773 */
3774 // (implicit) *vtable;
3775 union StackBufferOrFields {
3776 // fStackFields is used iff (fLengthAndFlags&kUsingStackBuffer) else fFields is used.
3777 // Each struct of the union must begin with fLengthAndFlags.
3778 struct {
3779 int16_t fLengthAndFlags; // bit fields: see constants above
3780 char16_t fBuffer[US_STACKBUF_SIZE]; // buffer for short strings
3781 } fStackFields;
3782 struct {
3783 int16_t fLengthAndFlags; // bit fields: see constants above
3784 int32_t fLength; // number of characters in fArray if >127; else undefined
3785 int32_t fCapacity; // capacity of fArray (in char16_ts)
3786 // array pointer last to minimize padding for machines with P128 data model
3787 // or pointer sizes that are not a power of 2
3788 char16_t *fArray; // the Unicode data
3789 } fFields;
3790 } fUnion;
3791};
3792
3793/**
3794 * Create a new UnicodeString with the concatenation of two others.
3795 *
3796 * @param s1 The first string to be copied to the new one.
3797 * @param s2 The second string to be copied to the new one, after s1.
3798 * @return UnicodeString(s1).append(s2)
3799 * @stable ICU 2.8
3800 */
3801U_COMMON_API UnicodeString U_EXPORT2
3802operator+ (const UnicodeString &s1, const UnicodeString &s2);
3803
3804//========================================
3805// Inline members
3806//========================================
3807
3808//========================================
3809// Privates
3810//========================================
3811
3812inline void
3813UnicodeString::pinIndex(int32_t& start) const
3814{
3815 // pin index
3816 if(start < 0) {
3817 start = 0;
3818 } else if(start > length()) {
3819 start = length();
3820 }
3821}
3822
3823inline void
3824UnicodeString::pinIndices(int32_t& start,
3825 int32_t& _length) const
3826{
3827 // pin indices
3828 int32_t len = length();
3829 if(start < 0) {
3830 start = 0;
3831 } else if(start > len) {
3832 start = len;
3833 }
3834 if(_length < 0) {
3835 _length = 0;
3836 } else if(_length > (len - start)) {
3837 _length = (len - start);
3838 }
3839}
3840
3841inline char16_t*
3842UnicodeString::getArrayStart() {
3843 return (fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) ?
3844 fUnion.fStackFields.fBuffer : fUnion.fFields.fArray;
3845}
3846
3847inline const char16_t*
3848UnicodeString::getArrayStart() const {
3849 return (fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) ?
3850 fUnion.fStackFields.fBuffer : fUnion.fFields.fArray;
3851}
3852
3853//========================================
3854// Default constructor
3855//========================================
3856
3857inline
3858UnicodeString::UnicodeString() {
3859 fUnion.fStackFields.fLengthAndFlags=kShortString;
3860}
3861
3862inline UnicodeString::UnicodeString(const std::nullptr_t /*text*/) {
3863 fUnion.fStackFields.fLengthAndFlags=kShortString;
3864}
3865
3866inline UnicodeString::UnicodeString(const std::nullptr_t /*text*/, int32_t /*length*/) {
3867 fUnion.fStackFields.fLengthAndFlags=kShortString;
3868}
3869
3870inline UnicodeString::UnicodeString(std::nullptr_t /*buffer*/, int32_t /*buffLength*/, int32_t /*buffCapacity*/) {
3871 fUnion.fStackFields.fLengthAndFlags=kShortString;
3872}
3873
3874//========================================
3875// Read-only implementation methods
3876//========================================
3877inline UBool
3878UnicodeString::hasShortLength() const {
3879 return fUnion.fFields.fLengthAndFlags>=0;
3880}
3881
3882inline int32_t
3883UnicodeString::getShortLength() const {
3884 // fLengthAndFlags must be non-negative -> short length >= 0
3885 // and arithmetic or logical shift does not matter.
3886 return fUnion.fFields.fLengthAndFlags>>kLengthShift;
3887}
3888
3889inline int32_t
3890UnicodeString::length() const {
3891 return hasShortLength() ? getShortLength() : fUnion.fFields.fLength;
3892}
3893
3894inline int32_t
3895UnicodeString::getCapacity() const {
3896 return (fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) ?
