fix crashes due to missing npos handling in several wxString methods in STL build...
[wxWidgets.git] / include / wx / string.h
1 ///////////////////////////////////////////////////////////////////////////////
2 // Name: wx/string.h
3 // Purpose: wxString class
4 // Author: Vadim Zeitlin
5 // Modified by:
6 // Created: 29/01/98
7 // RCS-ID: $Id$
8 // Copyright: (c) 1998 Vadim Zeitlin <zeitlin@dptmaths.ens-cachan.fr>
9 // Licence: wxWindows licence
10 ///////////////////////////////////////////////////////////////////////////////
11
12 /*
13 Efficient string class [more or less] compatible with MFC CString,
14 wxWidgets version 1 wxString and std::string and some handy functions
15 missing from string.h.
16 */
17
18 #ifndef _WX_WXSTRING_H__
19 #define _WX_WXSTRING_H__
20
21 // ----------------------------------------------------------------------------
22 // headers
23 // ----------------------------------------------------------------------------
24
25 #include "wx/defs.h" // everybody should include this
26
27 #ifndef __WXPALMOS5__
28 #if defined(__WXMAC__) || defined(__VISAGECPP__)
29 #include <ctype.h>
30 #endif
31
32 #if defined(__VISAGECPP__) && __IBMCPP__ >= 400
33 // problem in VACPP V4 with including stdlib.h multiple times
34 // strconv includes it anyway
35 # include <stdio.h>
36 # include <string.h>
37 # include <stdarg.h>
38 # include <limits.h>
39 #else
40 # include <string.h>
41 # include <stdio.h>
42 # include <stdarg.h>
43 # include <limits.h>
44 # include <stdlib.h>
45 #endif
46
47 #ifdef HAVE_STRCASECMP_IN_STRINGS_H
48 #include <strings.h> // for strcasecmp()
49 #endif // HAVE_STRCASECMP_IN_STRINGS_H
50 #endif // ! __WXPALMOS5__
51
52 #include "wx/wxcrtbase.h" // for wxChar, wxStrlen() etc.
53 #include "wx/strvararg.h"
54 #include "wx/buffer.h" // for wxCharBuffer
55 #include "wx/strconv.h" // for wxConvertXXX() macros and wxMBConv classes
56 #include "wx/stringimpl.h"
57 #include "wx/stringops.h"
58 #include "wx/unichar.h"
59
60 class WXDLLIMPEXP_FWD_BASE wxString;
61
62 // unless this symbol is predefined to disable the compatibility functions, do
63 // use them
64 #ifndef WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
65 #define WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER 1
66 #endif
67
68 // ---------------------------------------------------------------------------
69 // macros
70 // ---------------------------------------------------------------------------
71
72 // casts [unfortunately!] needed to call some broken functions which require
73 // "char *" instead of "const char *"
74 #define WXSTRINGCAST (wxChar *)(const wxChar *)
75 #define wxCSTRINGCAST (wxChar *)(const wxChar *)
76 #define wxMBSTRINGCAST (char *)(const char *)
77 #define wxWCSTRINGCAST (wchar_t *)(const wchar_t *)
78
79 // ----------------------------------------------------------------------------
80 // constants
81 // ----------------------------------------------------------------------------
82
83 #if WXWIN_COMPATIBILITY_2_6
84
85 // deprecated in favour of wxString::npos, don't use in new code
86 //
87 // maximum possible length for a string means "take all string" everywhere
88 #define wxSTRING_MAXLEN wxString::npos
89
90 #endif // WXWIN_COMPATIBILITY_2_6
91
92 // ---------------------------------------------------------------------------
93 // global functions complementing standard C string library replacements for
94 // strlen() and portable strcasecmp()
95 //---------------------------------------------------------------------------
96
97 #if WXWIN_COMPATIBILITY_2_8
98 // Use wxXXX() functions from wxcrt.h instead! These functions are for
99 // backwards compatibility only.
100
101 // checks whether the passed in pointer is NULL and if the string is empty
102 wxDEPRECATED( inline bool IsEmpty(const char *p) );
103 inline bool IsEmpty(const char *p) { return (!p || !*p); }
104
105 // safe version of strlen() (returns 0 if passed NULL pointer)
106 wxDEPRECATED( inline size_t Strlen(const char *psz) );
107 inline size_t Strlen(const char *psz)
108 { return psz ? strlen(psz) : 0; }
109
110 // portable strcasecmp/_stricmp
111 wxDEPRECATED( inline int Stricmp(const char *psz1, const char *psz2) );
112 inline int Stricmp(const char *psz1, const char *psz2)
113 {
114 #if defined(__VISUALC__) && defined(__WXWINCE__)
115 register char c1, c2;
116 do {
117 c1 = tolower(*psz1++);
118 c2 = tolower(*psz2++);
119 } while ( c1 && (c1 == c2) );
120
121 return c1 - c2;
122 #elif defined(__VISUALC__) || ( defined(__MWERKS__) && defined(__INTEL__) )
123 return _stricmp(psz1, psz2);
124 #elif defined(__SC__)
125 return _stricmp(psz1, psz2);
126 #elif defined(__BORLANDC__)
127 return stricmp(psz1, psz2);
128 #elif defined(__WATCOMC__)
129 return stricmp(psz1, psz2);
130 #elif defined(__DJGPP__)
131 return stricmp(psz1, psz2);
132 #elif defined(__EMX__)
133 return stricmp(psz1, psz2);
134 #elif defined(__WXPM__)
135 return stricmp(psz1, psz2);
136 #elif defined(__WXPALMOS__) || \
137 defined(HAVE_STRCASECMP_IN_STRING_H) || \
138 defined(HAVE_STRCASECMP_IN_STRINGS_H) || \
139 defined(__GNUWIN32__)
140 return strcasecmp(psz1, psz2);
141 #elif defined(__MWERKS__) && !defined(__INTEL__)
142 register char c1, c2;
143 do {
144 c1 = tolower(*psz1++);
145 c2 = tolower(*psz2++);
146 } while ( c1 && (c1 == c2) );
147
148 return c1 - c2;
149 #else
150 // almost all compilers/libraries provide this function (unfortunately under
151 // different names), that's why we don't implement our own which will surely
152 // be more efficient than this code (uncomment to use):
153 /*
154 register char c1, c2;
155 do {
156 c1 = tolower(*psz1++);
157 c2 = tolower(*psz2++);
158 } while ( c1 && (c1 == c2) );
159
160 return c1 - c2;
161 */
162
163 #error "Please define string case-insensitive compare for your OS/compiler"
164 #endif // OS/compiler
165 }
166
167 #endif // WXWIN_COMPATIBILITY_2_8
168
169 // ----------------------------------------------------------------------------
170 // wxCStrData
171 // ----------------------------------------------------------------------------
172
173 // Lightweight object returned by wxString::c_str() and implicitly convertible
174 // to either const char* or const wchar_t*.
175 class WXDLLIMPEXP_BASE wxCStrData
176 {
177 private:
178 // Ctors; for internal use by wxString and wxCStrData only
179 wxCStrData(const wxString *str, size_t offset = 0, bool owned = false)
180 : m_str(str), m_offset(offset), m_owned(owned) {}
181
182 public:
183 // Ctor constructs the object from char literal; they are needed to make
184 // operator?: compile and they intentionally take char*, not const char*
185 inline wxCStrData(char *buf);
186 inline wxCStrData(wchar_t *buf);
187 inline wxCStrData(const wxCStrData& data);
188
189 inline ~wxCStrData();
190
191 // methods defined inline below must be declared inline or mingw32 3.4.5
192 // warns about "<symbol> defined locally after being referenced with
193 // dllimport linkage"
194 #if wxUSE_UNICODE_WCHAR
195 inline
196 #endif
197 const wchar_t* AsWChar() const;
198 operator const wchar_t*() const { return AsWChar(); }
199
200 #if !wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY
201 inline
202 #endif
203 const char* AsChar() const;
204 const unsigned char* AsUnsignedChar() const
205 { return (const unsigned char *) AsChar(); }
206 operator const char*() const { return AsChar(); }
207 operator const unsigned char*() const { return AsUnsignedChar(); }
208
209 operator const void*() const { return AsChar(); }
210
211 inline const wxCharBuffer AsCharBuf() const;
212 inline const wxWCharBuffer AsWCharBuf() const;
213
214 inline wxString AsString() const;
215
216 // returns the value as C string in internal representation (equivalent
217 // to AsString().wx_str(), but more efficient)
218 const wxStringCharType *AsInternal() const;
219
220 // allow expressions like "c_str()[0]":
221 inline wxUniChar operator[](size_t n) const;
222 wxUniChar operator[](int n) const { return operator[](size_t(n)); }
223 wxUniChar operator[](long n) const { return operator[](size_t(n)); }
224 #ifndef wxSIZE_T_IS_UINT
225 wxUniChar operator[](unsigned int n) const { return operator[](size_t(n)); }
226 #endif // size_t != unsigned int
227
228 // these operators are needed to emulate the pointer semantics of c_str():
229 // expressions like "wxChar *p = str.c_str() + 1;" should continue to work
230 // (we need both versions to resolve ambiguities):
231 wxCStrData operator+(int n) const
232 { return wxCStrData(m_str, m_offset + n, m_owned); }
233 wxCStrData operator+(long n) const
234 { return wxCStrData(m_str, m_offset + n, m_owned); }
235 wxCStrData operator+(size_t n) const
236 { return wxCStrData(m_str, m_offset + n, m_owned); }
237
238 // and these for "str.c_str() + (p2 - p1)" (it also works for any integer
239 // expression but it must be ptrdiff_t and not e.g. int to work in this
240 // example):
241 wxCStrData operator-(ptrdiff_t n) const
242 {
243 wxASSERT_MSG( n <= (ptrdiff_t)m_offset,
244 _T("attempt to construct address before the beginning of the string") );
245 return wxCStrData(m_str, m_offset - n, m_owned);
246 }
247
248 // this operator is needed to make expressions like "*c_str()" or
249 // "*(c_str() + 2)" work
250 inline wxUniChar operator*() const;
251
252 private:
253 const wxString *m_str;
254 size_t m_offset;
255 bool m_owned;
256
257 friend class WXDLLIMPEXP_FWD_BASE wxString;
258 };
259
260 // ----------------------------------------------------------------------------
261 // wxStringPrintfMixin
262 // ---------------------------------------------------------------------------
263
264 // NB: VC6 has a bug that causes linker errors if you have template methods
265 // in a class using __declspec(dllimport). The solution is to split such
266 // class into two classes, one that contains the template methods and does
267 // *not* use WXDLLIMPEXP_BASE and another class that contains the rest
268 // (with DLL linkage).
269 //
270 // We only do this for VC6 here, because the code is less efficient
271 // (Printf() has to use dynamic_cast<>) and because OpenWatcom compiler
272 // cannot compile this code.
273
274 #if defined(__VISUALC__) && __VISUALC__ < 1300
275 #define wxNEEDS_WXSTRING_PRINTF_MIXIN
276 #endif
277
278 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
279 // this class contains implementation of wxString's vararg methods, it's
280 // exported from wxBase DLL
281 class WXDLLIMPEXP_BASE wxStringPrintfMixinBase
282 {
283 protected:
284 wxStringPrintfMixinBase() {}
285
286 #if !wxUSE_UTF8_LOCALE_ONLY
287 int DoPrintfWchar(const wxChar *format, ...);
288 static wxString DoFormatWchar(const wxChar *format, ...);
289 #endif
290 #if wxUSE_UNICODE_UTF8
291 int DoPrintfUtf8(const char *format, ...);
292 static wxString DoFormatUtf8(const char *format, ...);
293 #endif
294 };
295
296 // this class contains template wrappers for wxString's vararg methods, it's
297 // intentionally *not* exported from the DLL in order to fix the VC6 bug
298 // described above
299 class wxStringPrintfMixin : public wxStringPrintfMixinBase
300 {
301 private:
302 // to further complicate things, we can't return wxString from
303 // wxStringPrintfMixin::Format() because wxString is not yet declared at
304 // this point; the solution is to use this fake type trait template - this
305 // way the compiler won't know the return type until Format() is used
306 // (this doesn't compile with Watcom, but VC6 compiles it just fine):
307 template<typename T> struct StringReturnType
308 {
309 typedef wxString type;
310 };
311
312 public:
313 // these are duplicated wxString methods, they're also declared below
314 // if !wxNEEDS_WXSTRING_PRINTF_MIXIN:
315
316 // static wxString Format(const wString& format, ...) ATTRIBUTE_PRINTF_1;
317 WX_DEFINE_VARARG_FUNC_SANS_N0(static typename StringReturnType<T1>::type,
318 Format, 1, (const wxFormatString&),
319 DoFormatWchar, DoFormatUtf8)
320 // We have to implement the version without template arguments manually
321 // because of the StringReturnType<> hack, although WX_DEFINE_VARARG_FUNC
322 // normally does it itself. It has to be a template so that we can use
323 // the hack, even though there's no real template parameter. We can't move
324 // it to wxStrig, because it would shadow these versions of Format() then.
325 template<typename T>
326 inline static typename StringReturnType<T>::type
327 Format(const T& fmt)
328 {
329 // NB: this doesn't compile if T is not (some form of) a string;
330 // this makes Format's prototype equivalent to
331 // Format(const wxFormatString& fmt)
332 return DoFormatWchar(wxFormatString(fmt));
333 }
334
335 // int Printf(const wxString& format, ...);
336 WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&),
337 DoPrintfWchar, DoPrintfUtf8)
338 // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2;
339 WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&),
340 DoPrintfWchar, DoPrintfUtf8)
341
342 protected:
343 wxStringPrintfMixin() : wxStringPrintfMixinBase() {}
344 };
345 #endif // wxNEEDS_WXSTRING_PRINTF_MIXIN
346
347
348 // ----------------------------------------------------------------------------
349 // wxString: string class trying to be compatible with std::string, MFC
350 // CString and wxWindows 1.x wxString all at once
351 // ---------------------------------------------------------------------------
352
353 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
354 // "non dll-interface class 'wxStringPrintfMixin' used as base interface
355 // for dll-interface class 'wxString'" -- this is OK in our case
356 #pragma warning (disable:4275)
357 #endif
358
359 #if wxUSE_UNICODE_UTF8
360 // see the comment near wxString::iterator for why we need this
361 class WXDLLIMPEXP_BASE wxStringIteratorNode
362 {
363 public:
364 wxStringIteratorNode()
365 : m_str(NULL), m_citer(NULL), m_iter(NULL), m_prev(NULL), m_next(NULL) {}
366 wxStringIteratorNode(const wxString *str,
367 wxStringImpl::const_iterator *citer)
368 { DoSet(str, citer, NULL); }
369 wxStringIteratorNode(const wxString *str, wxStringImpl::iterator *iter)
370 { DoSet(str, NULL, iter); }
371 ~wxStringIteratorNode()
372 { clear(); }
373
374 inline void set(const wxString *str, wxStringImpl::const_iterator *citer)
375 { clear(); DoSet(str, citer, NULL); }
376 inline void set(const wxString *str, wxStringImpl::iterator *iter)
377 { clear(); DoSet(str, NULL, iter); }
378
379 const wxString *m_str;
380 wxStringImpl::const_iterator *m_citer;
381 wxStringImpl::iterator *m_iter;
382 wxStringIteratorNode *m_prev, *m_next;
383
384 private:
385 inline void clear();
386 inline void DoSet(const wxString *str,
387 wxStringImpl::const_iterator *citer,
388 wxStringImpl::iterator *iter);
389
390 // the node belongs to a particular iterator instance, it's not copied
391 // when a copy of the iterator is made
392 DECLARE_NO_COPY_CLASS(wxStringIteratorNode)
393 };
394 #endif // wxUSE_UNICODE_UTF8
395
396 class WXDLLIMPEXP_BASE wxString
397 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
398 : public wxStringPrintfMixin
399 #endif
400 {
401 // NB: special care was taken in arranging the member functions in such order
402 // that all inline functions can be effectively inlined, verify that all
403 // performance critical functions are still inlined if you change order!
404 public:
405 // an 'invalid' value for string index, moved to this place due to a CW bug
406 static const size_t npos;
407
408 private:
409 // if we hadn't made these operators private, it would be possible to
410 // compile "wxString s; s = 17;" without any warnings as 17 is implicitly
411 // converted to char in C and we do have operator=(char)
412 //
413 // NB: we don't need other versions (short/long and unsigned) as attempt
414 // to assign another numeric type to wxString will now result in
415 // ambiguity between operator=(char) and operator=(int)
416 wxString& operator=(int);
417
418 // these methods are not implemented - there is _no_ conversion from int to
419 // string, you're doing something wrong if the compiler wants to call it!
