]> git.saurik.com Git - wxWidgets.git/blame_incremental - include/wx/string.h
set eol-style
[wxWidgets.git] / include / wx / string.h
... / ...
CommitLineData
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#if defined(__WXMAC__) || defined(__VISAGECPP__)
28 #include <ctype.h>
29#endif
30
31#if defined(__VISAGECPP__) && __IBMCPP__ >= 400
32 // problem in VACPP V4 with including stdlib.h multiple times
33 // strconv includes it anyway
34# include <stdio.h>
35# include <string.h>
36# include <stdarg.h>
37# include <limits.h>
38#else
39# include <string.h>
40# include <stdio.h>
41# include <stdarg.h>
42# include <limits.h>
43# include <stdlib.h>
44#endif
45
46#ifdef HAVE_STRCASECMP_IN_STRINGS_H
47 #include <strings.h> // for strcasecmp()
48#endif // HAVE_STRCASECMP_IN_STRINGS_H
49
50#ifdef __WXPALMOS__
51 #include <StringMgr.h>
52#endif
53
54#include "wx/wxcrtbase.h" // for wxChar, wxStrlen() etc.
55#include "wx/strvararg.h"
56#include "wx/buffer.h" // for wxCharBuffer
57#include "wx/strconv.h" // for wxConvertXXX() macros and wxMBConv classes
58#include "wx/stringimpl.h"
59#include "wx/stringops.h"
60#include "wx/unichar.h"
61
62class WXDLLIMPEXP_FWD_BASE wxString;
63
64// unless this symbol is predefined to disable the compatibility functions, do
65// use them
66#ifndef WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
67 #define WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER 1
68#endif
69
70// ---------------------------------------------------------------------------
71// macros
72// ---------------------------------------------------------------------------
73
74// casts [unfortunately!] needed to call some broken functions which require
75// "char *" instead of "const char *"
76#define WXSTRINGCAST (wxChar *)(const wxChar *)
77#define wxCSTRINGCAST (wxChar *)(const wxChar *)
78#define wxMBSTRINGCAST (char *)(const char *)
79#define wxWCSTRINGCAST (wchar_t *)(const wchar_t *)
80
81// like _T(), but for literals in wxString's internal representation, i.e.
82// char* in UTF-8 build and wxChar* otherwise:
83#if wxUSE_UNICODE_UTF8
84 #define wxSTRING_TEXT(str) str
85#else
86 #define wxSTRING_TEXT(str) _T(str)
87#endif
88
89// ----------------------------------------------------------------------------
90// constants
91// ----------------------------------------------------------------------------
92
93#if WXWIN_COMPATIBILITY_2_6
94
95// deprecated in favour of wxString::npos, don't use in new code
96//
97// maximum possible length for a string means "take all string" everywhere
98#define wxSTRING_MAXLEN wxString::npos
99
100#endif // WXWIN_COMPATIBILITY_2_6
101
102// ---------------------------------------------------------------------------
103// global functions complementing standard C string library replacements for
104// strlen() and portable strcasecmp()
105//---------------------------------------------------------------------------
106
107#if WXWIN_COMPATIBILITY_2_8
108// Use wxXXX() functions from wxcrt.h instead! These functions are for
109// backwards compatibility only.