3897 US_STACKBUF_SIZE : fUnion.fFields.fCapacity;
3898}
3899
3900inline int32_t
3901UnicodeString::hashCode() const
3902{ return doHashCode(); }
3903
3904inline UBool
3905UnicodeString::isBogus() const
3906{ return (UBool)(fUnion.fFields.fLengthAndFlags & kIsBogus); }
3907
3908inline UBool
3909UnicodeString::isWritable() const
3910{ return (UBool)!(fUnion.fFields.fLengthAndFlags&(kOpenGetBuffer|kIsBogus)); }
3911
3912inline UBool
3913UnicodeString::isBufferWritable() const
3914{
3915 return (UBool)(
3916 !(fUnion.fFields.fLengthAndFlags&(kOpenGetBuffer|kIsBogus|kBufferIsReadonly)) &&
3917 (!(fUnion.fFields.fLengthAndFlags&kRefCounted) || refCount()==1));
3918}
3919
3920inline const char16_t *
3921UnicodeString::getBuffer() const {
3922 if(fUnion.fFields.fLengthAndFlags&(kIsBogus|kOpenGetBuffer)) {
3923 return nullptr;
3924 } else if(fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) {
3925 return fUnion.fStackFields.fBuffer;
3926 } else {
3927 return fUnion.fFields.fArray;
3928 }
3929}
3930
3931//========================================
3932// Read-only alias methods
3933//========================================
3934inline int8_t
3935UnicodeString::doCompare(int32_t start,
3936 int32_t thisLength,
3937 const UnicodeString& srcText,
3938 int32_t srcStart,
3939 int32_t srcLength) const
3940{
3941 if(srcText.isBogus()) {
3942 return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
3943 } else {
3944 srcText.pinIndices(srcStart, srcLength);
3945 return doCompare(start, thisLength, srcText.getArrayStart(), srcStart, srcLength);
3946 }
3947}
3948
3949inline bool
3950UnicodeString::operator== (const UnicodeString& text) const
3951{
3952 if(isBogus()) {
3953 return text.isBogus();
3954 } else {
3955 int32_t len = length(), textLength = text.length();
3956 return !text.isBogus() && len == textLength && doEquals(text, len);
3957 }
3958}
3959
3960inline bool
3961UnicodeString::operator!= (const UnicodeString& text) const
3962{ return (! operator==(text)); }
3963
3964inline UBool
3965UnicodeString::operator> (const UnicodeString& text) const
3966{ return doCompare(0, length(), text, 0, text.length()) == 1; }
3967
3968inline UBool
3969UnicodeString::operator< (const UnicodeString& text) const
3970{ return doCompare(0, length(), text, 0, text.length()) == -1; }
3971
3972inline UBool
3973UnicodeString::operator>= (const UnicodeString& text) const
3974{ return doCompare(0, length(), text, 0, text.length()) != -1; }
3975
3976inline UBool
3977UnicodeString::operator<= (const UnicodeString& text) const
3978{ return doCompare(0, length(), text, 0, text.length()) != 1; }
3979
3980inline int8_t
3981UnicodeString::compare(const UnicodeString& text) const
3982{ return doCompare(0, length(), text, 0, text.length()); }
3983
3984inline int8_t
3985UnicodeString::compare(int32_t start,
3986 int32_t _length,
3987 const UnicodeString& srcText) const
3988{ return doCompare(start, _length, srcText, 0, srcText.length()); }
3989
3990inline int8_t
3991UnicodeString::compare(ConstChar16Ptr srcChars,
3992 int32_t srcLength) const
3993{ return doCompare(0, length(), srcChars, 0, srcLength); }
3994
3995inline int8_t
3996UnicodeString::compare(int32_t start,
3997 int32_t _length,
3998 const UnicodeString& srcText,
3999 int32_t srcStart,
4000 int32_t srcLength) const
4001{ return doCompare(start, _length, srcText, srcStart, srcLength); }
4002
4003inline int8_t
4004UnicodeString::compare(int32_t start,
4005 int32_t _length,
4006 const char16_t *srcChars) const
4007{ return doCompare(start, _length, srcChars, 0, _length); }
4008
4009inline int8_t
4010UnicodeString::compare(int32_t start,
4011 int32_t _length,
4012 const char16_t *srcChars,
4013 int32_t srcStart,
4014 int32_t srcLength) const
4015{ return doCompare(start, _length, srcChars, srcStart, srcLength); }
4016
4017inline int8_t