420 //
421 // try `s << i' or `s.Printf("%d", i)' instead
422 wxString(int);
423
424
425 // buffer for holding temporary substring when using any of the methods
426 // that take (char*,size_t) or (wchar_t*,size_t) arguments:
427 template<typename T>
428 struct SubstrBufFromType
429 {
430 T data;
431 size_t len;
432
433 SubstrBufFromType(const T& data_, size_t len_)
434 : data(data_), len(len_)
435 {
436 wxASSERT_MSG( len != npos, "must have real length" );
437 }
438 };
439
440 #if wxUSE_UNICODE_UTF8
441 // even char* -> char* needs conversion, from locale charset to UTF-8
442 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromWC;
443 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromMB;
444 #elif wxUSE_UNICODE_WCHAR
445 typedef SubstrBufFromType<const wchar_t*> SubstrBufFromWC;
446 typedef SubstrBufFromType<wxWCharBuffer> SubstrBufFromMB;
447 #else
448 typedef SubstrBufFromType<const char*> SubstrBufFromMB;
449 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromWC;
450 #endif
451
452
453 // Functions implementing primitive operations on string data; wxString
454 // methods and iterators are implemented in terms of it. The differences
455 // between UTF-8 and wchar_t* representations of the string are mostly
456 // contained here.
457
458 #if wxUSE_UNICODE_UTF8
459 static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength,
460 const wxMBConv& conv);
461 static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength,
462 const wxMBConv& conv);
463 #elif wxUSE_UNICODE_WCHAR
464 static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength,
465 const wxMBConv& conv);
466 #else
467 static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength,
468 const wxMBConv& conv);
469 #endif
470
471 #if !wxUSE_UNICODE_UTF8 // wxUSE_UNICODE_WCHAR or !wxUSE_UNICODE
472 // returns C string encoded as the implementation expects:
473 #if wxUSE_UNICODE
474 static const wchar_t* ImplStr(const wchar_t* str)
475 { return str ? str : wxT(""); }
476 static const SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
477 { return SubstrBufFromWC(str, (str && n == npos) ? wxWcslen(str) : n); }
478 static wxWCharBuffer ImplStr(const char* str,
479 const wxMBConv& conv = wxConvLibc)
480 { return ConvertStr(str, npos, conv).data; }
481 static SubstrBufFromMB ImplStr(const char* str, size_t n,
482 const wxMBConv& conv = wxConvLibc)
483 { return ConvertStr(str, n, conv); }
484 #else
485 static const char* ImplStr(const char* str,
486 const wxMBConv& WXUNUSED(conv) = wxConvLibc)
487 { return str ? str : ""; }
488 static const SubstrBufFromMB ImplStr(const char* str, size_t n,
489 const wxMBConv& WXUNUSED(conv) = wxConvLibc)
490 { return SubstrBufFromMB(str, (str && n == npos) ? wxStrlen(str) : n); }
491 static wxCharBuffer ImplStr(const wchar_t* str)
492 { return ConvertStr(str, npos, wxConvLibc).data; }
493 static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
494 { return ConvertStr(str, n, wxConvLibc); }
495 #endif
496
497 // translates position index in wxString to/from index in underlying
498 // wxStringImpl:
499 static size_t PosToImpl(size_t pos) { return pos; }
500 static void PosLenToImpl(size_t pos, size_t len,
501 size_t *implPos, size_t *implLen)
502 { *implPos = pos; *implLen = len; }
503 static size_t LenToImpl(size_t len) { return len; }
504 static size_t PosFromImpl(size_t pos) { return pos; }
505
506 #else // wxUSE_UNICODE_UTF8
507
508 static wxCharBuffer ImplStr(const char* str,
509 const wxMBConv& conv = wxConvLibc)
510 { return ConvertStr(str, npos, conv).data; }
511 static SubstrBufFromMB ImplStr(const char* str, size_t n,
512 const wxMBConv& conv = wxConvLibc)
513 { return ConvertStr(str, n, conv); }
514
515 static wxCharBuffer ImplStr(const wchar_t* str)
516 { return ConvertStr(str, npos, wxMBConvUTF8()).data; }
517 static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
518 { return ConvertStr(str, n, wxMBConvUTF8()); }
519
520 size_t PosToImpl(size_t pos) const
521 {
522 if ( pos == 0 || pos == npos )
523 return pos;
524 else
525 return (begin() + pos).impl() - m_impl.begin();
526 }
527
528 void PosLenToImpl(size_t pos, size_t len, size_t *implPos, size_t *implLen) const;
529
530 size_t LenToImpl(size_t len) const
531 {
532 size_t pos, len2;
533 PosLenToImpl(0, len, &pos, &len2);
534 return len2;
535 }
536
537 size_t PosFromImpl(size_t pos) const
538 {
539 if ( pos == 0 || pos == npos )
540 return pos;
541 else
542 return const_iterator(this, m_impl.begin() + pos) - begin();
543 }
544 #endif // !wxUSE_UNICODE_UTF8/wxUSE_UNICODE_UTF8
545
546 public:
547 // standard types
548 typedef wxUniChar value_type;
549 typedef wxUniChar char_type;
550 typedef wxUniCharRef reference;
551 typedef wxChar* pointer;
552 typedef const wxChar* const_pointer;
553
554 typedef size_t size_type;
555 typedef wxUniChar const_reference;
556
557 #if wxUSE_STL
558 #if wxUSE_UNICODE_UTF8
559 // random access is not O(1), as required by Random Access Iterator
560 #define WX_STR_ITERATOR_TAG std::bidirectional_iterator_tag
561 #else
562 #define WX_STR_ITERATOR_TAG std::random_access_iterator_tag
563 #endif
564 #else
565 #define WX_STR_ITERATOR_TAG void /* dummy type */
566 #endif
567
568 #define WX_STR_ITERATOR_IMPL(iterator_name, pointer_type, reference_type) \
569 private: \
570 typedef wxStringImpl::iterator_name underlying_iterator; \
571 public: \
572 typedef WX_STR_ITERATOR_TAG iterator_category; \
573 typedef wxUniChar value_type; \
574 typedef int difference_type; \
575 typedef reference_type reference; \
576 typedef pointer_type pointer; \
577 \
578 reference operator[](size_t n) const { return *(*this + n); } \
579 \
580 iterator_name& operator++() \
581 { wxStringOperations::IncIter(m_cur); return *this; } \
582 iterator_name& operator--() \
583 { wxStringOperations::DecIter(m_cur); return *this; } \
584 iterator_name operator++(int) \
585 { \
586 iterator_name tmp = *this; \
587 wxStringOperations::IncIter(m_cur); \
588 return tmp; \
589 } \
590 iterator_name operator--(int) \
591 { \
592 iterator_name tmp = *this; \
593 wxStringOperations::DecIter(m_cur); \
594 return tmp; \
595 } \
596 \
597 iterator_name& operator+=(ptrdiff_t n) \
598 { \
599 m_cur = wxStringOperations::AddToIter(m_cur, n); \
600 return *this; \
601 } \
602 iterator_name& operator-=(ptrdiff_t n) \
603 { \
604 m_cur = wxStringOperations::AddToIter(m_cur, -n); \
605 return *this; \
606 } \
607 \
608 difference_type operator-(const iterator_name& i) const \
609 { return wxStringOperations::DiffIters(m_cur, i.m_cur); } \
610 \
611 bool operator==(const iterator_name& i) const \
612 { return m_cur == i.m_cur; } \
613 bool operator!=(const iterator_name& i) const \
614 { return m_cur != i.m_cur; } \
615 \
616 bool operator<(const iterator_name& i) const \
617 { return m_cur < i.m_cur; } \
618 bool operator>(const iterator_name& i) const \
619 { return m_cur > i.m_cur; } \
620 bool operator<=(const iterator_name& i) const \
621 { return m_cur <= i.m_cur; } \
622 bool operator>=(const iterator_name& i) const \
623 { return m_cur >= i.m_cur; } \
624 \
625 private: \
626 /* for internal wxString use only: */ \
627 underlying_iterator impl() const { return m_cur; } \
628 \
629 friend class wxString; \
630 friend class wxCStrData; \
631 \
632 private: \
633 underlying_iterator m_cur
634
635 class WXDLLIMPEXP_FWD_BASE const_iterator;
636
637 #if wxUSE_UNICODE_UTF8
638 // NB: In UTF-8 build, (non-const) iterator needs to keep reference
639 // to the underlying wxStringImpl, because UTF-8 is variable-length
640 // encoding and changing the value pointer to by an iterator (using
641 // its operator*) requires calling wxStringImpl::replace() if the old
642 // and new values differ in their encoding's length.
643 //
644 // Furthermore, the replace() call may invalid all iterators for the
645 // string, so we have to keep track of outstanding iterators and update
646 // them if replace() happens.
647 //
648 // This is implemented by maintaining linked list of iterators for every
649 // string and traversing it in wxUniCharRef::operator=(). Head of the
650 // list is stored in wxString. (FIXME-UTF8)
651
652 class WXDLLIMPEXP_BASE iterator
653 {
654 WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);
655
656 public:
657 iterator() {}
658 iterator(const iterator& i)
659 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
660 iterator& operator=(const iterator& i)
661 {
662 if (&i != this)
663 {
664 m_cur = i.m_cur;
665 m_node.set(i.str(), &m_cur);
666 }
667 return *this;
668 }
669
670 reference operator*()
671 { return wxUniCharRef::CreateForString(m_node, m_cur); }
672
673 iterator operator+(ptrdiff_t n) const
674 { return iterator(str(), wxStringOperations::AddToIter(m_cur, n)); }
675 iterator operator-(ptrdiff_t n) const
676 { return iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); }
677
678 private:
679 iterator(wxString *str, underlying_iterator ptr)
680 : m_cur(ptr), m_node(str, &m_cur) {}
681
682 wxString* str() const { return wx_const_cast(wxString*, m_node.m_str); }
683
684 wxStringIteratorNode m_node;
685
686 friend class const_iterator;
687 };
688
689 class WXDLLIMPEXP_BASE const_iterator
690 {
691 // NB: reference_type is intentionally value, not reference, the character
692 // may be encoded differently in wxString data:
693 WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar);
694
695 public:
696 const_iterator() {}
697 const_iterator(const const_iterator& i)
698 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
699 const_iterator(const iterator& i)
700 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
701
702 const_iterator& operator=(const const_iterator& i)
703 {
704 if (&i != this)
705 {
706 m_cur = i.m_cur;
707 m_node.set(i.str(), &m_cur);
708 }
709 return *this;
710 }
711 const_iterator& operator=(const iterator& i)
712 { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; }
713
714 reference operator*() const
715 { return wxStringOperations::DecodeChar(m_cur); }
716
717 const_iterator operator+(ptrdiff_t n) const
718 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, n)); }
719 const_iterator operator-(ptrdiff_t n) const
720 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); }
721
722 private:
723 // for internal wxString use only:
724 const_iterator(const wxString *str, underlying_iterator ptr)
725 : m_cur(ptr), m_node(str, &m_cur) {}
726
727 const wxString* str() const { return m_node.m_str; }
728
729 wxStringIteratorNode m_node;
730 };
731
732 size_t IterToImplPos(wxString::iterator i) const
733 { return wxStringImpl::const_iterator(i.impl()) - m_impl.begin(); }
734
735 #else // !wxUSE_UNICODE_UTF8
736
737 class WXDLLIMPEXP_BASE iterator
738 {
739 WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);
740
741 public:
742 iterator() {}
743 iterator(const iterator& i) : m_cur(i.m_cur) {}
744
745 reference operator*()
746 { return wxUniCharRef::CreateForString(m_cur); }
747
748 iterator operator+(ptrdiff_t n) const
749 { return iterator(wxStringOperations::AddToIter(m_cur, n)); }
750 iterator operator-(ptrdiff_t n) const
751 { return iterator(wxStringOperations::AddToIter(m_cur, -n)); }
752
753 private:
754 // for internal wxString use only:
755 iterator(underlying_iterator ptr) : m_cur(ptr) {}
756 iterator(wxString *WXUNUSED(str), underlying_iterator ptr) : m_cur(ptr) {}
757
758 friend class const_iterator;
759 };
760
761 class WXDLLIMPEXP_BASE const_iterator
762 {
763 // NB: reference_type is intentionally value, not reference, the character
764 // may be encoded differently in wxString data:
765 WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar);
766
767 public:
768 const_iterator() {}
769 const_iterator(const const_iterator& i) : m_cur(i.m_cur) {}
770 const_iterator(const iterator& i) : m_cur(i.m_cur) {}
771
772 reference operator*() const
773 { return wxStringOperations::DecodeChar(m_cur); }
774
775 const_iterator operator+(ptrdiff_t n) const
776 { return const_iterator(wxStringOperations::AddToIter(m_cur, n)); }
777 const_iterator operator-(ptrdiff_t n) const
778 { return const_iterator(wxStringOperations::AddToIter(m_cur, -n)); }
779
780 private:
781 // for internal wxString use only:
782 const_iterator(underlying_iterator ptr) : m_cur(ptr) {}
783 const_iterator(const wxString *WXUNUSED(str), underlying_iterator ptr)
784 : m_cur(ptr) {}
785 };
786 #endif // wxUSE_UNICODE_UTF8/!wxUSE_UNICODE_UTF8
787
788 #undef WX_STR_ITERATOR_TAG
789 #undef WX_STR_ITERATOR_IMPL
790
791 friend class iterator;
792 friend class const_iterator;
793
794 template <typename T>
795 class reverse_iterator_impl
796 {
797 public:
798 typedef T iterator_type;
799
800 typedef typename T::iterator_category iterator_category;
801 typedef typename T::value_type value_type;
802 typedef typename T::difference_type difference_type;
803 typedef typename T::reference reference;
804 typedef typename T::pointer *pointer;
805
806 reverse_iterator_impl() {}
807 reverse_iterator_impl(iterator_type i) : m_cur(i) {}
808 reverse_iterator_impl(const reverse_iterator_impl& ri)
809 : m_cur(ri.m_cur) {}
810
811 iterator_type base() const { return m_cur; }
812
813 reference operator*() const { return *(m_cur-1); }
814 reference operator[](size_t n) const { return *(*this + n); }
815
816 reverse_iterator_impl& operator++()
817 { --m_cur; return *this; }
818 reverse_iterator_impl operator++(int)
819 { reverse_iterator_impl tmp = *this; --m_cur; return tmp; }
820 reverse_iterator_impl& operator--()
821 { ++m_cur; return *this; }
822 reverse_iterator_impl operator--(int)
823 { reverse_iterator_impl tmp = *this; ++m_cur; return tmp; }
824
825 // NB: explicit <T> in the functions below is to keep BCC 5.5 happy
826 reverse_iterator_impl operator+(ptrdiff_t n) const
827 { return reverse_iterator_impl<T>(m_cur - n); }
828 reverse_iterator_impl operator-(ptrdiff_t n) const
829 { return reverse_iterator_impl<T>(m_cur + n); }
830 reverse_iterator_impl operator+=(ptrdiff_t n)
831 { m_cur -= n; return *this; }
832 reverse_iterator_impl operator-=(ptrdiff_t n)
833 { m_cur += n; return *this; }
834
835 unsigned operator-(const reverse_iterator_impl& i) const
836 { return i.m_cur - m_cur; }
837
838 bool operator==(const reverse_iterator_impl& ri) const
839 { return m_cur == ri.m_cur; }
840 bool operator!=(const reverse_iterator_impl& ri) const
841 { return !(*this == ri); }
842
843 bool operator<(const reverse_iterator_impl& i) const
844 { return m_cur > i.m_cur; }
845 bool operator>(const reverse_iterator_impl& i) const
846 { return m_cur < i.m_cur; }
847 bool operator<=(const reverse_iterator_impl& i) const
848 { return m_cur >= i.m_cur; }
849 bool operator>=(const reverse_iterator_impl& i) const
850 { return m_cur <= i.m_cur; }
851
852 private:
853 iterator_type m_cur;
854 };
855
856 typedef reverse_iterator_impl<iterator> reverse_iterator;
857 typedef reverse_iterator_impl<const_iterator> const_reverse_iterator;
858
859 private:
860 // used to transform an expression built using c_str() (and hence of type
861 // wxCStrData) to an iterator into the string
862 static const_iterator CreateConstIterator(const wxCStrData& data)
863 {
864 return const_iterator(data.m_str,
865 (data.m_str->begin() + data.m_offset).impl());
866 }
867
868 // in UTF-8 STL build, creation from std::string requires conversion under
869 // non-UTF8 locales, so we can't have and use wxString(wxStringImpl) ctor;
870 // instead we define dummy type that lets us have wxString ctor for creation
871 // from wxStringImpl that couldn't be used by user code (in all other builds,
872 // "standard" ctors can be used):
873 #if wxUSE_UNICODE_UTF8 && wxUSE_STL_BASED_WXSTRING
874 struct CtorFromStringImplTag {};
875
876 wxString(CtorFromStringImplTag* WXUNUSED(dummy), const wxStringImpl& src)
877 : m_impl(src) {}
878
879 static wxString FromImpl(const wxStringImpl& src)
880 { return wxString((CtorFromStringImplTag*)NULL, src); }
881 #else
882 #if !wxUSE_STL_BASED_WXSTRING