110
111// checks whether the passed in pointer is NULL and if the string is empty
112wxDEPRECATED( inline bool IsEmpty(const char *p) );
113inline bool IsEmpty(const char *p) { return (!p || !*p); }
114
115// safe version of strlen() (returns 0 if passed NULL pointer)
116wxDEPRECATED( inline size_t Strlen(const char *psz) );
117inline size_t Strlen(const char *psz)
118 { return psz ? strlen(psz) : 0; }
119
120// portable strcasecmp/_stricmp
121wxDEPRECATED( inline int Stricmp(const char *psz1, const char *psz2) );
122inline int Stricmp(const char *psz1, const char *psz2)
123{
124#if defined(__VISUALC__) && defined(__WXWINCE__)
125 register char c1, c2;
126 do {
127 c1 = tolower(*psz1++);
128 c2 = tolower(*psz2++);
129 } while ( c1 && (c1 == c2) );
130
131 return c1 - c2;
132#elif defined(__VISUALC__) || ( defined(__MWERKS__) && defined(__INTEL__) )
133 return _stricmp(psz1, psz2);
134#elif defined(__SC__)
135 return _stricmp(psz1, psz2);
136#elif defined(__SALFORDC__)
137 return stricmp(psz1, psz2);
138#elif defined(__BORLANDC__)
139 return stricmp(psz1, psz2);
140#elif defined(__WATCOMC__)
141 return stricmp(psz1, psz2);
142#elif defined(__DJGPP__)
143 return stricmp(psz1, psz2);
144#elif defined(__EMX__)
145 return stricmp(psz1, psz2);
146#elif defined(__WXPM__)
147 return stricmp(psz1, psz2);
148#elif defined(__WXPALMOS__) || \
149 defined(HAVE_STRCASECMP_IN_STRING_H) || \
150 defined(HAVE_STRCASECMP_IN_STRINGS_H) || \
151 defined(__GNUWIN32__)
152 return strcasecmp(psz1, psz2);
153#elif defined(__MWERKS__) && !defined(__INTEL__)
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#else
162 // almost all compilers/libraries provide this function (unfortunately under
163 // different names), that's why we don't implement our own which will surely
164 // be more efficient than this code (uncomment to use):
165 /*
166 register char c1, c2;
167 do {
168 c1 = tolower(*psz1++);
169 c2 = tolower(*psz2++);
170 } while ( c1 && (c1 == c2) );
171
172 return c1 - c2;
173 */
174
175 #error "Please define string case-insensitive compare for your OS/compiler"
176#endif // OS/compiler
177}
178
179#endif // WXWIN_COMPATIBILITY_2_8
180
181// ----------------------------------------------------------------------------
182// wxCStrData
183// ----------------------------------------------------------------------------
184
185// Lightweight object returned by wxString::c_str() and implicitly convertible
186// to either const char* or const wchar_t*.
187class WXDLLIMPEXP_BASE wxCStrData
188{
189private:
190 // Ctors; for internal use by wxString and wxCStrData only
191 wxCStrData(const wxString *str, size_t offset = 0, bool owned = false)
192 : m_str(str), m_offset(offset), m_owned(owned) {}
193
194public:
195 // Ctor constructs the object from char literal; they are needed to make
196 // operator?: compile and they intentionally take char*, not const char*
197 inline wxCStrData(char *buf);
198 inline wxCStrData(wchar_t *buf);
199 inline wxCStrData(const wxCStrData& data);
200
201 inline ~wxCStrData();
202
203 // methods defined inline below must be declared inline or mingw32 3.4.5
204 // warns about "<symbol> defined locally after being referenced with
205 // dllimport linkage"
206#if wxUSE_UNICODE_WCHAR
207 inline
208#endif
209 const wchar_t* AsWChar() const;
210 operator const wchar_t*() const { return AsWChar(); }
211
212#if !wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY
213 inline
214#endif
215 const char* AsChar() const;
216 const unsigned char* AsUnsignedChar() const
217 { return (const unsigned char *) AsChar(); }
218 operator const char*() const { return AsChar(); }
219 operator const unsigned char*() const { return AsUnsignedChar(); }
220
221 operator const void*() const { return AsChar(); }
222
223 inline const wxCharBuffer AsCharBuf() const;
224 inline const wxWCharBuffer AsWCharBuf() const;
225
226 inline wxString AsString() const;
227
228 // returns the value as C string in internal representation (equivalent
229 // to AsString().wx_str(), but more efficient)
230 const wxStringCharType *AsInternal() const;
231
232 // allow expressions like "c_str()[0]":
233 inline wxUniChar operator[](size_t n) const;
234 wxUniChar operator[](int n) const { return operator[](size_t(n)); }
235 wxUniChar operator[](long n) const { return operator[](size_t(n)); }
236#ifndef wxSIZE_T_IS_UINT
237 wxUniChar operator[](unsigned int n) const { return operator[](size_t(n)); }
238#endif // size_t != unsigned int
239
240 // these operators are needed to emulate the pointer semantics of c_str():