4018UnicodeString::compareBetween(int32_t start,
4019 int32_t limit,
4020 const UnicodeString& srcText,
4021 int32_t srcStart,
4022 int32_t srcLimit) const
4023{ return doCompare(start, limit - start,
4024 srcText, srcStart, srcLimit - srcStart); }
4025
4026inline int8_t
4027UnicodeString::doCompareCodePointOrder(int32_t start,
4028 int32_t thisLength,
4029 const UnicodeString& srcText,
4030 int32_t srcStart,
4031 int32_t srcLength) const
4032{
4033 if(srcText.isBogus()) {
4034 return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
4035 } else {
4036 srcText.pinIndices(srcStart, srcLength);
4037 return doCompareCodePointOrder(start, thisLength, srcText.getArrayStart(), srcStart, srcLength);
4038 }
4039}
4040
4041inline int8_t
4042UnicodeString::compareCodePointOrder(const UnicodeString& text) const
4043{ return doCompareCodePointOrder(0, length(), text, 0, text.length()); }
4044
4045inline int8_t
4046UnicodeString::compareCodePointOrder(int32_t start,
4047 int32_t _length,
4048 const UnicodeString& srcText) const
4049{ return doCompareCodePointOrder(start, _length, srcText, 0, srcText.length()); }
4050
4051inline int8_t
4052UnicodeString::compareCodePointOrder(ConstChar16Ptr srcChars,
4053 int32_t srcLength) const
4054{ return doCompareCodePointOrder(0, length(), srcChars, 0, srcLength); }
4055
4056inline int8_t
4057UnicodeString::compareCodePointOrder(int32_t start,
4058 int32_t _length,
4059 const UnicodeString& srcText,
4060 int32_t srcStart,
4061 int32_t srcLength) const
4062{ return doCompareCodePointOrder(start, _length, srcText, srcStart, srcLength); }
4063
4064inline int8_t
4065UnicodeString::compareCodePointOrder(int32_t start,
4066 int32_t _length,
4067 const char16_t *srcChars) const
4068{ return doCompareCodePointOrder(start, _length, srcChars, 0, _length); }
4069
4070inline int8_t
4071UnicodeString::compareCodePointOrder(int32_t start,
4072 int32_t _length,
4073 const char16_t *srcChars,
4074 int32_t srcStart,
4075 int32_t srcLength) const
4076{ return doCompareCodePointOrder(start, _length, srcChars, srcStart, srcLength); }
4077
4078inline int8_t
4079UnicodeString::compareCodePointOrderBetween(int32_t start,
4080 int32_t limit,
4081 const UnicodeString& srcText,
4082 int32_t srcStart,
4083 int32_t srcLimit) const
4084{ return doCompareCodePointOrder(start, limit - start,
4085 srcText, srcStart, srcLimit - srcStart); }
4086
4087inline int8_t
4088UnicodeString::doCaseCompare(int32_t start,
4089 int32_t thisLength,
4090 const UnicodeString &srcText,
4091 int32_t srcStart,
4092 int32_t srcLength,
4093 uint32_t options) const
4094{
4095 if(srcText.isBogus()) {
4096 return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
4097 } else {
4098 srcText.pinIndices(srcStart, srcLength);
4099 return doCaseCompare(start, thisLength, srcText.getArrayStart(), srcStart, srcLength, options);
4100 }
4101}
4102
4103inline int8_t
4104UnicodeString::caseCompare(const UnicodeString &text, uint32_t options) const {
4105 return doCaseCompare(0, length(), text, 0, text.length(), options);
4106}
4107
4108inline int8_t
4109UnicodeString::caseCompare(int32_t start,
4110 int32_t _length,
4111 const UnicodeString &srcText,
4112 uint32_t options) const {
4113 return doCaseCompare(start, _length, srcText, 0, srcText.length(), options);
4114}
4115
4116inline int8_t
4117UnicodeString::caseCompare(ConstChar16Ptr srcChars,
4118 int32_t srcLength,
4119 uint32_t options) const {
4120 return doCaseCompare(0, length(), srcChars, 0, srcLength, options);
4121}
4122
4123inline int8_t
4124UnicodeString::caseCompare(int32_t start,
4125 int32_t _length,
4126 const UnicodeString &srcText,
4127 int32_t srcStart,
4128 int32_t srcLength,
4129 uint32_t options) const {
4130 return doCaseCompare(start, _length, srcText, srcStart, srcLength, options);
4131}
4132
4133inline int8_t
4134UnicodeString::caseCompare(int32_t start,