883 wxString(const wxStringImpl& src) : m_impl(src) { }
884 // else: already defined as wxString(wxStdString) below
885 #endif
886 static wxString FromImpl(const wxStringImpl& src) { return wxString(src); }
887 #endif
888
889 public:
890 // constructors and destructor
891 // ctor for an empty string
892 wxString() {}
893
894 // copy ctor
895 wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { }
896
897 // string containing nRepeat copies of ch
898 wxString(wxUniChar ch, size_t nRepeat = 1 )
899 { assign(nRepeat, ch); }
900 wxString(size_t nRepeat, wxUniChar ch)
901 { assign(nRepeat, ch); }
902 wxString(wxUniCharRef ch, size_t nRepeat = 1)
903 { assign(nRepeat, ch); }
904 wxString(size_t nRepeat, wxUniCharRef ch)
905 { assign(nRepeat, ch); }
906 wxString(char ch, size_t nRepeat = 1)
907 { assign(nRepeat, ch); }
908 wxString(size_t nRepeat, char ch)
909 { assign(nRepeat, ch); }
910 wxString(wchar_t ch, size_t nRepeat = 1)
911 { assign(nRepeat, ch); }
912 wxString(size_t nRepeat, wchar_t ch)
913 { assign(nRepeat, ch); }
914
915 // ctors from char* strings:
916 wxString(const char *psz)
917 : m_impl(ImplStr(psz)) {}
918 wxString(const char *psz, const wxMBConv& conv)
919 : m_impl(ImplStr(psz, conv)) {}
920 wxString(const char *psz, size_t nLength)
921 { assign(psz, nLength); }
922 wxString(const char *psz, const wxMBConv& conv, size_t nLength)
923 {
924 SubstrBufFromMB str(ImplStr(psz, nLength, conv));
925 m_impl.assign(str.data, str.len);
926 }
927
928 // and unsigned char*:
929 wxString(const unsigned char *psz)
930 : m_impl(ImplStr((const char*)psz)) {}
931 wxString(const unsigned char *psz, const wxMBConv& conv)
932 : m_impl(ImplStr((const char*)psz, conv)) {}
933 wxString(const unsigned char *psz, size_t nLength)
934 { assign((const char*)psz, nLength); }
935 wxString(const unsigned char *psz, const wxMBConv& conv, size_t nLength)
936 {
937 SubstrBufFromMB str(ImplStr((const char*)psz, nLength, conv));
938 m_impl.assign(str.data, str.len);
939 }
940
941 // ctors from wchar_t* strings:
942 wxString(const wchar_t *pwz)
943 : m_impl(ImplStr(pwz)) {}
944 wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv))
945 : m_impl(ImplStr(pwz)) {}
946 wxString(const wchar_t *pwz, size_t nLength)
947 { assign(pwz, nLength); }
948 wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv), size_t nLength)
949 { assign(pwz, nLength); }
950
951 wxString(const wxCharBuffer& buf)
952 { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length
953 wxString(const wxWCharBuffer& buf)
954 { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length
955
956 wxString(const wxCStrData& cstr)
957 : m_impl(cstr.AsString().m_impl) { }
958
959 // as we provide both ctors with this signature for both char and unsigned
960 // char string, we need to provide one for wxCStrData to resolve ambiguity
961 wxString(const wxCStrData& cstr, size_t nLength)
962 : m_impl(cstr.AsString().Mid(0, nLength).m_impl) {}
963
964 // and because wxString is convertible to wxCStrData and const wxChar *
965 // we also need to provide this one
966 wxString(const wxString& str, size_t nLength)
967 { assign(str, nLength); }
968
969 // even if we're not built with wxUSE_STL == 1 it is very convenient to allow
970 // implicit conversions from std::string to wxString and vice verse as this
971 // allows to use the same strings in non-GUI and GUI code, however we don't
972 // want to unconditionally add this ctor as it would make wx lib dependent on
973 // libstdc++ on some Linux versions which is bad, so instead we ask the
974 // client code to define this wxUSE_STD_STRING symbol if they need it
975 #if wxUSE_STD_STRING
976 #if wxUSE_UNICODE_WCHAR
977 wxString(const wxStdWideString& str) : m_impl(str) {}
978 #else // UTF-8 or ANSI
979 wxString(const wxStdWideString& str)
980 { assign(str.c_str(), str.length()); }
981 #endif
982
983 #if !wxUSE_UNICODE // ANSI build
984 // FIXME-UTF8: do this in UTF8 build #if wxUSE_UTF8_LOCALE_ONLY, too
985 wxString(const std::string& str) : m_impl(str) {}
986 #else // Unicode
987 wxString(const std::string& str)
988 { assign(str.c_str(), str.length()); }
989 #endif
990 #endif // wxUSE_STD_STRING
991
992 // Unlike ctor from std::string, we provide conversion to std::string only
993 // if wxUSE_STL and not merely wxUSE_STD_STRING (which is on by default),
994 // because it conflicts with operator const char/wchar_t*:
995 #if wxUSE_STL
996 #if wxUSE_UNICODE_WCHAR && wxUSE_STL_BASED_WXSTRING
997 // wxStringImpl is std::string in the encoding we want
998 operator const wxStdWideString&() const { return m_impl; }
999 #else
1000 // wxStringImpl is either not std::string or needs conversion
1001 operator wxStdWideString() const
1002 // FIXME-UTF8: broken for embedded NULs
1003 { return wxStdWideString(wc_str()); }
1004 #endif
1005
1006 #if (!wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY) && wxUSE_STL_BASED_WXSTRING
1007 // wxStringImpl is std::string in the encoding we want
1008 operator const std::string&() const { return m_impl; }
1009 #else
1010 // wxStringImpl is either not std::string or needs conversion
1011 operator std::string() const
1012 // FIXME-UTF8: broken for embedded NULs
1013 { return std::string(mb_str()); }
1014 #endif
1015 #endif // wxUSE_STL
1016
1017 // first valid index position
1018 const_iterator begin() const { return const_iterator(this, m_impl.begin()); }
1019 iterator begin() { return iterator(this, m_impl.begin()); }
1020 // position one after the last valid one
1021 const_iterator end() const { return const_iterator(this, m_impl.end()); }
1022 iterator end() { return iterator(this, m_impl.end()); }
1023
1024 // first element of the reversed string
1025 const_reverse_iterator rbegin() const
1026 { return const_reverse_iterator(end()); }
1027 reverse_iterator rbegin()
1028 { return reverse_iterator(end()); }
1029 // one beyond the end of the reversed string
1030 const_reverse_iterator rend() const
1031 { return const_reverse_iterator(begin()); }
1032 reverse_iterator rend()
1033 { return reverse_iterator(begin()); }
1034
1035 // std::string methods:
1036 #if wxUSE_UNICODE_UTF8
1037 size_t length() const { return end() - begin(); } // FIXME-UTF8: optimize!
1038 #else
1039 size_t length() const { return m_impl.length(); }
1040 #endif
1041
1042 size_type size() const { return length(); }
1043 size_type max_size() const { return npos; }
1044
1045 bool empty() const { return m_impl.empty(); }
1046
1047 size_type capacity() const { return m_impl.capacity(); } // FIXME-UTF8
1048 void reserve(size_t sz) { m_impl.reserve(sz); } // FIXME-UTF8
1049
1050 void resize(size_t nSize, wxUniChar ch = wxT('\0'))
1051 {
1052 const size_t len = length();
1053 if ( nSize == len)
1054 return;
1055
1056 #if wxUSE_UNICODE_UTF8
1057 if ( nSize < len )
1058 {
1059 // we can't use wxStringImpl::resize() for truncating the string as it
1060 // counts in bytes, not characters
1061 erase(nSize);
1062 return;
1063 }
1064
1065 // we also can't use (presumably more efficient) resize() if we have to
1066 // append characters taking more than one byte
1067 if ( !ch.IsAscii() )
1068 append(nSize - len, ch);
1069 else
1070 #endif // wxUSE_UNICODE_UTF8
1071 m_impl.resize(nSize, (wxStringCharType)ch);
1072 }
1073
1074 wxString substr(size_t nStart = 0, size_t nLen = npos) const
1075 {
1076 size_t pos, len;
1077 PosLenToImpl(nStart, nLen, &pos, &len);
1078 return FromImpl(m_impl.substr(pos, len));
1079 }
1080
1081 // generic attributes & operations
1082 // as standard strlen()
1083 size_t Len() const { return length(); }
1084 // string contains any characters?
1085 bool IsEmpty() const { return empty(); }
1086 // empty string is "false", so !str will return true
1087 bool operator!() const { return empty(); }
1088 // truncate the string to given length
1089 wxString& Truncate(size_t uiLen);
1090 // empty string contents
1091 void Empty()
1092 {
1093 Truncate(0);
1094
1095 wxASSERT_MSG( empty(), _T("string not empty after call to Empty()?") );
1096 }
1097 // empty the string and free memory
1098 void Clear()
1099 {
1100 wxString tmp(wxEmptyString);
1101 swap(tmp);
1102 }
1103
1104 // contents test
1105 // Is an ascii value
1106 bool IsAscii() const;
1107 // Is a number
1108 bool IsNumber() const;
1109 // Is a word
1110 bool IsWord() const;
1111
1112 // data access (all indexes are 0 based)
1113 // read access
1114 wxUniChar at(size_t n) const
1115 { return *(begin() + n); } // FIXME-UTF8: optimize?
1116 wxUniChar GetChar(size_t n) const
1117 { return at(n); }
1118 // read/write access
1119 wxUniCharRef at(size_t n)
1120 { return *(begin() + n); } // FIXME-UTF8: optimize?
1121 wxUniCharRef GetWritableChar(size_t n)
1122 { return at(n); }
1123 // write access
1124 void SetChar(size_t n, wxUniChar ch)
1125 { at(n) = ch; }
1126
1127 // get last character
1128 wxUniChar Last() const
1129 {
1130 wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
1131 return *rbegin();
1132 }
1133
1134 // get writable last character
1135 wxUniCharRef Last()
1136 {
1137 wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
1138 return *rbegin();
1139 }
1140
1141 /*
1142 Note that we we must define all of the overloads below to avoid
1143 ambiguity when using str[0].
1144 */
1145 wxUniChar operator[](int n) const
1146 { return at(n); }
1147 wxUniChar operator[](long n) const
1148 { return at(n); }
1149 wxUniChar operator[](size_t n) const
1150 { return at(n); }
1151 #ifndef wxSIZE_T_IS_UINT
1152 wxUniChar operator[](unsigned int n) const
1153 { return at(n); }
1154 #endif // size_t != unsigned int
1155
1156 // operator versions of GetWriteableChar()
1157 wxUniCharRef operator[](int n)
1158 { return at(n); }
1159 wxUniCharRef operator[](long n)
1160 { return at(n); }
1161 wxUniCharRef operator[](size_t n)
1162 { return at(n); }
1163 #ifndef wxSIZE_T_IS_UINT
1164 wxUniCharRef operator[](unsigned int n)
1165 { return at(n); }
1166 #endif // size_t != unsigned int
1167
1168 // explicit conversion to C string (use this with printf()!)
1169 wxCStrData c_str() const { return wxCStrData(this); }
1170 wxCStrData data() const { return c_str(); }
1171
1172 // implicit conversion to C string
1173 operator wxCStrData() const { return c_str(); }
1174
1175 // the first two operators conflict with operators for conversion to
1176 // std::string and they must be disabled in STL build; the next one only
1177 // makes sense if conversions to char* are also defined and not defining it
1178 // in STL build also helps us to get more clear error messages for the code
1179 // which relies on implicit conversion to char* in STL build
1180 #if !wxUSE_STL
1181 operator const char*() const { return c_str(); }
1182 operator const wchar_t*() const { return c_str(); }
1183
1184 // implicit conversion to untyped pointer for compatibility with previous
1185 // wxWidgets versions: this is the same as conversion to const char * so it
1186 // may fail!
1187 operator const void*() const { return c_str(); }
1188 #endif // wxUSE_STL
1189
1190 // identical to c_str(), for MFC compatibility
1191 const wxCStrData GetData() const { return c_str(); }
1192
1193 // explicit conversion to C string in internal representation (char*,
1194 // wchar_t*, UTF-8-encoded char*, depending on the build):
1195 const wxStringCharType *wx_str() const { return m_impl.c_str(); }
1196
1197 // conversion to *non-const* multibyte or widestring buffer; modifying
1198 // returned buffer won't affect the string, these methods are only useful
1199 // for passing values to const-incorrect functions
1200 wxWritableCharBuffer char_str(const wxMBConv& conv = wxConvLibc) const
1201 { return mb_str(conv); }
1202 wxWritableWCharBuffer wchar_str() const { return wc_str(); }
1203
1204 // conversion to/from plain (i.e. 7 bit) ASCII: this is useful for
1205 // converting numbers or strings which are certain not to contain special
1206 // chars (typically system functions, X atoms, environment variables etc.)
1207 //
1208 // the behaviour of these functions with the strings containing anything
1209 // else than 7 bit ASCII characters is undefined, use at your own risk.
1210 #if wxUSE_UNICODE
1211 static wxString FromAscii(const char *ascii, size_t len);
1212 static wxString FromAscii(const char *ascii);
1213 static wxString FromAscii(char ascii);
1214 const wxCharBuffer ToAscii() const;
1215 #else // ANSI
1216 static wxString FromAscii(const char *ascii) { return wxString( ascii ); }
1217 static wxString FromAscii(const char *ascii, size_t len)
1218 { return wxString( ascii, len ); }
1219 static wxString FromAscii(char ascii) { return wxString( ascii ); }
1220 const char *ToAscii() const { return c_str(); }
1221 #endif // Unicode/!Unicode
1222
1223 // also provide unsigned char overloads as signed/unsigned doesn't matter
1224 // for 7 bit ASCII characters
1225 static wxString FromAscii(const unsigned char *ascii)
1226 { return FromAscii((const char *)ascii); }
1227 static wxString FromAscii(const unsigned char *ascii, size_t len)
1228 { return FromAscii((const char *)ascii, len); }
1229
1230 // conversion to/from UTF-8:
1231 #if wxUSE_UNICODE_UTF8
1232 static wxString FromUTF8(const char *utf8)
1233 {
1234 if ( !utf8 )
1235 return wxEmptyString;
1236
1237 wxASSERT( wxStringOperations::IsValidUtf8String(utf8) );
1238 return FromImpl(wxStringImpl(utf8));
1239 }
1240 static wxString FromUTF8(const char *utf8, size_t len)
1241 {
1242 if ( !utf8 )
1243 return wxEmptyString;
1244 if ( len == npos )
1245 return FromUTF8(utf8);
1246
1247 wxASSERT( wxStringOperations::IsValidUtf8String(utf8, len) );
1248 return FromImpl(wxStringImpl(utf8, len));
1249 }
1250 const char* utf8_str() const { return wx_str(); }
1251 const char* ToUTF8() const { return wx_str(); }
1252 #elif wxUSE_UNICODE_WCHAR
1253 static wxString FromUTF8(const char *utf8)
1254 { return wxString(utf8, wxMBConvUTF8()); }
1255 static wxString FromUTF8(const char *utf8, size_t len)
1256 { return wxString(utf8, wxMBConvUTF8(), len); }
1257 const wxCharBuffer utf8_str() const { return mb_str(wxMBConvUTF8()); }
1258 const wxCharBuffer ToUTF8() const { return utf8_str(); }
1259 #else // ANSI
1260 static wxString FromUTF8(const char *utf8)
1261 { return wxString(wxMBConvUTF8().cMB2WC(utf8)); }
1262 static wxString FromUTF8(const char *utf8, size_t len)
1263 {
1264 size_t wlen;
1265 wxWCharBuffer buf(wxMBConvUTF8().cMB2WC(utf8, len == npos ? wxNO_LEN : len, &wlen));
1266 return wxString(buf.data(), wlen);
1267 }
1268 const wxCharBuffer utf8_str() const
1269 { return wxMBConvUTF8().cWC2MB(wc_str()); }
1270 const wxCharBuffer ToUTF8() const { return utf8_str(); }
1271 #endif
1272
1273 // functions for storing binary data in wxString:
1274 #if wxUSE_UNICODE
1275 static wxString From8BitData(const char *data, size_t len)
1276 { return wxString(data, wxConvISO8859_1, len); }
1277 // version for NUL-terminated data:
1278 static wxString From8BitData(const char *data)
1279 { return wxString(data, wxConvISO8859_1); }
1280 const wxCharBuffer To8BitData() const { return mb_str(wxConvISO8859_1); }
1281 #else // ANSI
1282 static wxString From8BitData(const char *data, size_t len)
1283 { return wxString(data, len); }
1284 // version for NUL-terminated data:
1285 static wxString From8BitData(const char *data)
1286 { return wxString(data); }
1287 const char *To8BitData() const { return c_str(); }
1288 #endif // Unicode/ANSI
1289
1290 // conversions with (possible) format conversions: have to return a
1291 // buffer with temporary data
1292 //
1293 // the functions defined (in either Unicode or ANSI) mode are mb_str() to
1294 // return an ANSI (multibyte) string, wc_str() to return a wide string and
1295 // fn_str() to return a string which should be used with the OS APIs
1296 // accepting the file names. The return value is always the same, but the
1297 // type differs because a function may either return pointer to the buffer
1298 // directly or have to use intermediate buffer for translation.