241 // expressions like "wxChar *p = str.c_str() + 1;" should continue to work
242 // (we need both versions to resolve ambiguities):
243 wxCStrData operator+(int n) const
244 { return wxCStrData(m_str, m_offset + n, m_owned); }
245 wxCStrData operator+(long n) const
246 { return wxCStrData(m_str, m_offset + n, m_owned); }
247 wxCStrData operator+(size_t n) const
248 { return wxCStrData(m_str, m_offset + n, m_owned); }
249
250 // and these for "str.c_str() + n - 2":
251 wxCStrData operator-(int n) const
252 {
253 wxASSERT_MSG( n <= (int)m_offset,
254 _T("attempt to construct address before the beginning of the string") );
255 return wxCStrData(m_str, m_offset - n, m_owned);
256 }
257 wxCStrData operator-(long n) const
258 {
259 wxASSERT_MSG( n <= (int)m_offset,
260 _T("attempt to construct address before the beginning of the string") );
261 return wxCStrData(m_str, m_offset - n, m_owned);
262 }
263 wxCStrData operator-(size_t n) const
264 {
265 wxASSERT_MSG( n <= m_offset,
266 _T("attempt to construct address before the beginning of the string") );
267 return wxCStrData(m_str, m_offset - n, m_owned);
268 }
269
270 // this operator is needed to make expressions like "*c_str()" or
271 // "*(c_str() + 2)" work
272 inline wxUniChar operator*() const;
273
274private:
275 const wxString *m_str;
276 size_t m_offset;
277 bool m_owned;
278
279 friend class WXDLLIMPEXP_FWD_BASE wxString;
280};
281
282// ----------------------------------------------------------------------------
283// wxStringPrintfMixin
284// ---------------------------------------------------------------------------
285
286// NB: VC6 has a bug that causes linker errors if you have template methods
287// in a class using __declspec(dllimport). The solution is to split such
288// class into two classes, one that contains the template methods and does
289// *not* use WXDLLIMPEXP_BASE and another class that contains the rest
290// (with DLL linkage).
291//
292// We only do this for VC6 here, because the code is less efficient
293// (Printf() has to use dynamic_cast<>) and because OpenWatcom compiler
294// cannot compile this code.
295
296#if defined(__VISUALC__) && __VISUALC__ < 1300
297 #define wxNEEDS_WXSTRING_PRINTF_MIXIN
298#endif
299
300#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
301// this class contains implementation of wxString's vararg methods, it's
302// exported from wxBase DLL
303class WXDLLIMPEXP_BASE wxStringPrintfMixinBase
304{
305protected:
306 wxStringPrintfMixinBase() {}
307
308#if !wxUSE_UTF8_LOCALE_ONLY
309 int DoPrintfWchar(const wxChar *format, ...);
310 static wxString DoFormatWchar(const wxChar *format, ...);
311#endif
312#if wxUSE_UNICODE_UTF8
313 int DoPrintfUtf8(const char *format, ...);
314 static wxString DoFormatUtf8(const char *format, ...);
315#endif
316};
317
318// this class contains template wrappers for wxString's vararg methods, it's
319// intentionally *not* exported from the DLL in order to fix the VC6 bug
320// described above
321class wxStringPrintfMixin : public wxStringPrintfMixinBase
322{
323private:
324 // to further complicate things, we can't return wxString from
325 // wxStringPrintfMixin::Format() because wxString is not yet declared at
326 // this point; the solution is to use this fake type trait template - this
327 // way the compiler won't know the return type until Format() is used
328 // (this doesn't compile with Watcom, but VC6 compiles it just fine):
329 template<typename T> struct StringReturnType
330 {
331 typedef wxString type;
332 };
333
334public:
335 // these are duplicated wxString methods, they're also declared below
336 // if !wxNEEDS_WXSTRING_PRINTF_MIXIN:
337
338 // static wxString Format(const wString& format, ...) ATTRIBUTE_PRINTF_1;
339 WX_DEFINE_VARARG_FUNC_SANS_N0(static typename StringReturnType<T1>::type,
340 Format, 1, (const wxFormatString&),
341 DoFormatWchar, DoFormatUtf8)
342 // We have to implement the version without template arguments manually
343 // because of the StringReturnType<> hack, although WX_DEFINE_VARARG_FUNC
344 // normally does it itself. It has to be a template so that we can use
345 // the hack, even though there's no real template parameter:
346 struct FormatDummyArg {} ;
347
348 template<typename T>
349 inline static typename StringReturnType<T>::type
350 Format(const wxFormatString& fmt, FormatDummyArg dummy = FormatDummyArg())
351 {
352 return DoFormatWchar(fmt);
353 }
354
355 // int Printf(const wxString& format, ...);