4135 int32_t _length,
4136 const char16_t *srcChars,
4137 uint32_t options) const {
4138 return doCaseCompare(start, _length, srcChars, 0, _length, options);
4139}
4140
4141inline int8_t
4142UnicodeString::caseCompare(int32_t start,
4143 int32_t _length,
4144 const char16_t *srcChars,
4145 int32_t srcStart,
4146 int32_t srcLength,
4147 uint32_t options) const {
4148 return doCaseCompare(start, _length, srcChars, srcStart, srcLength, options);
4149}
4150
4151inline int8_t
4152UnicodeString::caseCompareBetween(int32_t start,
4153 int32_t limit,
4154 const UnicodeString &srcText,
4155 int32_t srcStart,
4156 int32_t srcLimit,
4157 uint32_t options) const {
4158 return doCaseCompare(start, limit - start, srcText, srcStart, srcLimit - srcStart, options);
4159}
4160
4161inline int32_t
4162UnicodeString::indexOf(const UnicodeString& srcText,
4163 int32_t srcStart,
4164 int32_t srcLength,
4165 int32_t start,
4166 int32_t _length) const
4167{
4168 if(!srcText.isBogus()) {
4169 srcText.pinIndices(srcStart, srcLength);
4170 if(srcLength > 0) {
4171 return indexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
4172 }
4173 }
4174 return -1;
4175}
4176
4177inline int32_t
4178UnicodeString::indexOf(const UnicodeString& text) const
4179{ return indexOf(text, 0, text.length(), 0, length()); }
4180
4181inline int32_t
4182UnicodeString::indexOf(const UnicodeString& text,
4183 int32_t start) const {
4184 pinIndex(start);
4185 return indexOf(text, 0, text.length(), start, length() - start);
4186}
4187
4188inline int32_t
4189UnicodeString::indexOf(const UnicodeString& text,
4190 int32_t start,
4191 int32_t _length) const
4192{ return indexOf(text, 0, text.length(), start, _length); }
4193
4194inline int32_t
4195UnicodeString::indexOf(const char16_t *srcChars,
4196 int32_t srcLength,
4197 int32_t start) const {
4198 pinIndex(start);
4199 return indexOf(srcChars, 0, srcLength, start, length() - start);
4200}
4201
4202inline int32_t
4203UnicodeString::indexOf(ConstChar16Ptr srcChars,
4204 int32_t srcLength,
4205 int32_t start,
4206 int32_t _length) const
4207{ return indexOf(srcChars, 0, srcLength, start, _length); }
4208
4209inline int32_t
4210UnicodeString::indexOf(char16_t c,
4211 int32_t start,
4212 int32_t _length) const
4213{ return doIndexOf(c, start, _length); }
4214
4215inline int32_t
4216UnicodeString::indexOf(UChar32 c,
4217 int32_t start,
4218 int32_t _length) const
4219{ return doIndexOf(c, start, _length); }
4220
4221inline int32_t
4222UnicodeString::indexOf(char16_t c) const
4223{ return doIndexOf(c, 0, length()); }
4224
4225inline int32_t
4226UnicodeString::indexOf(UChar32 c) const
4227{ return indexOf(c, 0, length()); }
4228
4229inline int32_t
4230UnicodeString::indexOf(char16_t c,
4231 int32_t start) const {
4232 pinIndex(start);
4233 return doIndexOf(c, start, length() - start);
4234}
4235
4236inline int32_t
4237UnicodeString::indexOf(UChar32 c,
4238 int32_t start) const {
4239 pinIndex(start);
4240 return indexOf(c, start, length() - start);
4241}
4242
4243inline int32_t
4244UnicodeString::lastIndexOf(ConstChar16Ptr srcChars,
4245 int32_t srcLength,
4246 int32_t start,
4247 int32_t _length) const
4248{ return lastIndexOf(srcChars, 0, srcLength, start, _length); }
4249
4250inline int32_t
4251UnicodeString::lastIndexOf(const char16_t *srcChars,
4252 int32_t srcLength,
4253 int32_t start) const {
4254 pinIndex(start);
4255 return lastIndexOf(srcChars, 0, srcLength, start, length() - start);
4256}
4257
4258inline int32_t
4259UnicodeString::lastIndexOf(const UnicodeString& srcText,
4260 int32_t srcStart,
4261 int32_t srcLength,
4262 int32_t start,
4263 int32_t _length) const
4264{
4265 if(!srcText.isBogus()) {
4266 srcText.pinIndices(srcStart, srcLength);
4267 if(srcLength > 0) {
4268 return lastIndexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
4269 }
4270 }
4271 return -1;