1299 #if wxUSE_UNICODE
1300
1301 #if wxUSE_UTF8_LOCALE_ONLY
1302 const char* mb_str() const { return wx_str(); }
1303 const wxCharBuffer mb_str(const wxMBConv& conv) const;
1304 #else
1305 const wxCharBuffer mb_str(const wxMBConv& conv = wxConvLibc) const;
1306 #endif
1307
1308 const wxWX2MBbuf mbc_str() const { return mb_str(*wxConvCurrent); }
1309
1310 #if wxUSE_UNICODE_WCHAR
1311 const wxChar* wc_str() const { return wx_str(); }
1312 #elif wxUSE_UNICODE_UTF8
1313 const wxWCharBuffer wc_str() const;
1314 #endif
1315 // for compatibility with !wxUSE_UNICODE version
1316 const wxWX2WCbuf wc_str(const wxMBConv& WXUNUSED(conv)) const
1317 { return wc_str(); }
1318
1319 #if wxMBFILES
1320 const wxCharBuffer fn_str() const { return mb_str(wxConvFile); }
1321 #else // !wxMBFILES
1322 const wxWX2WCbuf fn_str() const { return wc_str(); }
1323 #endif // wxMBFILES/!wxMBFILES
1324
1325 #else // ANSI
1326 const wxChar* mb_str() const { return wx_str(); }
1327
1328 // for compatibility with wxUSE_UNICODE version
1329 const wxChar* mb_str(const wxMBConv& WXUNUSED(conv)) const { return wx_str(); }
1330
1331 const wxWX2MBbuf mbc_str() const { return mb_str(); }
1332
1333 #if wxUSE_WCHAR_T
1334 const wxWCharBuffer wc_str(const wxMBConv& conv = wxConvLibc) const;
1335 #endif // wxUSE_WCHAR_T
1336 const wxCharBuffer fn_str() const { return wxConvFile.cWC2WX( wc_str( wxConvLibc ) ); }
1337 #endif // Unicode/ANSI
1338
1339 // overloaded assignment
1340 // from another wxString
1341 wxString& operator=(const wxString& stringSrc)
1342 { if (&stringSrc != this) m_impl = stringSrc.m_impl; return *this; }
1343 wxString& operator=(const wxCStrData& cstr)
1344 { return *this = cstr.AsString(); }
1345 // from a character
1346 wxString& operator=(wxUniChar ch)
1347 {
1348 #if wxUSE_UNICODE_UTF8
1349 if ( !ch.IsAscii() )
1350 m_impl = wxStringOperations::EncodeChar(ch);
1351 else
1352 #endif
1353 m_impl = (wxStringCharType)ch;
1354 return *this;
1355 }
1356 wxString& operator=(wxUniCharRef ch)
1357 { return operator=((wxUniChar)ch); }
1358 wxString& operator=(char ch)
1359 { return operator=(wxUniChar(ch)); }
1360 wxString& operator=(unsigned char ch)
1361 { return operator=(wxUniChar(ch)); }
1362 wxString& operator=(wchar_t ch)
1363 { return operator=(wxUniChar(ch)); }
1364 // from a C string - STL probably will crash on NULL,
1365 // so we need to compensate in that case
1366 #if wxUSE_STL_BASED_WXSTRING
1367 wxString& operator=(const char *psz)
1368 { if (psz) m_impl = ImplStr(psz); else Clear(); return *this; }
1369 wxString& operator=(const wchar_t *pwz)
1370 { if (pwz) m_impl = ImplStr(pwz); else Clear(); return *this; }
1371 #else
1372 wxString& operator=(const char *psz)
1373 { m_impl = ImplStr(psz); return *this; }
1374 wxString& operator=(const wchar_t *pwz)
1375 { m_impl = ImplStr(pwz); return *this; }
1376 #endif
1377 wxString& operator=(const unsigned char *psz)
1378 { return operator=((const char*)psz); }
1379
1380 // from wxWCharBuffer
1381 wxString& operator=(const wxWCharBuffer& s)
1382 { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs
1383 // from wxCharBuffer
1384 wxString& operator=(const wxCharBuffer& s)
1385 { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs
1386
1387 // string concatenation
1388 // in place concatenation
1389 /*
1390 Concatenate and return the result. Note that the left to right
1391 associativity of << allows to write things like "str << str1 << str2
1392 << ..." (unlike with +=)
1393 */
1394 // string += string
1395 wxString& operator<<(const wxString& s)
1396 {
1397 #if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
1398 wxASSERT_MSG( s.IsValid(),
1399 _T("did you forget to call UngetWriteBuf()?") );
1400 #endif
1401
1402 append(s);
1403 return *this;
1404 }
1405 // string += C string
1406 wxString& operator<<(const char *psz)
1407 { append(psz); return *this; }
1408 wxString& operator<<(const wchar_t *pwz)
1409 { append(pwz); return *this; }
1410 wxString& operator<<(const wxCStrData& psz)
1411 { append(psz.AsString()); return *this; }
1412 // string += char
1413 wxString& operator<<(wxUniChar ch) { append(1, ch); return *this; }
1414 wxString& operator<<(wxUniCharRef ch) { append(1, ch); return *this; }
1415 wxString& operator<<(char ch) { append(1, ch); return *this; }
1416 wxString& operator<<(unsigned char ch) { append(1, ch); return *this; }
1417 wxString& operator<<(wchar_t ch) { append(1, ch); return *this; }
1418
1419 // string += buffer (i.e. from wxGetString)
1420 wxString& operator<<(const wxWCharBuffer& s)
1421 { return operator<<((const wchar_t *)s); }
1422 wxString& operator<<(const wxCharBuffer& s)
1423 { return operator<<((const char *)s); }
1424
1425 // string += C string
1426 wxString& Append(const wxString& s)
1427 {
1428 // test for empty() to share the string if possible
1429 if ( empty() )
1430 *this = s;
1431 else
1432 append(s);
1433 return *this;
1434 }
1435 wxString& Append(const char* psz)
1436 { append(psz); return *this; }
1437 wxString& Append(const wchar_t* pwz)
1438 { append(pwz); return *this; }
1439 wxString& Append(const wxCStrData& psz)
1440 { append(psz); return *this; }
1441 wxString& Append(const wxCharBuffer& psz)
1442 { append(psz); return *this; }
1443 wxString& Append(const wxWCharBuffer& psz)
1444 { append(psz); return *this; }
1445 // append count copies of given character
1446 wxString& Append(wxUniChar ch, size_t count = 1u)
1447 { append(count, ch); return *this; }
1448 wxString& Append(wxUniCharRef ch, size_t count = 1u)
1449 { append(count, ch); return *this; }
1450 wxString& Append(char ch, size_t count = 1u)
1451 { append(count, ch); return *this; }
1452 wxString& Append(unsigned char ch, size_t count = 1u)
1453 { append(count, ch); return *this; }
1454 wxString& Append(wchar_t ch, size_t count = 1u)
1455 { append(count, ch); return *this; }
1456 wxString& Append(const char* psz, size_t nLen)
1457 { append(psz, nLen); return *this; }
1458 wxString& Append(const wchar_t* pwz, size_t nLen)
1459 { append(pwz, nLen); return *this; }
1460
1461 // prepend a string, return the string itself
1462 wxString& Prepend(const wxString& str)
1463 { *this = str + *this; return *this; }
1464
1465 // non-destructive concatenation
1466 // two strings
1467 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string1,
1468 const wxString& string2);
1469 // string with a single char
1470 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxUniChar ch);
1471 // char with a string
1472 friend wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string);
1473 // string with C string
1474 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string,
1475 const char *psz);
1476 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string,
1477 const wchar_t *pwz);
1478 // C string with string
1479 friend wxString WXDLLIMPEXP_BASE operator+(const char *psz,
1480 const wxString& string);
1481 friend wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz,
1482 const wxString& string);
1483
1484 // stream-like functions
1485 // insert an int into string
1486 wxString& operator<<(int i)
1487 { return (*this) << Format(_T("%d"), i); }
1488 // insert an unsigned int into string
1489 wxString& operator<<(unsigned int ui)
1490 { return (*this) << Format(_T("%u"), ui); }
1491 // insert a long into string
1492 wxString& operator<<(long l)
1493 { return (*this) << Format(_T("%ld"), l); }
1494 // insert an unsigned long into string
1495 wxString& operator<<(unsigned long ul)
1496 { return (*this) << Format(_T("%lu"), ul); }
1497 #if defined wxLongLong_t && !defined wxLongLongIsLong
1498 // insert a long long if they exist and aren't longs
1499 wxString& operator<<(wxLongLong_t ll)
1500 {
1501 const wxChar *fmt = _T("%") wxLongLongFmtSpec _T("d");
1502 return (*this) << Format(fmt, ll);
1503 }
1504 // insert an unsigned long long
1505 wxString& operator<<(wxULongLong_t ull)
1506 {
1507 const wxChar *fmt = _T("%") wxLongLongFmtSpec _T("u");
1508 return (*this) << Format(fmt , ull);
1509 }
1510 #endif
1511 // insert a float into string
1512 wxString& operator<<(float f)
1513 { return (*this) << Format(_T("%f"), f); }
1514 // insert a double into string
1515 wxString& operator<<(double d)
1516 { return (*this) << Format(_T("%g"), d); }
1517
1518 // string comparison
1519 // case-sensitive comparison (returns a value < 0, = 0 or > 0)
1520 int Cmp(const char *psz) const
1521 { return compare(psz); }
1522 int Cmp(const wchar_t *pwz) const
1523 { return compare(pwz); }
1524 int Cmp(const wxString& s) const
1525 { return compare(s); }
1526 int Cmp(const wxCStrData& s) const
1527 { return compare(s); }
1528 int Cmp(const wxCharBuffer& s) const
1529 { return compare(s); }
1530 int Cmp(const wxWCharBuffer& s) const
1531 { return compare(s); }
1532 // same as Cmp() but not case-sensitive
1533 int CmpNoCase(const wxString& s) const;
1534
1535 // test for the string equality, either considering case or not
1536 // (if compareWithCase then the case matters)
1537 bool IsSameAs(const wxString& str, bool compareWithCase = true) const
1538 {
1539 #if !wxUSE_UNICODE_UTF8
1540 // in UTF-8 build, length() is O(n) and doing this would be _slower_
1541 if ( length() != str.length() )
1542 return false;
1543 #endif
1544 return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0;
1545 }
1546 bool IsSameAs(const char *str, bool compareWithCase = true) const
1547 { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
1548 bool IsSameAs(const wchar_t *str, bool compareWithCase = true) const
1549 { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
1550
1551 bool IsSameAs(const wxCStrData& str, bool compareWithCase = true) const
1552 { return IsSameAs(str.AsString(), compareWithCase); }
1553 bool IsSameAs(const wxCharBuffer& str, bool compareWithCase = true) const
1554 { return IsSameAs(str.data(), compareWithCase); }
1555 bool IsSameAs(const wxWCharBuffer& str, bool compareWithCase = true) const
1556 { return IsSameAs(str.data(), compareWithCase); }
1557 // comparison with a single character: returns true if equal
1558 bool IsSameAs(wxUniChar c, bool compareWithCase = true) const;
1559 // FIXME-UTF8: remove these overloads
1560 bool IsSameAs(wxUniCharRef c, bool compareWithCase = true) const
1561 { return IsSameAs(wxUniChar(c), compareWithCase); }
1562 bool IsSameAs(char c, bool compareWithCase = true) const
1563 { return IsSameAs(wxUniChar(c), compareWithCase); }
1564 bool IsSameAs(unsigned char c, bool compareWithCase = true) const
1565 { return IsSameAs(wxUniChar(c), compareWithCase); }
1566 bool IsSameAs(wchar_t c, bool compareWithCase = true) const
1567 { return IsSameAs(wxUniChar(c), compareWithCase); }
1568 bool IsSameAs(int c, bool compareWithCase = true) const
1569 { return IsSameAs(wxUniChar(c), compareWithCase); }
1570
1571 // simple sub-string extraction
1572 // return substring starting at nFirst of length nCount (or till the end
1573 // if nCount = default value)
1574 wxString Mid(size_t nFirst, size_t nCount = npos) const;
1575
1576 // operator version of Mid()
1577 wxString operator()(size_t start, size_t len) const
1578 { return Mid(start, len); }
1579
1580 // check if the string starts with the given prefix and return the rest
1581 // of the string in the provided pointer if it is not NULL; otherwise
1582 // return false
1583 bool StartsWith(const wxString& prefix, wxString *rest = NULL) const;
1584 // check if the string ends with the given suffix and return the
1585 // beginning of the string before the suffix in the provided pointer if
1586 // it is not NULL; otherwise return false
1587 bool EndsWith(const wxString& suffix, wxString *rest = NULL) const;
1588
1589 // get first nCount characters
1590 wxString Left(size_t nCount) const;
1591 // get last nCount characters
1592 wxString Right(size_t nCount) const;
1593 // get all characters before the first occurance of ch
1594 // (returns the whole string if ch not found)
1595 wxString BeforeFirst(wxUniChar ch) const;
1596 // get all characters before the last occurence of ch
1597 // (returns empty string if ch not found)
1598 wxString BeforeLast(wxUniChar ch) const;
1599 // get all characters after the first occurence of ch
1600 // (returns empty string if ch not found)
1601 wxString AfterFirst(wxUniChar ch) const;
1602 // get all characters after the last occurence of ch
1603 // (returns the whole string if ch not found)
1604 wxString AfterLast(wxUniChar ch) const;
1605
1606 // for compatibility only, use more explicitly named functions above
1607 wxString Before(wxUniChar ch) const { return BeforeLast(ch); }
1608 wxString After(wxUniChar ch) const { return AfterFirst(ch); }
1609
1610 // case conversion
1611 // convert to upper case in place, return the string itself
1612 wxString& MakeUpper();
1613 // convert to upper case, return the copy of the string
1614 // Here's something to remember: BC++ doesn't like returns in inlines.
1615 wxString Upper() const ;
1616 // convert to lower case in place, return the string itself
1617 wxString& MakeLower();
1618 // convert to lower case, return the copy of the string
1619 wxString Lower() const ;
1620
1621 // trimming/padding whitespace (either side) and truncating
1622 // remove spaces from left or from right (default) side
1623 wxString& Trim(bool bFromRight = true);
1624 // add nCount copies chPad in the beginning or at the end (default)
1625 wxString& Pad(size_t nCount, wxUniChar chPad = wxT(' '), bool bFromRight = true);
1626
1627 // searching and replacing
1628 // searching (return starting index, or -1 if not found)
1629 int Find(wxUniChar ch, bool bFromEnd = false) const; // like strchr/strrchr
1630 int Find(wxUniCharRef ch, bool bFromEnd = false) const
1631 { return Find(wxUniChar(ch), bFromEnd); }
1632 int Find(char ch, bool bFromEnd = false) const
1633 { return Find(wxUniChar(ch), bFromEnd); }
1634 int Find(unsigned char ch, bool bFromEnd = false) const
1635 { return Find(wxUniChar(ch), bFromEnd); }
1636 int Find(wchar_t ch, bool bFromEnd = false) const
1637 { return Find(wxUniChar(ch), bFromEnd); }
1638 // searching (return starting index, or -1 if not found)
1639 int Find(const wxString& sub) const // like strstr
1640 {
1641 size_type idx = find(sub);
1642 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1643 }
1644 int Find(const char *sub) const // like strstr
1645 {
1646 size_type idx = find(sub);
1647 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1648 }
1649 int Find(const wchar_t *sub) const // like strstr
1650 {
1651 size_type idx = find(sub);
1652 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1653 }
1654
1655 int Find(const wxCStrData& sub) const
1656 { return Find(sub.AsString()); }
1657 int Find(const wxCharBuffer& sub) const
1658 { return Find(sub.data()); }
1659 int Find(const wxWCharBuffer& sub) const
1660 { return Find(sub.data()); }
1661
1662 // replace first (or all of bReplaceAll) occurences of substring with
1663 // another string, returns the number of replacements made
1664 size_t Replace(const wxString& strOld,
1665 const wxString& strNew,
1666 bool bReplaceAll = true);
1667
1668 // check if the string contents matches a mask containing '*' and '?'