356 WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&),
357 DoPrintfWchar, DoPrintfUtf8)
358 // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2;
359 WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&),
360 DoPrintfWchar, DoPrintfUtf8)
361
362protected:
363 wxStringPrintfMixin() : wxStringPrintfMixinBase() {}
364};
365#endif // wxNEEDS_WXSTRING_PRINTF_MIXIN
366
367
368// ----------------------------------------------------------------------------
369// wxString: string class trying to be compatible with std::string, MFC
370// CString and wxWindows 1.x wxString all at once
371// ---------------------------------------------------------------------------
372
373#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
374 // "non dll-interface class 'wxStringPrintfMixin' used as base interface
375 // for dll-interface class 'wxString'" -- this is OK in our case
376 #pragma warning (disable:4275)
377#endif
378
379#if wxUSE_UNICODE_UTF8
380// see the comment near wxString::iterator for why we need this
381struct WXDLLIMPEXP_BASE wxStringIteratorNode
382{
383 wxStringIteratorNode()
384 : m_str(NULL), m_citer(NULL), m_iter(NULL), m_prev(NULL), m_next(NULL) {}
385 wxStringIteratorNode(const wxString *str,
386 wxStringImpl::const_iterator *citer)
387 { DoSet(str, citer, NULL); }
388 wxStringIteratorNode(const wxString *str, wxStringImpl::iterator *iter)
389 { DoSet(str, NULL, iter); }
390 ~wxStringIteratorNode()
391 { clear(); }
392
393 inline void set(const wxString *str, wxStringImpl::const_iterator *citer)
394 { clear(); DoSet(str, citer, NULL); }
395 inline void set(const wxString *str, wxStringImpl::iterator *iter)
396 { clear(); DoSet(str, NULL, iter); }
397
398 const wxString *m_str;
399 wxStringImpl::const_iterator *m_citer;
400 wxStringImpl::iterator *m_iter;
401 wxStringIteratorNode *m_prev, *m_next;
402
403private:
404 inline void clear();
405 inline void DoSet(const wxString *str,
406 wxStringImpl::const_iterator *citer,
407 wxStringImpl::iterator *iter);
408
409 // the node belongs to a particular iterator instance, it's not copied
410 // when a copy of the iterator is made
411 DECLARE_NO_COPY_CLASS(wxStringIteratorNode)
412};
413#endif // wxUSE_UNICODE_UTF8
414
415class WXDLLIMPEXP_BASE wxString
416#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
417 : public wxStringPrintfMixin
418#endif
419{
420 // NB: special care was taken in arranging the member functions in such order
421 // that all inline functions can be effectively inlined, verify that all
422 // performance critical functions are still inlined if you change order!
423public:
424 // an 'invalid' value for string index, moved to this place due to a CW bug
425 static const size_t npos;
426
427private:
428 // if we hadn't made these operators private, it would be possible to
429 // compile "wxString s; s = 17;" without any warnings as 17 is implicitly
430 // converted to char in C and we do have operator=(char)
431 //
432 // NB: we don't need other versions (short/long and unsigned) as attempt
433 // to assign another numeric type to wxString will now result in
434 // ambiguity between operator=(char) and operator=(int)
435 wxString& operator=(int);
436
437 // these methods are not implemented - there is _no_ conversion from int to
438 // string, you're doing something wrong if the compiler wants to call it!
439 //
440 // try `s << i' or `s.Printf("%d", i)' instead
441 wxString(int);
442
443
444 // buffer for holding temporary substring when using any of the methods
445 // that take (char*,size_t) or (wchar_t*,size_t) arguments:
446 template<typename T>
447 struct SubstrBufFromType
448 {
449 T data;
450 size_t len;
451
452 SubstrBufFromType(const T& data_, size_t len_)
453 : data(data_), len(len_) {}
454 };
455
456#if wxUSE_UNICODE_UTF8
457 // even char* -> char* needs conversion, from locale charset to UTF-8
458 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromWC;
459 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromMB;
460#elif wxUSE_UNICODE_WCHAR
461 typedef SubstrBufFromType<const wchar_t*> SubstrBufFromWC;
462 typedef SubstrBufFromType<wxWCharBuffer> SubstrBufFromMB;
463#else
464 typedef SubstrBufFromType<const char*> SubstrBufFromMB;
465 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromWC;
466#endif
467
468
469 // Functions implementing primitive operations on string data; wxString
470 // methods and iterators are implemented in terms of it. The differences
471 // between UTF-8 and wchar_t* representations of the string are mostly
472 // contained here.