4272}
4273
4274inline int32_t
4275UnicodeString::lastIndexOf(const UnicodeString& text,
4276 int32_t start,
4277 int32_t _length) const
4278{ return lastIndexOf(text, 0, text.length(), start, _length); }
4279
4280inline int32_t
4281UnicodeString::lastIndexOf(const UnicodeString& text,
4282 int32_t start) const {
4283 pinIndex(start);
4284 return lastIndexOf(text, 0, text.length(), start, length() - start);
4285}
4286
4287inline int32_t
4288UnicodeString::lastIndexOf(const UnicodeString& text) const
4289{ return lastIndexOf(text, 0, text.length(), 0, length()); }
4290
4291inline int32_t
4292UnicodeString::lastIndexOf(char16_t c,
4293 int32_t start,
4294 int32_t _length) const
4295{ return doLastIndexOf(c, start, _length); }
4296
4297inline int32_t
4298UnicodeString::lastIndexOf(UChar32 c,
4299 int32_t start,
4300 int32_t _length) const {
4301 return doLastIndexOf(c, start, _length);
4302}
4303
4304inline int32_t
4305UnicodeString::lastIndexOf(char16_t c) const
4306{ return doLastIndexOf(c, 0, length()); }
4307
4308inline int32_t
4309UnicodeString::lastIndexOf(UChar32 c) const {
4310 return lastIndexOf(c, 0, length());
4311}
4312
4313inline int32_t
4314UnicodeString::lastIndexOf(char16_t c,
4315 int32_t start) const {
4316 pinIndex(start);
4317 return doLastIndexOf(c, start, length() - start);
4318}
4319
4320inline int32_t
4321UnicodeString::lastIndexOf(UChar32 c,
4322 int32_t start) const {
4323 pinIndex(start);
4324 return lastIndexOf(c, start, length() - start);
4325}
4326
4327inline UBool
4328UnicodeString::startsWith(const UnicodeString& text) const
4329{ return compare(0, text.length(), text, 0, text.length()) == 0; }
4330
4331inline UBool
4332UnicodeString::startsWith(const UnicodeString& srcText,
4333 int32_t srcStart,
4334 int32_t srcLength) const
4335{ return doCompare(0, srcLength, srcText, srcStart, srcLength) == 0; }
4336
4337inline UBool
4338UnicodeString::startsWith(ConstChar16Ptr srcChars, int32_t srcLength) const {
4339 if(srcLength < 0) {
4340 srcLength = u_strlenu_strlen_71(toUCharPtr(srcChars));
4341 }
4342 return doCompare(0, srcLength, srcChars, 0, srcLength) == 0;
4343}
4344
4345inline UBool
4346UnicodeString::startsWith(const char16_t *srcChars, int32_t srcStart, int32_t srcLength) const {
4347 if(srcLength < 0) {
4348 srcLength = u_strlenu_strlen_71(toUCharPtr(srcChars));
4349 }
4350 return doCompare(0, srcLength, srcChars, srcStart, srcLength) == 0;
4351}
4352
4353inline UBool
4354UnicodeString::endsWith(const UnicodeString& text) const
4355{ return doCompare(length() - text.length(), text.length(),
4356 text, 0, text.length()) == 0; }
4357
4358inline UBool
4359UnicodeString::endsWith(const UnicodeString& srcText,
4360 int32_t srcStart,
4361 int32_t srcLength) const {
4362 srcText.pinIndices(srcStart, srcLength);
4363 return doCompare(length() - srcLength, srcLength,
4364 srcText, srcStart, srcLength) == 0;
4365}
4366
4367inline UBool
4368UnicodeString::endsWith(ConstChar16Ptr srcChars,
4369 int32_t srcLength) const {
4370 if(srcLength < 0) {
4371 srcLength = u_strlenu_strlen_71(toUCharPtr(srcChars));
4372 }
4373 return doCompare(length() - srcLength, srcLength,
4374 srcChars, 0, srcLength) == 0;
4375}
4376
4377inline UBool
4378UnicodeString::endsWith(const char16_t *srcChars,
4379 int32_t srcStart,
4380 int32_t srcLength) const {
4381 if(srcLength < 0) {
4382 srcLength = u_strlenu_strlen_71(toUCharPtr(srcChars + srcStart));
4383 }
4384 return doCompare(length() - srcLength, srcLength,
4385 srcChars, srcStart, srcLength) == 0;
4386}
4387
4388//========================================
4389// replace
4390//========================================
4391inline UnicodeString&
4392UnicodeString::replace(int32_t start,
4393 int32_t _length,
4394 const UnicodeString& srcText)
4395{ return doReplace(start, _length, srcText, 0, srcText.length()); }
4396