1669 bool Matches(const wxString& mask) const;
1670
1671 // conversion to numbers: all functions return true only if the whole
1672 // string is a number and put the value of this number into the pointer
1673 // provided, the base is the numeric base in which the conversion should be
1674 // done and must be comprised between 2 and 36 or be 0 in which case the
1675 // standard C rules apply (leading '0' => octal, "0x" => hex)
1676 // convert to a signed integer
1677 bool ToLong(long *val, int base = 10) const;
1678 // convert to an unsigned integer
1679 bool ToULong(unsigned long *val, int base = 10) const;
1680 // convert to wxLongLong
1681 #if defined(wxLongLong_t)
1682 bool ToLongLong(wxLongLong_t *val, int base = 10) const;
1683 // convert to wxULongLong
1684 bool ToULongLong(wxULongLong_t *val, int base = 10) const;
1685 #endif // wxLongLong_t
1686 // convert to a double
1687 bool ToDouble(double *val) const;
1688
1689
1690 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1691 // formatted input/output
1692 // as sprintf(), returns the number of characters written or < 0 on error
1693 // (take 'this' into account in attribute parameter count)
1694 // int Printf(const wxString& format, ...);
1695 WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&),
1696 DoPrintfWchar, DoPrintfUtf8)
1697 #ifdef __WATCOMC__
1698 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1699 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxString&),
1700 (wxFormatString(f1)));
1701 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxCStrData&),
1702 (wxFormatString(f1)));
1703 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const char*),
1704 (wxFormatString(f1)));
1705 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wchar_t*),
1706 (wxFormatString(f1)));
1707 #endif
1708 #endif // !wxNEEDS_WXSTRING_PRINTF_MIXIN
1709 // as vprintf(), returns the number of characters written or < 0 on error
1710 int PrintfV(const wxString& format, va_list argptr);
1711
1712 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1713 // returns the string containing the result of Printf() to it
1714 // static wxString Format(const wxString& format, ...) ATTRIBUTE_PRINTF_1;
1715 WX_DEFINE_VARARG_FUNC(static wxString, Format, 1, (const wxFormatString&),
1716 DoFormatWchar, DoFormatUtf8)
1717 #ifdef __WATCOMC__
1718 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1719 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxString&),
1720 (wxFormatString(f1)));
1721 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxCStrData&),
1722 (wxFormatString(f1)));
1723 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const char*),
1724 (wxFormatString(f1)));
1725 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wchar_t*),
1726 (wxFormatString(f1)));
1727 #endif
1728 #endif
1729 // the same as above, but takes a va_list
1730 static wxString FormatV(const wxString& format, va_list argptr);
1731
1732 // raw access to string memory
1733 // ensure that string has space for at least nLen characters
1734 // only works if the data of this string is not shared
1735 bool Alloc(size_t nLen) { reserve(nLen); return capacity() >= nLen; }
1736 // minimize the string's memory
1737 // only works if the data of this string is not shared
1738 bool Shrink();
1739 #if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
1740 // These are deprecated, use wxStringBuffer or wxStringBufferLength instead
1741 //
1742 // get writable buffer of at least nLen bytes. Unget() *must* be called
1743 // a.s.a.p. to put string back in a reasonable state!
1744 wxDEPRECATED( wxStringCharType *GetWriteBuf(size_t nLen) );
1745 // call this immediately after GetWriteBuf() has been used
1746 wxDEPRECATED( void UngetWriteBuf() );
1747 wxDEPRECATED( void UngetWriteBuf(size_t nLen) );
1748 #endif // WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && wxUSE_UNICODE_UTF8
1749
1750 // wxWidgets version 1 compatibility functions
1751
1752 // use Mid()
1753 wxString SubString(size_t from, size_t to) const
1754 { return Mid(from, (to - from + 1)); }
1755 // values for second parameter of CompareTo function
1756 enum caseCompare {exact, ignoreCase};
1757 // values for first parameter of Strip function
1758 enum stripType {leading = 0x1, trailing = 0x2, both = 0x3};
1759
1760 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1761 // use Printf()
1762 // (take 'this' into account in attribute parameter count)
1763 // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2;
1764 WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&),
1765 DoPrintfWchar, DoPrintfUtf8)
1766 #ifdef __WATCOMC__
1767 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1768 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxString&),
1769 (wxFormatString(f1)));
1770 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxCStrData&),
1771 (wxFormatString(f1)));
1772 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const char*),
1773 (wxFormatString(f1)));
1774 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wchar_t*),
1775 (wxFormatString(f1)));
1776 #endif
1777 #endif // wxNEEDS_WXSTRING_PRINTF_MIXIN
1778
1779 // use Cmp()
1780 inline int CompareTo(const wxChar* psz, caseCompare cmp = exact) const
1781 { return cmp == exact ? Cmp(psz) : CmpNoCase(psz); }
1782
1783 // use length()
1784 size_t Length() const { return length(); }
1785 // Count the number of characters
1786 int Freq(wxUniChar ch) const;
1787 // use MakeLower
1788 void LowerCase() { MakeLower(); }
1789 // use MakeUpper
1790 void UpperCase() { MakeUpper(); }
1791 // use Trim except that it doesn't change this string
1792 wxString Strip(stripType w = trailing) const;
1793
1794 // use Find (more general variants not yet supported)
1795 size_t Index(const wxChar* psz) const { return Find(psz); }
1796 size_t Index(wxUniChar ch) const { return Find(ch); }
1797 // use Truncate
1798 wxString& Remove(size_t pos) { return Truncate(pos); }
1799 wxString& RemoveLast(size_t n = 1) { return Truncate(length() - n); }
1800
1801 wxString& Remove(size_t nStart, size_t nLen)
1802 { return (wxString&)erase( nStart, nLen ); }
1803
1804 // use Find()
1805 int First( wxUniChar ch ) const { return Find(ch); }
1806 int First( wxUniCharRef ch ) const { return Find(ch); }
1807 int First( char ch ) const { return Find(ch); }
1808 int First( unsigned char ch ) const { return Find(ch); }
1809 int First( wchar_t ch ) const { return Find(ch); }
1810 int First( const wxString& str ) const { return Find(str); }
1811 int Last( wxUniChar ch ) const { return Find(ch, true); }
1812 bool Contains(const wxString& str) const { return Find(str) != wxNOT_FOUND; }
1813
1814 // use empty()
1815 bool IsNull() const { return empty(); }
1816
1817 // std::string compatibility functions
1818
1819 // take nLen chars starting at nPos
1820 wxString(const wxString& str, size_t nPos, size_t nLen)
1821 { assign(str, nPos, nLen); }
1822 // take all characters from first to last
1823 wxString(const_iterator first, const_iterator last)
1824 : m_impl(first.impl(), last.impl()) { }
1825 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1826 // the 2 overloads below are for compatibility with the existing code using
1827 // pointers instead of iterators
1828 wxString(const char *first, const char *last)
1829 {
1830 SubstrBufFromMB str(ImplStr(first, last - first));
1831 m_impl.assign(str.data, str.len);
1832 }
1833 wxString(const wchar_t *first, const wchar_t *last)
1834 {
1835 SubstrBufFromWC str(ImplStr(first, last - first));
1836 m_impl.assign(str.data, str.len);
1837 }
1838 // and this one is needed to compile code adding offsets to c_str() result
1839 wxString(const wxCStrData& first, const wxCStrData& last)
1840 : m_impl(CreateConstIterator(first).impl(),
1841 CreateConstIterator(last).impl())
1842 {
1843 wxASSERT_MSG( first.m_str == last.m_str,
1844 _T("pointers must be into the same string") );
1845 }
1846 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1847
1848 // lib.string.modifiers
1849 // append elements str[pos], ..., str[pos+n]
1850 wxString& append(const wxString& str, size_t pos, size_t n)
1851 {
1852 size_t from, len;
1853 str.PosLenToImpl(pos, n, &from, &len);
1854 m_impl.append(str.m_impl, from, len);
1855 return *this;
1856 }
1857 // append a string
1858 wxString& append(const wxString& str)
1859 { m_impl.append(str.m_impl); return *this; }
1860 // append first n (or all if n == npos) characters of sz
1861 wxString& append(const char *sz)
1862 { m_impl.append(ImplStr(sz)); return *this; }
1863 wxString& append(const wchar_t *sz)
1864 { m_impl.append(ImplStr(sz)); return *this; }
1865 wxString& append(const char *sz, size_t n)
1866 {
1867 SubstrBufFromMB str(ImplStr(sz, n));
1868 m_impl.append(str.data, str.len);
1869 return *this;
1870 }
1871 wxString& append(const wchar_t *sz, size_t n)
1872 {
1873 SubstrBufFromWC str(ImplStr(sz, n));
1874 m_impl.append(str.data, str.len);
1875 return *this;
1876 }
1877
1878 wxString& append(const wxCStrData& str)
1879 { return append(str.AsString()); }
1880 wxString& append(const wxCharBuffer& str)
1881 { return append(str.data()); }
1882 wxString& append(const wxWCharBuffer& str)
1883 { return append(str.data()); }
1884
1885 // append n copies of ch
1886 wxString& append(size_t n, wxUniChar ch)
1887 {
1888 #if wxUSE_UNICODE_UTF8
1889 if ( !ch.IsAscii() )
1890 m_impl.append(wxStringOperations::EncodeNChars(n, ch));
1891 else
1892 #endif
1893 m_impl.append(n, (wxStringCharType)ch);
1894 return *this;
1895 }
1896 // append from first to last
1897 wxString& append(const_iterator first, const_iterator last)
1898 { m_impl.append(first.impl(), last.impl()); return *this; }
1899 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1900 wxString& append(const char *first, const char *last)
1901 { return append(first, last - first); }
1902 wxString& append(const wchar_t *first, const wchar_t *last)
1903 { return append(first, last - first); }
1904 wxString& append(const wxCStrData& first, const wxCStrData& last)
1905 { return append(CreateConstIterator(first), CreateConstIterator(last)); }
1906 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1907
1908 // same as `this_string = str'
1909 wxString& assign(const wxString& str)
1910 { m_impl = str.m_impl; return *this; }
1911 wxString& assign(const wxString& str, size_t len)
1912 {
1913 m_impl.assign(str.m_impl, 0, str.LenToImpl(len));
1914 return *this;
1915 }
1916 // same as ` = str[pos..pos + n]
1917 wxString& assign(const wxString& str, size_t pos, size_t n)
1918 {
1919 size_t from, len;
1920 str.PosLenToImpl(pos, n, &from, &len);
1921 m_impl.assign(str.m_impl, from, len);
1922 return *this;
1923 }
1924 // same as `= first n (or all if n == npos) characters of sz'
1925 wxString& assign(const char *sz)
1926 { m_impl.assign(ImplStr(sz)); return *this; }
1927 wxString& assign(const wchar_t *sz)
1928 { m_impl.assign(ImplStr(sz)); return *this; }
1929 wxString& assign(const char *sz, size_t n)
1930 {
1931 SubstrBufFromMB str(ImplStr(sz, n));
1932 m_impl.assign(str.data, str.len);
1933 return *this;
1934 }
1935 wxString& assign(const wchar_t *sz, size_t n)
1936 {
1937 SubstrBufFromWC str(ImplStr(sz, n));
1938 m_impl.assign(str.data, str.len);
1939 return *this;
1940 }
1941
1942 wxString& assign(const wxCStrData& str)
1943 { return assign(str.AsString()); }
1944 wxString& assign(const wxCharBuffer& str)
1945 { return assign(str.data()); }
1946 wxString& assign(const wxWCharBuffer& str)
1947 { return assign(str.data()); }
1948 wxString& assign(const wxCStrData& str, size_t len)
1949 { return assign(str.AsString(), len); }
1950 wxString& assign(const wxCharBuffer& str, size_t len)
1951 { return assign(str.data(), len); }
1952 wxString& assign(const wxWCharBuffer& str, size_t len)
1953 { return assign(str.data(), len); }
1954
1955 // same as `= n copies of ch'
1956 wxString& assign(size_t n, wxUniChar ch)
1957 {
1958 #if wxUSE_UNICODE_UTF8
1959 if ( !ch.IsAscii() )
1960 m_impl.assign(wxStringOperations::EncodeNChars(n, ch));
1961 else
1962 #endif
1963 m_impl.assign(n, (wxStringCharType)ch);
1964 return *this;
1965 }
1966
1967 wxString& assign(size_t n, wxUniCharRef ch)
1968 { return assign(n, wxUniChar(ch)); }
1969 wxString& assign(size_t n, char ch)
1970 { return assign(n, wxUniChar(ch)); }
1971 wxString& assign(size_t n, unsigned char ch)
1972 { return assign(n, wxUniChar(ch)); }
1973 wxString& assign(size_t n, wchar_t ch)
1974 { return assign(n, wxUniChar(ch)); }
1975
1976 // assign from first to last
1977 wxString& assign(const_iterator first, const_iterator last)
1978 { m_impl.assign(first.impl(), last.impl()); return *this; }
1979 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1980 wxString& assign(const char *first, const char *last)
1981 { return assign(first, last - first); }
1982 wxString& assign(const wchar_t *first, const wchar_t *last)
1983 { return assign(first, last - first); }
1984 wxString& assign(const wxCStrData& first, const wxCStrData& last)
1985 { return assign(CreateConstIterator(first), CreateConstIterator(last)); }
1986 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1987
1988 // string comparison
1989 int compare(const wxString& str) const;
1990 int compare(const char* sz) const;
1991 int compare(const wchar_t* sz) const;
1992 int compare(const wxCStrData& str) const
1993 { return compare(str.AsString()); }
1994 int compare(const wxCharBuffer& str) const
1995 { return compare(str.data()); }
1996 int compare(const wxWCharBuffer& str) const
1997 { return compare(str.data()); }
1998 // comparison with a substring
1999 int compare(size_t nStart, size_t nLen, const wxString& str) const;
2000 // comparison of 2 substrings
2001 int compare(size_t nStart, size_t nLen,
2002 const wxString& str, size_t nStart2, size_t nLen2) const;
2003 // substring comparison with first nCount characters of sz
2004 int compare(size_t nStart, size_t nLen,
2005 const char* sz, size_t nCount = npos) const;
2006 int compare(size_t nStart, size_t nLen,
2007 const wchar_t* sz, size_t nCount = npos) const;
2008
2009 // insert another string
2010 wxString& insert(size_t nPos, const wxString& str)
2011 { insert(begin() + nPos, str.begin(), str.end()); return *this; }
2012 // insert n chars of str starting at nStart (in str)
2013 wxString& insert(size_t nPos, const wxString& str, size_t nStart, size_t n)
2014 {
2015 size_t from, len;
2016 str.PosLenToImpl(nStart, n, &from, &len);
2017 m_impl.insert(PosToImpl(nPos), str.m_impl, from, len);
2018 return *this;
2019 }
2020 // insert first n (or all if n == npos) characters of sz
2021 wxString& insert(size_t nPos, const char *sz)
2022 { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; }
2023 wxString& insert(size_t nPos, const wchar_t *sz)
2024 { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; }
2025 wxString& insert(size_t nPos, const char *sz, size_t n)
2026 {
2027 SubstrBufFromMB str(ImplStr(sz, n));
2028 m_impl.insert(PosToImpl(nPos), str.data, str.len);
2029 return *this;
2030 }
2031 wxString& insert(size_t nPos, const wchar_t *sz, size_t n)
2032 {
2033 SubstrBufFromWC str(ImplStr(sz, n));
2034 m_impl.insert(PosToImpl(nPos), str.data, str.len);
2035 return *this;
2036 }
2037 // insert n copies of ch
2038 wxString& insert(size_t nPos, size_t n, wxUniChar ch)
2039 {
2040 #if wxUSE_UNICODE_UTF8
2041 if ( !ch.IsAscii() )
2042 m_impl.insert(PosToImpl(nPos), wxStringOperations::EncodeNChars(n, ch));
2043 else
2044 #endif
2045 m_impl.insert(PosToImpl(nPos), n, (wxStringCharType)ch);
2046 return *this;
2047 }
2048 iterator insert(iterator it, wxUniChar ch)
2049 {
2050 #if wxUSE_UNICODE_UTF8
2051 if ( !ch.IsAscii() )
2052 {
2053 size_t pos = IterToImplPos(it);
2054 m_impl.insert(pos, wxStringOperations::EncodeChar(ch));
2055 return iterator(this, m_impl.begin() + pos);
2056 }
2057 else
2058 #endif
2059 return iterator(this, m_impl.insert(it.impl(), (wxStringCharType)ch));
2060 }
2061 void insert(iterator it, const_iterator first, const_iterator last)
2062 { m_impl.insert(it.impl(), first.impl(), last.impl()); }
2063 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2064 void insert(iterator it, const char *first, const char *last)
2065 { insert(it - begin(), first, last - first); }
2066 void insert(iterator it, const wchar_t *first, const wchar_t *last)
2067 { insert(it - begin(), first, last - first); }
2068 void insert(iterator it, const wxCStrData& first, const wxCStrData& last)
2069 { insert(it, CreateConstIterator(first), CreateConstIterator(last)); }
2070 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2071
2072 void insert(iterator it, size_type n, wxUniChar ch)
2073 {
2074 #if wxUSE_UNICODE_UTF8
2075 if ( !ch.IsAscii() )
2076 m_impl.insert(IterToImplPos(it), wxStringOperations::EncodeNChars(n, ch));
2077 else
2078 #endif
2079 m_impl.insert(it.impl(), n, (wxStringCharType)ch);
2080 }
2081
2082 // delete characters from nStart to nStart + nLen
2083 wxString& erase(size_type pos = 0, size_type n = npos)
2084 {
2085 size_t from, len;
2086 PosLenToImpl(pos, n, &from, &len);
2087 m_impl.erase(from, len);
2088 return *this;
2089 }
2090 // delete characters from first up to last
2091 iterator erase(iterator first, iterator last)
2092 { return iterator(this, m_impl.erase(first.impl(), last.impl())); }
2093 iterator erase(iterator first)
2094 { return iterator(this, m_impl.erase(first.impl())); }
2095
2096 #ifdef wxSTRING_BASE_HASNT_CLEAR
2097 void clear() { erase(); }
2098 #else
2099 void clear() { m_impl.clear(); }
2100 #endif
2101
2102 // replaces the substring of length nLen starting at nStart
2103 wxString& replace(size_t nStart, size_t nLen, const char* sz)
2104 {
2105 size_t from, len;
2106 PosLenToImpl(nStart, nLen, &from, &len);
2107 m_impl.replace(from, len, ImplStr(sz));
2108 return *this;
2109 }
2110 wxString& replace(size_t nStart, size_t nLen, const wchar_t* sz)
2111 {
2112 size_t from, len;
2113 PosLenToImpl(nStart, nLen, &from, &len);
2114 m_impl.replace(from, len, ImplStr(sz));
2115 return *this;
2116 }
2117 // replaces the substring of length nLen starting at nStart
2118 wxString& replace(size_t nStart, size_t nLen, const wxString& str)
2119 {
2120 size_t from, len;
2121 PosLenToImpl(nStart, nLen, &from, &len);
2122 m_impl.replace(from, len, str.m_impl);
2123 return *this;
2124 }
2125 // replaces the substring with nCount copies of ch
2126 wxString& replace(size_t nStart, size_t nLen, size_t nCount, wxUniChar ch)
2127 {
2128 size_t from, len;
2129 PosLenToImpl(nStart, nLen, &from, &len);
2130 #if wxUSE_UNICODE_UTF8
2131 if ( !ch.IsAscii() )
2132 m_impl.replace(from, len, wxStringOperations::EncodeNChars(nCount, ch));
2133 else
2134 #endif
2135 m_impl.replace(from, len, nCount, (wxStringCharType)ch);
2136 return *this;
2137 }
2138 // replaces a substring with another substring
2139 wxString& replace(size_t nStart, size_t nLen,
2140 const wxString& str, size_t nStart2, size_t nLen2)
2141 {
2142 size_t from, len;
2143 PosLenToImpl(nStart, nLen, &from, &len);
2144
2145 size_t from2, len2;
2146 str.PosLenToImpl(nStart2, nLen2, &from2, &len2);
2147
2148 m_impl.replace(from, len, str.m_impl, from2, len2);
2149 return *this;
2150 }
2151 // replaces the substring with first nCount chars of sz
2152 wxString& replace(size_t nStart, size_t nLen,
2153 const char* sz, size_t nCount)
2154 {
2155 size_t from, len;
2156 PosLenToImpl(nStart, nLen, &from, &len);
2157
2158 SubstrBufFromMB str(ImplStr(sz, nCount));
2159
2160 m_impl.replace(from, len, str.data, str.len);
2161 return *this;
2162 }
2163 wxString& replace(size_t nStart, size_t nLen,
2164 const wchar_t* sz, size_t nCount)
2165 {
2166 size_t from, len;
2167 PosLenToImpl(nStart, nLen, &from, &len);
2168
2169 SubstrBufFromWC str(ImplStr(sz, nCount));
2170
2171 m_impl.replace(from, len, str.data, str.len);
2172 return *this;
2173 }
2174 wxString& replace(size_t nStart, size_t nLen,
2175 const wxString& s, size_t nCount)
2176 {
2177 size_t from, len;
2178 PosLenToImpl(nStart, nLen, &from, &len);
2179 m_impl.replace(from, len, s.m_impl.c_str(), s.LenToImpl(nCount));
2180 return *this;
2181 }
2182
2183 wxString& replace(iterator first, iterator last, const char* s)
2184 { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; }
2185 wxString& replace(iterator first, iterator last, const wchar_t* s)
2186 { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; }
2187 wxString& replace(iterator first, iterator last, const char* s, size_type n)
2188 {
2189 SubstrBufFromMB str(ImplStr(s, n));
2190 m_impl.replace(first.impl(), last.impl(), str.data, str.len);
2191 return *this;
2192 }
2193 wxString& replace(iterator first, iterator last, const wchar_t* s, size_type n)
2194 {
2195 SubstrBufFromWC str(ImplStr(s, n));
2196 m_impl.replace(first.impl(), last.impl(), str.data, str.len);
2197 return *this;
2198 }
2199 wxString& replace(iterator first, iterator last, const wxString& s)
2200 { m_impl.replace(first.impl(), last.impl(), s.m_impl); return *this; }
2201 wxString& replace(iterator first, iterator last, size_type n, wxUniChar ch)
2202 {
2203 #if wxUSE_UNICODE_UTF8
2204 if ( !ch.IsAscii() )
2205 m_impl.replace(first.impl(), last.impl(),
2206 wxStringOperations::EncodeNChars(n, ch));
2207 else
2208 #endif
2209 m_impl.replace(first.impl(), last.impl(), n, (wxStringCharType)ch);
2210 return *this;
2211 }
2212 wxString& replace(iterator first, iterator last,
2213 const_iterator first1, const_iterator last1)
2214 {
2215 m_impl.replace(first.impl(), last.impl(), first1.impl(), last1.impl());
2216 return *this;
2217 }
2218 wxString& replace(iterator first, iterator last,
2219 const char *first1, const char *last1)
2220 { replace(first, last, first1, last1 - first1); return *this; }
2221 wxString& replace(iterator first, iterator last,
2222 const wchar_t *first1, const wchar_t *last1)
2223 { replace(first, last, first1, last1 - first1); return *this; }
2224
2225 // swap two strings
2226 void swap(wxString& str)
2227 { m_impl.swap(str.m_impl); }
2228
2229 // find a substring
2230 size_t find(const wxString& str, size_t nStart = 0) const
2231 { return PosFromImpl(m_impl.find(str.m_impl, PosToImpl(nStart))); }
2232
2233 // find first n characters of sz
2234 size_t find(const char* sz, size_t nStart = 0, size_t n = npos) const
2235 {
2236 SubstrBufFromMB str(ImplStr(sz, n));
2237 return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len));
2238 }
2239 size_t find(const wchar_t* sz, size_t nStart = 0, size_t n = npos) const
2240 {
2241 SubstrBufFromWC str(ImplStr(sz, n));
2242 return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len));
2243 }
2244 size_t find(const wxCharBuffer& s, size_t nStart = 0, size_t n = npos) const
2245 { return find(s.data(), nStart, n); }
2246 size_t find(const wxWCharBuffer& s, size_t nStart = 0, size_t n = npos) const
2247 { return find(s.data(), nStart, n); }
2248 size_t find(const wxCStrData& s, size_t nStart = 0, size_t n = npos) const
2249 { return find(s.AsWChar(), nStart, n); }
2250
2251 // find the first occurence of character ch after nStart
2252 size_t find(wxUniChar ch, size_t nStart = 0) const
2253 {
2254 #if wxUSE_UNICODE_UTF8
2255 if ( !ch.IsAscii() )
2256 return PosFromImpl(m_impl.find(wxStringOperations::EncodeChar(ch),
2257 PosToImpl(nStart)));
2258 else
2259 #endif
2260 return PosFromImpl(m_impl.find((wxStringCharType)ch,
2261 PosToImpl(nStart)));
2262
2263 }
2264 size_t find(wxUniCharRef ch, size_t nStart = 0) const
2265 { return find(wxUniChar(ch), nStart); }
2266 size_t find(char ch, size_t nStart = 0) const
2267 { return find(wxUniChar(ch), nStart); }
2268 size_t find(unsigned char ch, size_t nStart = 0) const
2269 { return find(wxUniChar(ch), nStart); }
2270 size_t find(wchar_t ch, size_t nStart = 0) const
2271 { return find(wxUniChar(ch), nStart); }
2272
2273 // rfind() family is exactly like find() but works right to left
2274
2275 // as find, but from the end
2276 size_t rfind(const wxString& str, size_t nStart = npos) const
2277 { return PosFromImpl(m_impl.rfind(str.m_impl, PosToImpl(nStart))); }
2278
2279 // as find, but from the end
2280 size_t rfind(const char* sz, size_t nStart = npos, size_t n = npos) const
2281 {
2282 SubstrBufFromMB str(ImplStr(sz, n));
2283 return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len));
2284 }
2285 size_t rfind(const wchar_t* sz, size_t nStart = npos, size_t n = npos) const
2286 {
2287 SubstrBufFromWC str(ImplStr(sz, n));
2288 return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len));
2289 }
2290 size_t rfind(const wxCharBuffer& s, size_t nStart = npos, size_t n = npos) const
2291 { return rfind(s.data(), nStart, n); }
2292 size_t rfind(const wxWCharBuffer& s, size_t nStart = npos, size_t n = npos) const
2293 { return rfind(s.data(), nStart, n); }
2294 size_t rfind(const wxCStrData& s, size_t nStart = npos, size_t n = npos) const
2295 { return rfind(s.AsWChar(), nStart, n); }
2296 // as find, but from the end
2297 size_t rfind(wxUniChar ch, size_t nStart = npos) const
2298 {
2299 #if wxUSE_UNICODE_UTF8
2300 if ( !ch.IsAscii() )
2301 return PosFromImpl(m_impl.rfind(wxStringOperations::EncodeChar(ch),
2302 PosToImpl(nStart)));
2303 else
2304 #endif
2305 return PosFromImpl(m_impl.rfind((wxStringCharType)ch,
2306 PosToImpl(nStart)));
2307 }
2308 size_t rfind(wxUniCharRef ch, size_t nStart = npos) const
2309 { return rfind(wxUniChar(ch), nStart); }
2310 size_t rfind(char ch, size_t nStart = npos) const
2311 { return rfind(wxUniChar(ch), nStart); }
2312 size_t rfind(unsigned char ch, size_t nStart = npos) const
2313 { return rfind(wxUniChar(ch), nStart); }
2314 size_t rfind(wchar_t ch, size_t nStart = npos) const
2315 { return rfind(wxUniChar(ch), nStart); }
2316
2317 // find first/last occurence of any character (not) in the set:
2318 #if wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2319 // FIXME-UTF8: this is not entirely correct, because it doesn't work if
2320 // sizeof(wchar_t)==2 and surrogates are present in the string;
2321 // should we care? Probably not.
2322 size_t find_first_of(const wxString& str, size_t nStart = 0) const
2323 { return m_impl.find_first_of(str.m_impl, nStart); }
2324 size_t find_first_of(const char* sz, size_t nStart = 0) const
2325 { return m_impl.find_first_of(ImplStr(sz), nStart); }
2326 size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const
2327 { return m_impl.find_first_of(ImplStr(sz), nStart); }
2328 size_t find_first_of(const char* sz, size_t nStart, size_t n) const
2329 { return m_impl.find_first_of(ImplStr(sz), nStart, n); }
2330 size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const
2331 { return m_impl.find_first_of(ImplStr(sz), nStart, n); }
2332 size_t find_first_of(wxUniChar c, size_t nStart = 0) const
2333 { return m_impl.find_first_of((wxChar)c, nStart); }
2334
2335 size_t find_last_of(const wxString& str, size_t nStart = npos) const
2336 { return m_impl.find_last_of(str.m_impl, nStart); }
2337 size_t find_last_of(const char* sz, size_t nStart = npos) const
2338 { return m_impl.find_last_of(ImplStr(sz), nStart); }
2339 size_t find_last_of(const wchar_t* sz, size_t nStart = npos) const
2340 { return m_impl.find_last_of(ImplStr(sz), nStart); }
2341 size_t find_last_of(const char* sz, size_t nStart, size_t n) const
2342 { return m_impl.find_last_of(ImplStr(sz), nStart, n); }
2343 size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const
2344 { return m_impl.find_last_of(ImplStr(sz), nStart, n); }
2345 size_t find_last_of(wxUniChar c, size_t nStart = npos) const
2346 { return m_impl.find_last_of((wxChar)c, nStart); }
2347
2348 size_t find_first_not_of(const wxString& str, size_t nStart = 0) const
2349 { return m_impl.find_first_not_of(str.m_impl, nStart); }
2350 size_t find_first_not_of(const char* sz, size_t nStart = 0) const
2351 { return m_impl.find_first_not_of(ImplStr(sz), nStart); }
2352 size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const
2353 { return m_impl.find_first_not_of(ImplStr(sz), nStart); }
2354 size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const
2355 { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); }
2356 size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const
2357 { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); }
2358 size_t find_first_not_of(wxUniChar c, size_t nStart = 0) const
2359 { return m_impl.find_first_not_of((wxChar)c, nStart); }
2360
2361 size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
2362 { return m_impl.find_last_not_of(str.m_impl, nStart); }
2363 size_t find_last_not_of(const char* sz, size_t nStart = npos) const
2364 { return m_impl.find_last_not_of(ImplStr(sz), nStart); }
2365 size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const
2366 { return m_impl.find_last_not_of(ImplStr(sz), nStart); }
2367 size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const
2368 { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); }
2369 size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const
2370 { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); }
2371 size_t find_last_not_of(wxUniChar c, size_t nStart = npos) const
2372 { return m_impl.find_last_not_of((wxChar)c, nStart); }
2373 #else
2374 // we can't use std::string implementation in UTF-8 build, because the
2375 // character sets would be interpreted wrongly:
2376
2377 // as strpbrk() but starts at nStart, returns npos if not found
2378 size_t find_first_of(const wxString& str, size_t nStart = 0) const
2379 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2380 { return find_first_of(str.wc_str(), nStart); }
2381 #else
2382 { return find_first_of(str.mb_str(), nStart); }
2383 #endif
2384 // same as above
2385 size_t find_first_of(const char* sz, size_t nStart = 0) const;
2386 size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const;
2387 size_t find_first_of(const char* sz, size_t nStart, size_t n) const;
2388 size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const;
2389 // same as find(char, size_t)
2390 size_t find_first_of(wxUniChar c, size_t nStart = 0) const
2391 { return find(c, nStart); }
2392 // find the last (starting from nStart) char from str in this string
2393 size_t find_last_of (const wxString& str, size_t nStart = npos) const
2394 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2395 { return find_last_of(str.wc_str(), nStart); }
2396 #else
2397 { return find_last_of(str.mb_str(), nStart); }
2398 #endif
2399 // same as above
2400 size_t find_last_of (const char* sz, size_t nStart = npos) const;
2401 size_t find_last_of (const wchar_t* sz, size_t nStart = npos) const;
2402 size_t find_last_of(const char* sz, size_t nStart, size_t n) const;
2403 size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const;
2404 // same as above
2405 size_t find_last_of(wxUniChar c, size_t nStart = npos) const
2406 { return rfind(c, nStart); }
2407
2408 // find first/last occurence of any character not in the set
2409
2410 // as strspn() (starting from nStart), returns npos on failure