4397inline UnicodeString&
4398UnicodeString::replace(int32_t start,
4399 int32_t _length,
4400 const UnicodeString& srcText,
4401 int32_t srcStart,
4402 int32_t srcLength)
4403{ return doReplace(start, _length, srcText, srcStart, srcLength); }
4404
4405inline UnicodeString&
4406UnicodeString::replace(int32_t start,
4407 int32_t _length,
4408 ConstChar16Ptr srcChars,
4409 int32_t srcLength)
4410{ return doReplace(start, _length, srcChars, 0, srcLength); }
4411
4412inline UnicodeString&
4413UnicodeString::replace(int32_t start,
4414 int32_t _length,
4415 const char16_t *srcChars,
4416 int32_t srcStart,
4417 int32_t srcLength)
4418{ return doReplace(start, _length, srcChars, srcStart, srcLength); }
4419
4420inline UnicodeString&
4421UnicodeString::replace(int32_t start,
4422 int32_t _length,
4423 char16_t srcChar)
4424{ return doReplace(start, _length, &srcChar, 0, 1); }
4425
4426inline UnicodeString&
4427UnicodeString::replaceBetween(int32_t start,
4428 int32_t limit,
4429 const UnicodeString& srcText)
4430{ return doReplace(start, limit - start, srcText, 0, srcText.length()); }
4431
4432inline UnicodeString&
4433UnicodeString::replaceBetween(int32_t start,
4434 int32_t limit,
4435 const UnicodeString& srcText,
4436 int32_t srcStart,
4437 int32_t srcLimit)
4438{ return doReplace(start, limit - start, srcText, srcStart, srcLimit - srcStart); }
4439
4440inline UnicodeString&
4441UnicodeString::findAndReplace(const UnicodeString& oldText,
4442 const UnicodeString& newText)
4443{ return findAndReplace(0, length(), oldText, 0, oldText.length(),
4444 newText, 0, newText.length()); }
4445
4446inline UnicodeString&
4447UnicodeString::findAndReplace(int32_t start,
4448 int32_t _length,
4449 const UnicodeString& oldText,
4450 const UnicodeString& newText)
4451{ return findAndReplace(start, _length, oldText, 0, oldText.length(),
4452 newText, 0, newText.length()); }
4453
4454// ============================
4455// extract
4456// ============================
4457inline void
4458UnicodeString::doExtract(int32_t start,
4459 int32_t _length,
4460 UnicodeString& target) const
4461{ target.replace(0, target.length(), *this, start, _length); }
4462
4463inline void
4464UnicodeString::extract(int32_t start,
4465 int32_t _length,
4466 Char16Ptr target,
4467 int32_t targetStart) const
4468{ doExtract(start, _length, target, targetStart); }
4469
4470inline void
4471UnicodeString::extract(int32_t start,
4472 int32_t _length,
4473 UnicodeString& target) const
4474{ doExtract(start, _length, target); }
4475
4476#if !UCONFIG_NO_CONVERSION0
4477
4478inline int32_t
4479UnicodeString::extract(int32_t start,
4480 int32_t _length,
4481 char *dst,
4482 const char *codepage) const
4483
4484{
4485 // This dstSize value will be checked explicitly
4486 return extract(start, _length, dst, dst!=0 ? 0xffffffff : 0, codepage);
4487}
4488
4489#endif
4490
4491inline void
4492UnicodeString::extractBetween(int32_t start,
4493 int32_t limit,
4494 char16_t *dst,
4495 int32_t dstStart) const {
4496 pinIndex(start);
4497 pinIndex(limit);
4498 doExtract(start, limit - start, dst, dstStart);
4499}
4500
4501inline UnicodeString
4502UnicodeString::tempSubStringBetween(int32_t start, int32_t limit) const {
4503 return tempSubString(start, limit - start);
4504}
4505
4506inline char16_t
4507UnicodeString::doCharAt(int32_t offset) const
4508{
4509 if((uint32_t)offset < (uint32_t)length()) {
4510 return getArrayStart()[offset];
4511 } else {
4512 return kInvalidUChar;
4513 }
4514}
4515
4516inline char16_t
4517UnicodeString::charAt(int32_t offset) const
4518{ return doCharAt(offset); }
4519
4520inline char16_t
4521UnicodeString::operator[] (int32_t offset) const
4522{ return doCharAt(offset); }
4523
4524inline UBool
4525UnicodeString::isEmpty() const {
4526 // Arithmetic or logical right shift does not matter: only testing for 0.