2411 size_t find_first_not_of(const wxString& str, size_t nStart = 0) const
2412 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2413 { return find_first_not_of(str.wc_str(), nStart); }
2414 #else
2415 { return find_first_not_of(str.mb_str(), nStart); }
2416 #endif
2417 // same as above
2418 size_t find_first_not_of(const char* sz, size_t nStart = 0) const;
2419 size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const;
2420 size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const;
2421 size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const;
2422 // same as above
2423 size_t find_first_not_of(wxUniChar ch, size_t nStart = 0) const;
2424 // as strcspn()
2425 size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
2426 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2427 { return find_last_not_of(str.wc_str(), nStart); }
2428 #else
2429 { return find_last_not_of(str.mb_str(), nStart); }
2430 #endif
2431 // same as above
2432 size_t find_last_not_of(const char* sz, size_t nStart = npos) const;
2433 size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const;
2434 size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const;
2435 size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const;
2436 // same as above
2437 size_t find_last_not_of(wxUniChar ch, size_t nStart = npos) const;
2438 #endif // wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 or not
2439
2440 // provide char/wchar_t/wxUniCharRef overloads for char-finding functions
2441 // above to resolve ambiguities:
2442 size_t find_first_of(wxUniCharRef ch, size_t nStart = 0) const
2443 { return find_first_of(wxUniChar(ch), nStart); }
2444 size_t find_first_of(char ch, size_t nStart = 0) const
2445 { return find_first_of(wxUniChar(ch), nStart); }
2446 size_t find_first_of(unsigned char ch, size_t nStart = 0) const
2447 { return find_first_of(wxUniChar(ch), nStart); }
2448 size_t find_first_of(wchar_t ch, size_t nStart = 0) const
2449 { return find_first_of(wxUniChar(ch), nStart); }
2450 size_t find_last_of(wxUniCharRef ch, size_t nStart = npos) const
2451 { return find_last_of(wxUniChar(ch), nStart); }
2452 size_t find_last_of(char ch, size_t nStart = npos) const
2453 { return find_last_of(wxUniChar(ch), nStart); }
2454 size_t find_last_of(unsigned char ch, size_t nStart = npos) const
2455 { return find_last_of(wxUniChar(ch), nStart); }
2456 size_t find_last_of(wchar_t ch, size_t nStart = npos) const
2457 { return find_last_of(wxUniChar(ch), nStart); }
2458 size_t find_first_not_of(wxUniCharRef ch, size_t nStart = 0) const
2459 { return find_first_not_of(wxUniChar(ch), nStart); }
2460 size_t find_first_not_of(char ch, size_t nStart = 0) const
2461 { return find_first_not_of(wxUniChar(ch), nStart); }
2462 size_t find_first_not_of(unsigned char ch, size_t nStart = 0) const
2463 { return find_first_not_of(wxUniChar(ch), nStart); }
2464 size_t find_first_not_of(wchar_t ch, size_t nStart = 0) const
2465 { return find_first_not_of(wxUniChar(ch), nStart); }
2466 size_t find_last_not_of(wxUniCharRef ch, size_t nStart = npos) const
2467 { return find_last_not_of(wxUniChar(ch), nStart); }
2468 size_t find_last_not_of(char ch, size_t nStart = npos) const
2469 { return find_last_not_of(wxUniChar(ch), nStart); }
2470 size_t find_last_not_of(unsigned char ch, size_t nStart = npos) const
2471 { return find_last_not_of(wxUniChar(ch), nStart); }
2472 size_t find_last_not_of(wchar_t ch, size_t nStart = npos) const
2473 { return find_last_not_of(wxUniChar(ch), nStart); }
2474
2475 // and additional overloads for the versions taking strings:
2476 size_t find_first_of(const wxCStrData& sz, size_t nStart = 0) const
2477 { return find_first_of(sz.AsString(), nStart); }
2478 size_t find_first_of(const wxCharBuffer& sz, size_t nStart = 0) const
2479 { return find_first_of(sz.data(), nStart); }
2480 size_t find_first_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2481 { return find_first_of(sz.data(), nStart); }
2482 size_t find_first_of(const wxCStrData& sz, size_t nStart, size_t n) const
2483 { return find_first_of(sz.AsWChar(), nStart, n); }
2484 size_t find_first_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2485 { return find_first_of(sz.data(), nStart, n); }
2486 size_t find_first_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2487 { return find_first_of(sz.data(), nStart, n); }
2488
2489 size_t find_last_of(const wxCStrData& sz, size_t nStart = 0) const
2490 { return find_last_of(sz.AsString(), nStart); }
2491 size_t find_last_of(const wxCharBuffer& sz, size_t nStart = 0) const
2492 { return find_last_of(sz.data(), nStart); }
2493 size_t find_last_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2494 { return find_last_of(sz.data(), nStart); }
2495 size_t find_last_of(const wxCStrData& sz, size_t nStart, size_t n) const
2496 { return find_last_of(sz.AsWChar(), nStart, n); }
2497 size_t find_last_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2498 { return find_last_of(sz.data(), nStart, n); }
2499 size_t find_last_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2500 { return find_last_of(sz.data(), nStart, n); }
2501
2502 size_t find_first_not_of(const wxCStrData& sz, size_t nStart = 0) const
2503 { return find_first_not_of(sz.AsString(), nStart); }
2504 size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart = 0) const
2505 { return find_first_not_of(sz.data(), nStart); }
2506 size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2507 { return find_first_not_of(sz.data(), nStart); }
2508 size_t find_first_not_of(const wxCStrData& sz, size_t nStart, size_t n) const
2509 { return find_first_not_of(sz.AsWChar(), nStart, n); }
2510 size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2511 { return find_first_not_of(sz.data(), nStart, n); }
2512 size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2513 { return find_first_not_of(sz.data(), nStart, n); }
2514
2515 size_t find_last_not_of(const wxCStrData& sz, size_t nStart = 0) const
2516 { return find_last_not_of(sz.AsString(), nStart); }
2517 size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart = 0) const
2518 { return find_last_not_of(sz.data(), nStart); }
2519 size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2520 { return find_last_not_of(sz.data(), nStart); }
2521 size_t find_last_not_of(const wxCStrData& sz, size_t nStart, size_t n) const
2522 { return find_last_not_of(sz.AsWChar(), nStart, n); }
2523 size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2524 { return find_last_not_of(sz.data(), nStart, n); }
2525 size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2526 { return find_last_not_of(sz.data(), nStart, n); }
2527
2528 // string += string
2529 wxString& operator+=(const wxString& s)
2530 { m_impl += s.m_impl; return *this; }
2531 // string += C string
2532 wxString& operator+=(const char *psz)
2533 { m_impl += ImplStr(psz); return *this; }
2534 wxString& operator+=(const wchar_t *pwz)
2535 { m_impl += ImplStr(pwz); return *this; }
2536 wxString& operator+=(const wxCStrData& s)
2537 { m_impl += s.AsString().m_impl; return *this; }
2538 wxString& operator+=(const wxCharBuffer& s)
2539 { return operator+=(s.data()); }
2540 wxString& operator+=(const wxWCharBuffer& s)
2541 { return operator+=(s.data()); }
2542 // string += char
2543 wxString& operator+=(wxUniChar ch)
2544 {
2545 #if wxUSE_UNICODE_UTF8
2546 if ( !ch.IsAscii() )
2547 m_impl += wxStringOperations::EncodeChar(ch);
2548 else
2549 #endif
2550 m_impl += (wxStringCharType)ch;
2551 return *this;
2552 }
2553 wxString& operator+=(wxUniCharRef ch) { return *this += wxUniChar(ch); }
2554 wxString& operator+=(int ch) { return *this += wxUniChar(ch); }
2555 wxString& operator+=(char ch) { return *this += wxUniChar(ch); }
2556 wxString& operator+=(unsigned char ch) { return *this += wxUniChar(ch); }
2557 wxString& operator+=(wchar_t ch) { return *this += wxUniChar(ch); }
2558
2559 private:
2560 #if !wxUSE_STL_BASED_WXSTRING
2561 // helpers for wxStringBuffer and wxStringBufferLength
2562 wxStringCharType *DoGetWriteBuf(size_t nLen)
2563 { return m_impl.DoGetWriteBuf(nLen); }
2564 void DoUngetWriteBuf()
2565 { m_impl.DoUngetWriteBuf(); }
2566 void DoUngetWriteBuf(size_t nLen)
2567 { m_impl.DoUngetWriteBuf(nLen); }
2568 #endif // !wxUSE_STL_BASED_WXSTRING
2569
2570 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
2571 #if !wxUSE_UTF8_LOCALE_ONLY
2572 int DoPrintfWchar(const wxChar *format, ...);
2573 static wxString DoFormatWchar(const wxChar *format, ...);
2574 #endif
2575 #if wxUSE_UNICODE_UTF8
2576 int DoPrintfUtf8(const char *format, ...);
2577 static wxString DoFormatUtf8(const char *format, ...);
2578 #endif
2579 #endif
2580
2581 #if !wxUSE_STL_BASED_WXSTRING
2582 // check string's data validity
2583 bool IsValid() const { return m_impl.GetStringData()->IsValid(); }
2584 #endif
2585
2586 private:
2587 wxStringImpl m_impl;
2588
2589 // buffers for compatibility conversion from (char*)c_str() and
2590 // (wchar_t*)c_str():
2591 // FIXME-UTF8: bechmark various approaches to keeping compatibility buffers
2592 template<typename T>
2593 struct ConvertedBuffer
2594 {
2595 ConvertedBuffer() : m_buf(NULL) {}
2596 ~ConvertedBuffer()
2597 { free(m_buf); }
2598
2599 operator T*() const { return m_buf; }
2600
2601 ConvertedBuffer& operator=(T *str)
2602 {
2603 free(m_buf);
2604 m_buf = str;
2605 return *this;
2606 }
2607
2608 T *m_buf;
2609 };
2610 #if wxUSE_UNICODE && !wxUSE_UTF8_LOCALE_ONLY
2611 ConvertedBuffer<char> m_convertedToChar;
2612 #endif
2613 #if !wxUSE_UNICODE_WCHAR
2614 ConvertedBuffer<wchar_t> m_convertedToWChar;
2615 #endif
2616
2617 #if wxUSE_UNICODE_UTF8
2618 // FIXME-UTF8: (try to) move this elsewhere (TLS) or solve differently
2619 // assigning to character pointer to by wxString::interator may
2620 // change the underlying wxStringImpl iterator, so we have to
2621 // keep track of all iterators and update them as necessary:
2622 struct wxStringIteratorNodeHead
2623 {
2624 wxStringIteratorNodeHead() : ptr(NULL) {}
2625 wxStringIteratorNode *ptr;
2626
2627 // copying is disallowed as it would result in more than one pointer into
2628 // the same linked list
2629 DECLARE_NO_COPY_CLASS(wxStringIteratorNodeHead)
2630 };
2631
2632 wxStringIteratorNodeHead m_iterators;
2633
2634 friend class WXDLLIMPEXP_FWD_BASE wxStringIteratorNode;
2635 friend class WXDLLIMPEXP_FWD_BASE wxUniCharRef;
2636 #endif // wxUSE_UNICODE_UTF8
2637
2638 friend class WXDLLIMPEXP_FWD_BASE wxCStrData;
2639 friend class wxStringInternalBuffer;
2640 friend class wxStringInternalBufferLength;
2641 };
2642
2643 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
2644 #pragma warning (default:4275)
2645 #endif
2646
2647 // string iterator operators that satisfy STL Random Access Iterator
2648 // requirements:
2649 inline wxString::iterator operator+(ptrdiff_t n, wxString::iterator i)
2650 { return i + n; }
2651 inline wxString::const_iterator operator+(ptrdiff_t n, wxString::const_iterator i)
2652 { return i + n; }
2653 inline wxString::reverse_iterator operator+(ptrdiff_t n, wxString::reverse_iterator i)
2654 { return i + n; }
2655 inline wxString::const_reverse_iterator operator+(ptrdiff_t n, wxString::const_reverse_iterator i)
2656 { return i + n; }
2657
2658 // notice that even though for many compilers the friend declarations above are
2659 // enough, from the point of view of C++ standard we must have the declarations
2660 // here as friend ones are not injected in the enclosing namespace and without
2661 // them the code fails to compile with conforming compilers such as xlC or g++4
2662 wxString WXDLLIMPEXP_BASE operator+(const wxString& string1, const wxString& string2);
2663 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const char *psz);
2664 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const wchar_t *pwz);
2665 wxString WXDLLIMPEXP_BASE operator+(const char *psz, const wxString& string);
2666 wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz, const wxString& string);
2667
2668 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxUniChar ch);
2669 wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string);
2670
2671 inline wxString operator+(const wxString& string, wxUniCharRef ch)
2672 { return string + (wxUniChar)ch; }
2673 inline wxString operator+(const wxString& string, char ch)
2674 { return string + wxUniChar(ch); }
2675 inline wxString operator+(const wxString& string, wchar_t ch)
2676 { return string + wxUniChar(ch); }
2677 inline wxString operator+(wxUniCharRef ch, const wxString& string)
2678 { return (wxUniChar)ch + string; }
2679 inline wxString operator+(char ch, const wxString& string)
2680 { return wxUniChar(ch) + string; }
2681 inline wxString operator+(wchar_t ch, const wxString& string)
2682 { return wxUniChar(ch) + string; }
2683
2684
2685 #define wxGetEmptyString() wxString()
2686
2687 // ----------------------------------------------------------------------------
2688 // wxStringBuffer: a tiny class allowing to get a writable pointer into string
2689 // ----------------------------------------------------------------------------
2690
2691 #if !wxUSE_STL_BASED_WXSTRING
2692 // string buffer for direct access to string data in their native
2693 // representation:
2694 class wxStringInternalBuffer
2695 {
2696 public:
2697 typedef wxStringCharType CharType;
2698
2699 wxStringInternalBuffer(wxString& str, size_t lenWanted = 1024)
2700 : m_str(str), m_buf(NULL)
2701 { m_buf = m_str.DoGetWriteBuf(lenWanted); }
2702
2703 ~wxStringInternalBuffer() { m_str.DoUngetWriteBuf(); }
2704
2705 operator wxStringCharType*() const { return m_buf; }
2706
2707 private:
2708 wxString& m_str;
2709 wxStringCharType *m_buf;
2710
2711 DECLARE_NO_COPY_CLASS(wxStringInternalBuffer)
2712 };
2713
2714 class wxStringInternalBufferLength
2715 {
2716 public:
2717 typedef wxStringCharType CharType;
2718
2719 wxStringInternalBufferLength(wxString& str, size_t lenWanted = 1024)
2720 : m_str(str), m_buf(NULL), m_len(0), m_lenSet(false)
2721 {
2722 m_buf = m_str.DoGetWriteBuf(lenWanted);
2723 wxASSERT(m_buf != NULL);
2724 }
2725
2726 ~wxStringInternalBufferLength()
2727 {
2728 wxASSERT(m_lenSet);
2729 m_str.DoUngetWriteBuf(m_len);
2730 }
2731
2732 operator wxStringCharType*() const { return m_buf; }
2733 void SetLength(size_t length) { m_len = length; m_lenSet = true; }
2734
2735 private:
2736 wxString& m_str;
2737 wxStringCharType *m_buf;
2738 size_t m_len;
2739 bool m_lenSet;
2740
2741 DECLARE_NO_COPY_CLASS(wxStringInternalBufferLength)
2742 };
2743
2744 #endif // !wxUSE_STL_BASED_WXSTRING
2745
2746 template<typename T>
2747 class WXDLLIMPEXP_BASE wxStringTypeBufferBase
2748 {
2749 public:
2750 typedef T CharType;
2751
2752 wxStringTypeBufferBase(wxString& str, size_t lenWanted = 1024)
2753 : m_str(str), m_buf(lenWanted)
2754 { }
2755
2756
2757 operator CharType*() { return m_buf.data(); }
2758
2759 protected:
2760 wxString& m_str;
2761 wxCharTypeBuffer<CharType> m_buf;
2762 };
2763
2764 template<typename T>
2765 class WXDLLIMPEXP_BASE wxStringTypeBufferLengthBase
2766 {
2767 public:
2768 typedef T CharType;
2769
2770 wxStringTypeBufferLengthBase(wxString& str, size_t lenWanted = 1024)
2771 : m_str(str), m_buf(lenWanted), m_len(0), m_lenSet(false)
2772 { }
2773
2774 operator CharType*() { return m_buf.data(); }
2775 void SetLength(size_t length) { m_len = length; m_lenSet = true; }
2776
2777 protected:
2778 wxString& m_str;
2779 wxCharTypeBuffer<CharType> m_buf;
2780 size_t m_len;
2781 bool m_lenSet;