4527 return (fUnion.fFields.fLengthAndFlags>>kLengthShift) == 0;
4528}
4529
4530//========================================
4531// Write implementation methods
4532//========================================
4533inline void
4534UnicodeString::setZeroLength() {
4535 fUnion.fFields.fLengthAndFlags &= kAllStorageFlags;
4536}
4537
4538inline void
4539UnicodeString::setShortLength(int32_t len) {
4540 // requires 0 <= len <= kMaxShortLength
4541 fUnion.fFields.fLengthAndFlags =
4542 (int16_t)((fUnion.fFields.fLengthAndFlags & kAllStorageFlags) | (len << kLengthShift));
19
The result of the left shift is undefined because the left operand is negative
4543}
4544
4545inline void
4546UnicodeString::setLength(int32_t len) {
4547 if(len
15.1
'len' is <= kMaxShortLength
15.1
'len' is <= kMaxShortLength
<= kMaxShortLength) {
16
Taking true branch
4548 setShortLength(len);
17
Passing the value -1 via 1st parameter 'len'
18
Calling 'UnicodeString::setShortLength'
4549 } else {
4550 fUnion.fFields.fLengthAndFlags |= kLengthIsLarge;
4551 fUnion.fFields.fLength = len;
4552 }
4553}
4554
4555inline void
4556UnicodeString::setToEmpty() {
4557 fUnion.fFields.fLengthAndFlags = kShortString;
4558}
4559
4560inline void
4561UnicodeString::setArray(char16_t *array, int32_t len, int32_t capacity) {
4562 setLength(len);
4563 fUnion.fFields.fArray = array;
4564 fUnion.fFields.fCapacity = capacity;
4565}
4566
4567inline UnicodeString&
4568UnicodeString::operator= (char16_t ch)
4569{ return doReplace(0, length(), &ch, 0, 1); }
4570
4571inline UnicodeString&
4572UnicodeString::operator= (UChar32 ch)
4573{ return replace(0, length(), ch); }
4574
4575inline UnicodeString&
4576UnicodeString::setTo(const UnicodeString& srcText,
4577 int32_t srcStart,
4578 int32_t srcLength)
4579{
4580 unBogus();
4581 return doReplace(0, length(), srcText, srcStart, srcLength);
4582}
4583
4584inline UnicodeString&
4585UnicodeString::setTo(const UnicodeString& srcText,
4586 int32_t srcStart)
4587{
4588 unBogus();
4589 srcText.pinIndex(srcStart);
4590 return doReplace(0, length(), srcText, srcStart, srcText.length() - srcStart);
4591}
4592
4593inline UnicodeString&
4594UnicodeString::setTo(const UnicodeString& srcText)
4595{
4596 return copyFrom(srcText);
4597}
4598
4599inline UnicodeString&
4600UnicodeString::setTo(const char16_t *srcChars,
4601 int32_t srcLength)
4602{
4603 unBogus();
4604 return doReplace(0, length(), srcChars, 0, srcLength);
4605}
4606
4607inline UnicodeString&
4608UnicodeString::setTo(char16_t srcChar)
4609{
4610 unBogus();
4611 return doReplace(0, length(), &srcChar, 0, 1);
4612}
4613
4614inline UnicodeString&
4615UnicodeString::setTo(UChar32 srcChar)
4616{
4617 unBogus();
4618 return replace(0, length(), srcChar);
4619}
4620
4621inline UnicodeString&
4622UnicodeString::append(const UnicodeString& srcText,
4623 int32_t srcStart,
4624 int32_t srcLength)
4625{ return doAppend(srcText, srcStart, srcLength); }
4626
4627inline UnicodeString&
4628UnicodeString::append(const UnicodeString& srcText)
4629{ return doAppend(srcText, 0, srcText.length()); }
4630
4631inline UnicodeString&
4632UnicodeString::append(const char16_t *srcChars,
4633 int32_t srcStart,
4634 int32_t srcLength)
4635{ return doAppend(srcChars, srcStart, srcLength); }
4636
4637inline UnicodeString&
4638UnicodeString::append(ConstChar16Ptr srcChars,
4639 int32_t srcLength)
4640{ return doAppend(srcChars, 0, srcLength); }
4641
4642inline UnicodeString&
4643UnicodeString::append(char16_t srcChar)
4644{ return doAppend(&srcChar, 0, 1); }
4645
4646inline UnicodeString&
4647UnicodeString::operator+= (char16_t ch)
4648{ return doAppend(&ch, 0, 1); }
4649
4650inline UnicodeString&
4651UnicodeString::operator+= (UChar32 ch) {
4652 return append(ch);
4653}
4654
4655inline UnicodeString&
4656UnicodeString::operator+= (const UnicodeString& srcText)
4657{ return doAppend(srcText, 0, srcText.length()); }
4658
4659inline UnicodeString&
4660UnicodeString::insert(int32_t start,
4661 const UnicodeString& srcText,
4662 int32_t srcStart,
4663 int32_t srcLength)
4664{ return doReplace(start, 0, srcText, srcStart, srcLength); }
4665
4666inline UnicodeString&
4667UnicodeString::insert(int32_t start,
4668 const UnicodeString& srcText)
4669{ return doReplace(start, 0, srcText, 0, srcText.length()); }
4670
4671inline UnicodeString&
4672UnicodeString::insert(int32_t start,
4673 const char16_t *srcChars,
4674 int32_t srcStart,
4675 int32_t srcLength)
4676{ return doReplace(start, 0, srcChars, srcStart, srcLength); }
4677
4678inline UnicodeString&
4679UnicodeString::insert(int32_t start,
4680 ConstChar16Ptr srcChars,
4681 int32_t srcLength)
4682{ return doReplace(start, 0, srcChars, 0, srcLength); }
4683
4684inline UnicodeString&
4685UnicodeString::insert(int32_t start,
4686 char16_t srcChar)
4687{ return doReplace(start, 0, &srcChar, 0, 1); }
4688
4689inline UnicodeString&
4690UnicodeString::insert(int32_t start,
4691 UChar32 srcChar)
4692{ return replace(start, 0, srcChar); }
4693
4694
4695inline UnicodeString&
4696UnicodeString::remove()
4697{
4698 // remove() of a bogus string makes the string empty and non-bogus
4699 if(isBogus()) {
4700 setToEmpty();
4701 } else {
4702 setZeroLength();
4703 }
4704 return *this;
4705}
4706
4707inline UnicodeString&
4708UnicodeString::remove(int32_t start,
4709 int32_t _length)
4710{
4711 if(start <= 0 && _length == INT32_MAX(2147483647)) {
4712 // remove(guaranteed everything) of a bogus string makes the string empty and non-bogus
4713 return remove();
4714 }
4715 return doReplace(start, _length, NULL__null, 0, 0);
4716}
4717
4718inline UnicodeString&
4719UnicodeString::removeBetween(int32_t start,
4720 int32_t limit)
4721{ return doReplace(start, limit - start, NULL__null, 0, 0); }
4722
4723inline UnicodeString &
4724UnicodeString::retainBetween(int32_t start, int32_t limit) {
4725 truncate(limit);
4726 return doReplace(0, start, NULL__null, 0, 0);
4727}
4728
4729inline UBool
4730UnicodeString::truncate(int32_t targetLength)
4731{
4732 if(isBogus() && targetLength == 0) {
4733 // truncate(0) of a bogus string makes the string empty and non-bogus
4734 unBogus();
4735 return false;
4736 } else if((uint32_t)targetLength < (uint32_t)length()) {
4737 setLength(targetLength);
4738 return true;
4739 } else {
4740 return false;
4741 }
4742}
4743
4744inline UnicodeString&
4745UnicodeString::reverse()
4746{ return doReverse(0, length()); }
4747
4748inline UnicodeString&
4749UnicodeString::reverse(int32_t start,
4750 int32_t _length)
4751{ return doReverse(start, _length); }
4752
4753U_NAMESPACE_END}
4754
4755#endif /* U_SHOW_CPLUSPLUS_API */
4756
4757#endif