2782 };
2783
2784 template<typename T>
2785 class wxStringTypeBuffer : public wxStringTypeBufferBase<T>
2786 {
2787 public:
2788 wxStringTypeBuffer(wxString& str, size_t lenWanted = 1024)
2789 : wxStringTypeBufferBase<T>(str, lenWanted) {}
2790 ~wxStringTypeBuffer()
2791 {
2792 this->m_str.assign(this->m_buf.data());
2793 }
2794
2795 DECLARE_NO_COPY_CLASS(wxStringTypeBuffer)
2796 };
2797
2798 template<typename T>
2799 class wxStringTypeBufferLength : public wxStringTypeBufferLengthBase<T>
2800 {
2801 public:
2802 wxStringTypeBufferLength(wxString& str, size_t lenWanted = 1024)
2803 : wxStringTypeBufferLengthBase<T>(str, lenWanted) {}
2804
2805 ~wxStringTypeBufferLength()
2806 {
2807 wxASSERT(this->m_lenSet);
2808 this->m_str.assign(this->m_buf.data(), this->m_len);
2809 }
2810
2811 DECLARE_NO_COPY_CLASS(wxStringTypeBufferLength)
2812 };
2813
2814 #if wxUSE_STL_BASED_WXSTRING
2815
2816 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferBase<wxStringCharType> )
2817
2818 class wxStringInternalBuffer : public wxStringTypeBufferBase<wxStringCharType>
2819 {
2820 public:
2821 wxStringInternalBuffer(wxString& str, size_t lenWanted = 1024)
2822 : wxStringTypeBufferBase<wxStringCharType>(str, lenWanted) {}
2823 ~wxStringInternalBuffer()
2824 { m_str.m_impl.assign(m_buf.data()); }
2825
2826 DECLARE_NO_COPY_CLASS(wxStringInternalBuffer)
2827 };
2828
2829 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE(
2830 wxStringTypeBufferLengthBase<wxStringCharType> )
2831
2832 class wxStringInternalBufferLength
2833 : public wxStringTypeBufferLengthBase<wxStringCharType>
2834 {
2835 public:
2836 wxStringInternalBufferLength(wxString& str, size_t lenWanted = 1024)
2837 : wxStringTypeBufferLengthBase<wxStringCharType>(str, lenWanted) {}
2838
2839 ~wxStringInternalBufferLength()
2840 {
2841 wxASSERT(m_lenSet);
2842 m_str.m_impl.assign(m_buf.data(), m_len);
2843 }
2844
2845 DECLARE_NO_COPY_CLASS(wxStringInternalBufferLength)
2846 };
2847 #endif // wxUSE_STL_BASED_WXSTRING
2848
2849
2850 #if wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8
2851 typedef wxStringTypeBuffer<wxChar> wxStringBuffer;
2852 typedef wxStringTypeBufferLength<wxChar> wxStringBufferLength;
2853 #else // if !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2854 typedef wxStringInternalBuffer wxStringBuffer;
2855 typedef wxStringInternalBufferLength wxStringBufferLength;
2856 #endif // !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2857
2858 #if wxUSE_UNICODE_UTF8
2859 typedef wxStringInternalBuffer wxUTF8StringBuffer;
2860 typedef wxStringInternalBufferLength wxUTF8StringBufferLength;
2861 #elif wxUSE_UNICODE_WCHAR
2862
2863 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferBase<char> )
2864
2865 class WXDLLIMPEXP_BASE wxUTF8StringBuffer : public wxStringTypeBufferBase<char>
2866 {
2867 public:
2868 wxUTF8StringBuffer(wxString& str, size_t lenWanted = 1024)
2869 : wxStringTypeBufferBase<char>(str, lenWanted) {}
2870 ~wxUTF8StringBuffer();
2871
2872 DECLARE_NO_COPY_CLASS(wxUTF8StringBuffer)
2873 };
2874
2875 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferLengthBase<char> )
2876
2877 class WXDLLIMPEXP_BASE wxUTF8StringBufferLength
2878 : public wxStringTypeBufferLengthBase<char>
2879 {
2880 public:
2881 wxUTF8StringBufferLength(wxString& str, size_t lenWanted = 1024)
2882 : wxStringTypeBufferLengthBase<char>(str, lenWanted) {}
2883 ~wxUTF8StringBufferLength();
2884
2885 DECLARE_NO_COPY_CLASS(wxUTF8StringBufferLength)
2886 };
2887 #endif // wxUSE_UNICODE_UTF8/wxUSE_UNICODE_WCHAR
2888
2889
2890 // ---------------------------------------------------------------------------
2891 // wxString comparison functions: operator versions are always case sensitive
2892 // ---------------------------------------------------------------------------
2893
2894 #define wxCMP_WXCHAR_STRING(p, s, op) 0 op s.Cmp(p)
2895
2896 wxDEFINE_ALL_COMPARISONS(const wxChar *, const wxString&, wxCMP_WXCHAR_STRING)
2897
2898 #undef wxCMP_WXCHAR_STRING
2899
2900 inline bool operator==(const wxString& s1, const wxString& s2)
2901 { return s1.IsSameAs(s2); }
2902 inline bool operator!=(const wxString& s1, const wxString& s2)
2903 { return !s1.IsSameAs(s2); }
2904 inline bool operator< (const wxString& s1, const wxString& s2)
2905 { return s1.Cmp(s2) < 0; }
2906 inline bool operator> (const wxString& s1, const wxString& s2)
2907 { return s1.Cmp(s2) > 0; }
2908 inline bool operator<=(const wxString& s1, const wxString& s2)
2909 { return s1.Cmp(s2) <= 0; }
2910 inline bool operator>=(const wxString& s1, const wxString& s2)
2911 { return s1.Cmp(s2) >= 0; }
2912
2913 inline bool operator==(const wxString& s1, const wxCStrData& s2)
2914 { return s1 == s2.AsString(); }
2915 inline bool operator==(const wxCStrData& s1, const wxString& s2)
2916 { return s1.AsString() == s2; }
2917 inline bool operator!=(const wxString& s1, const wxCStrData& s2)
2918 { return s1 != s2.AsString(); }
2919 inline bool operator!=(const wxCStrData& s1, const wxString& s2)
2920 { return s1.AsString() != s2; }
2921
2922 inline bool operator==(const wxString& s1, const wxWCharBuffer& s2)
2923 { return (s1.Cmp((const wchar_t *)s2) == 0); }
2924 inline bool operator==(const wxWCharBuffer& s1, const wxString& s2)
2925 { return (s2.Cmp((const wchar_t *)s1) == 0); }
2926 inline bool operator!=(const wxString& s1, const wxWCharBuffer& s2)
2927 { return (s1.Cmp((const wchar_t *)s2) != 0); }
2928 inline bool operator!=(const wxWCharBuffer& s1, const wxString& s2)
2929 { return (s2.Cmp((const wchar_t *)s1) != 0); }
2930
2931 inline bool operator==(const wxString& s1, const wxCharBuffer& s2)
2932 { return (s1.Cmp((const char *)s2) == 0); }
2933 inline bool operator==(const wxCharBuffer& s1, const wxString& s2)
2934 { return (s2.Cmp((const char *)s1) == 0); }
2935 inline bool operator!=(const wxString& s1, const wxCharBuffer& s2)
2936 { return (s1.Cmp((const char *)s2) != 0); }
2937 inline bool operator!=(const wxCharBuffer& s1, const wxString& s2)
2938 { return (s2.Cmp((const char *)s1) != 0); }
2939
2940 inline wxString operator+(const wxString& string, const wxWCharBuffer& buf)
2941 { return string + (const wchar_t *)buf; }
2942 inline wxString operator+(const wxWCharBuffer& buf, const wxString& string)
2943 { return (const wchar_t *)buf + string; }
2944
2945 inline wxString operator+(const wxString& string, const wxCharBuffer& buf)
2946 { return string + (const char *)buf; }
2947 inline wxString operator+(const wxCharBuffer& buf, const wxString& string)
2948 { return (const char *)buf + string; }
2949
2950 // comparison with char
2951 inline bool operator==(const wxUniChar& c, const wxString& s) { return s.IsSameAs(c); }
2952 inline bool operator==(const wxUniCharRef& c, const wxString& s) { return s.IsSameAs(c); }
2953 inline bool operator==(char c, const wxString& s) { return s.IsSameAs(c); }
2954 inline bool operator==(wchar_t c, const wxString& s) { return s.IsSameAs(c); }
2955 inline bool operator==(int c, const wxString& s) { return s.IsSameAs(c); }
2956 inline bool operator==(const wxString& s, const wxUniChar& c) { return s.IsSameAs(c); }
2957 inline bool operator==(const wxString& s, const wxUniCharRef& c) { return s.IsSameAs(c); }
2958 inline bool operator==(const wxString& s, char c) { return s.IsSameAs(c); }
2959 inline bool operator==(const wxString& s, wchar_t c) { return s.IsSameAs(c); }
2960 inline bool operator!=(const wxUniChar& c, const wxString& s) { return !s.IsSameAs(c); }
2961 inline bool operator!=(const wxUniCharRef& c, const wxString& s) { return !s.IsSameAs(c); }
2962 inline bool operator!=(char c, const wxString& s) { return !s.IsSameAs(c); }
2963 inline bool operator!=(wchar_t c, const wxString& s) { return !s.IsSameAs(c); }
2964 inline bool operator!=(int c, const wxString& s) { return !s.IsSameAs(c); }
2965 inline bool operator!=(const wxString& s, const wxUniChar& c) { return !s.IsSameAs(c); }
2966 inline bool operator!=(const wxString& s, const wxUniCharRef& c) { return !s.IsSameAs(c); }
2967 inline bool operator!=(const wxString& s, char c) { return !s.IsSameAs(c); }
2968 inline bool operator!=(const wxString& s, wchar_t c) { return !s.IsSameAs(c); }
2969
2970 // comparison with C string in Unicode build
2971 #if wxUSE_UNICODE
2972
2973 #define wxCMP_CHAR_STRING(p, s, op) wxString(p) op s
2974
2975 wxDEFINE_ALL_COMPARISONS(const char *, const wxString&, wxCMP_CHAR_STRING)
2976
2977 #undef wxCMP_CHAR_STRING
2978
2979 #endif // wxUSE_UNICODE
2980
2981 // we also need to provide the operators for comparison with wxCStrData to
2982 // resolve ambiguity between operator(const wxChar *,const wxString &) and
2983 // operator(const wxChar *, const wxChar *) for "p == s.c_str()"
2984 //
2985 // notice that these are (shallow) pointer comparisons, not (deep) string ones
2986 #define wxCMP_CHAR_CSTRDATA(p, s, op) p op s.AsChar()
2987 #define wxCMP_WCHAR_CSTRDATA(p, s, op) p op s.AsWChar()
2988
2989 wxDEFINE_ALL_COMPARISONS(const wchar_t *, const wxCStrData&, wxCMP_WCHAR_CSTRDATA)
2990 wxDEFINE_ALL_COMPARISONS(const char *, const wxCStrData&, wxCMP_CHAR_CSTRDATA)
2991
2992 #undef wxCMP_CHAR_CSTRDATA
2993 #undef wxCMP_WCHAR_CSTRDATA
2994
2995 // ---------------------------------------------------------------------------
2996 // Implementation only from here until the end of file
2997 // ---------------------------------------------------------------------------
2998
2999 #if wxUSE_STD_IOSTREAM
3000
3001 #include "wx/iosfwrap.h"
3002
3003 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxString&);
3004 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCStrData&);
3005 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCharBuffer&);
3006 #ifndef __BORLANDC__
3007 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxWCharBuffer&);
3008 #endif
3009
3010 #if wxUSE_UNICODE && defined(HAVE_WOSTREAM)
3011
3012 WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxString&);
3013 WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxCStrData&);
3014 WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxWCharBuffer&);
3015
3016 #endif // wxUSE_UNICODE && defined(HAVE_WOSTREAM)
3017
3018 #endif // wxUSE_STD_IOSTREAM
3019
3020 // ---------------------------------------------------------------------------
3021 // wxCStrData implementation
3022 // ---------------------------------------------------------------------------
3023
3024 inline wxCStrData::wxCStrData(char *buf)
3025 : m_str(new wxString(buf)), m_offset(0), m_owned(true) {}
3026 inline wxCStrData::wxCStrData(wchar_t *buf)
3027 : m_str(new wxString(buf)), m_offset(0), m_owned(true) {}
3028
3029 inline wxCStrData::wxCStrData(const wxCStrData& data)
3030 : m_str(data.m_owned ? new wxString(*data.m_str) : data.m_str),
3031 m_offset(data.m_offset),
3032 m_owned(data.m_owned)
3033 {
3034 }
3035
3036 inline wxCStrData::~wxCStrData()
3037 {
3038 if ( m_owned )
3039 delete wx_const_cast(wxString*, m_str); // cast to silence warnings
3040 }
3041
3042 // simple cases for AsChar() and AsWChar(), the complicated ones are
3043 // in string.cpp
3044 #if wxUSE_UNICODE_WCHAR
3045 inline const wchar_t* wxCStrData::AsWChar() const
3046 {
3047 return m_str->wx_str() + m_offset;
3048 }
3049 #endif // wxUSE_UNICODE_WCHAR
3050
3051 #if !wxUSE_UNICODE
3052 inline const char* wxCStrData::AsChar() const
3053 {
3054 return m_str->wx_str() + m_offset;
3055 }
3056 #endif // !wxUSE_UNICODE
3057
3058 #if wxUSE_UTF8_LOCALE_ONLY
3059 inline const char* wxCStrData::AsChar() const
3060 {
3061 return wxStringOperations::AddToIter(m_str->wx_str(), m_offset);
3062 }
3063 #endif // wxUSE_UTF8_LOCALE_ONLY
3064
3065 inline const wxCharBuffer wxCStrData::AsCharBuf() const
3066 {
3067 #if !wxUSE_UNICODE
3068 return wxCharBuffer::CreateNonOwned(AsChar());
3069 #else
3070 return AsString().mb_str();
3071 #endif
3072 }
3073
3074 inline const wxWCharBuffer wxCStrData::AsWCharBuf() const
3075 {
3076 #if wxUSE_UNICODE_WCHAR
3077 return wxWCharBuffer::CreateNonOwned(AsWChar());
3078 #else
3079 return AsString().wc_str();
3080 #endif
3081 }
3082
3083 inline wxString wxCStrData::AsString() const
3084 {
3085 if ( m_offset == 0 )
3086 return *m_str;
3087 else
3088 return m_str->Mid(m_offset);
3089 }
3090
3091 inline const wxStringCharType *wxCStrData::AsInternal() const
3092 {
3093 #if wxUSE_UNICODE_UTF8
3094 return wxStringOperations::AddToIter(m_str->wx_str(), m_offset);
3095 #else
3096 return m_str->wx_str() + m_offset;
3097 #endif
3098 }
3099
3100 inline wxUniChar wxCStrData::operator*() const
3101 {
3102 if ( m_str->empty() )
3103 return wxUniChar(_T('\0'));
3104 else
3105 return (*m_str)[m_offset];
3106 }
3107
3108 inline wxUniChar wxCStrData::operator[](size_t n) const
3109 {
3110 // NB: we intentionally use operator[] and not at() here because the former
3111 // works for the terminating NUL while the latter does not
3112 return (*m_str)[m_offset + n];
3113 }
3114
3115 // ----------------------------------------------------------------------------
3116 // more wxCStrData operators
3117 // ----------------------------------------------------------------------------
3118
3119 // we need to define those to allow "size_t pos = p - s.c_str()" where p is
3120 // some pointer into the string
3121 inline size_t operator-(const char *p, const wxCStrData& cs)
3122 {
3123 return p - cs.AsChar();
3124 }
3125
3126 inline size_t operator-(const wchar_t *p, const wxCStrData& cs)
3127 {
3128 return p - cs.AsWChar();
3129 }
3130
3131 // ----------------------------------------------------------------------------
3132 // implementation of wx[W]CharBuffer inline methods using wxCStrData
3133 // ----------------------------------------------------------------------------
3134
3135 // FIXME-UTF8: move this to buffer.h
3136 inline wxCharBuffer::wxCharBuffer(const wxCStrData& cstr)
3137 : wxCharTypeBufferBase(cstr.AsCharBuf())
3138 {
3139 }
3140
3141 inline wxWCharBuffer::wxWCharBuffer(const wxCStrData& cstr)
3142 : wxCharTypeBufferBase(cstr.AsWCharBuf())
3143 {
3144 }
3145
3146 #if wxUSE_UNICODE_UTF8
3147 // ----------------------------------------------------------------------------
3148 // implementation of wxStringIteratorNode inline methods
3149 // ----------------------------------------------------------------------------
3150
3151 void wxStringIteratorNode::DoSet(const wxString *str,
3152 wxStringImpl::const_iterator *citer,
3153 wxStringImpl::iterator *iter)
3154 {
3155 m_prev = NULL;
3156 m_iter = iter;
3157 m_citer = citer;
3158 m_str = str;
3159 if ( str )
3160 {
3161 m_next = str->m_iterators.ptr;
3162 wx_const_cast(wxString*, m_str)->m_iterators.ptr = this;
3163 if ( m_next )
3164 m_next->m_prev = this;
3165 }
3166 else
3167 {
3168 m_next = NULL;
3169 }
3170 }
3171
3172 void wxStringIteratorNode::clear()
3173 {
3174 if ( m_next )
3175 m_next->m_prev = m_prev;
3176 if ( m_prev )
3177 m_prev->m_next = m_next;
3178 else if ( m_str ) // first in the list
3179 wx_const_cast(wxString*, m_str)->m_iterators.ptr = m_next;
3180
3181 m_next = m_prev = NULL;
3182 m_citer = NULL;
3183 m_iter = NULL;
3184 m_str = NULL;
3185 }
3186 #endif // wxUSE_UNICODE_UTF8
3187
3188 #if WXWIN_COMPATIBILITY_2_8
3189 // lot of code out there doesn't explicitly include wx/crt.h, but uses
3190 // CRT wrappers that are now declared in wx/wxcrt.h and wx/wxcrtvararg.h,
3191 // so let's include this header now that wxString is defined and it's safe
3192 // to do it:
3193 #include "wx/crt.h"
3194 #endif
3195
3196 #endif // _WX_WXSTRING_H_