473
474#if wxUSE_UNICODE_UTF8
475 static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength,
476 const wxMBConv& conv);
477 static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength,
478 const wxMBConv& conv);
479#elif wxUSE_UNICODE_WCHAR
480 static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength,
481 const wxMBConv& conv);
482#else
483 static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength,
484 const wxMBConv& conv);
485#endif
486
487#if !wxUSE_UNICODE_UTF8 // wxUSE_UNICODE_WCHAR or !wxUSE_UNICODE
488 // returns C string encoded as the implementation expects:
489 #if wxUSE_UNICODE
490 static const wchar_t* ImplStr(const wchar_t* str)
491 { return str ? str : wxT(""); }
492 static const SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
493 { return SubstrBufFromWC(str, (str && n == npos) ? wxWcslen(str) : n); }
494 static wxWCharBuffer ImplStr(const char* str,
495 const wxMBConv& conv = wxConvLibc)
496 { return ConvertStr(str, npos, conv).data; }
497 static SubstrBufFromMB ImplStr(const char* str, size_t n,
498 const wxMBConv& conv = wxConvLibc)
499 { return ConvertStr(str, n, conv); }
500 #else
501 static const char* ImplStr(const char* str,
502 const wxMBConv& WXUNUSED(conv) = wxConvLibc)
503 { return str ? str : ""; }
504 static const SubstrBufFromMB ImplStr(const char* str, size_t n,
505 const wxMBConv& WXUNUSED(conv) = wxConvLibc)
506 { return SubstrBufFromMB(str, (str && n == npos) ? wxStrlen(str) : n); }
507 static wxCharBuffer ImplStr(const wchar_t* str)
508 { return ConvertStr(str, npos, wxConvLibc).data; }
509 static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
510 { return ConvertStr(str, n, wxConvLibc); }
511 #endif
512
513 // translates position index in wxString to/from index in underlying
514 // wxStringImpl:
515 static size_t PosToImpl(size_t pos) { return pos; }
516 static void PosLenToImpl(size_t pos, size_t len,
517 size_t *implPos, size_t *implLen)
518 { *implPos = pos; *implLen = len; }
519 static size_t LenToImpl(size_t len) { return len; }
520 static size_t PosFromImpl(size_t pos) { return pos; }
521
522#else // wxUSE_UNICODE_UTF8
523
524 static wxCharBuffer ImplStr(const char* str,
525 const wxMBConv& conv = wxConvLibc)
526 { return ConvertStr(str, npos, conv).data; }
527 static SubstrBufFromMB ImplStr(const char* str, size_t n,
528 const wxMBConv& conv = wxConvLibc)
529 { return ConvertStr(str, n, conv); }
530
531 static wxCharBuffer ImplStr(const wchar_t* str)
532 { return ConvertStr(str, npos, wxMBConvUTF8()).data; }
533 static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
534 { return ConvertStr(str, n, wxMBConvUTF8()); }
535
536 size_t PosToImpl(size_t pos) const
537 {
538 if ( pos == 0 || pos == npos )
539 return pos;
540 else
541 return (begin() + pos).impl() - m_impl.begin();
542 }
543
544 void PosLenToImpl(size_t pos, size_t len, size_t *implPos, size_t *implLen) const;
545
546 size_t LenToImpl(size_t len) const
547 {
548 size_t pos, len2;
549 PosLenToImpl(0, len, &pos, &len2);
550 return len2;
551 }
552
553 size_t PosFromImpl(size_t pos) const
554 {
555 if ( pos == 0 || pos == npos )
556 return pos;
557 else
558 return const_iterator(this, m_impl.begin() + pos) - begin();
559 }
560#endif // !wxUSE_UNICODE_UTF8/wxUSE_UNICODE_UTF8
561
562public:
563 // standard types
564 typedef wxUniChar value_type;
565 typedef wxUniChar char_type;
566 typedef wxUniCharRef reference;
567 typedef wxChar* pointer;
568 typedef const wxChar* const_pointer;
569
570 typedef size_t size_type;
571 typedef wxUniChar const_reference;
572
573#if wxUSE_STL
574 #if wxUSE_UNICODE_UTF8
575 // random access is not O(1), as required by Random Access Iterator
576 #define WX_STR_ITERATOR_TAG std::bidirectional_iterator_tag
577 #else
578 #define WX_STR_ITERATOR_TAG std::random_access_iterator_tag
579 #endif
580#else
581 #define WX_STR_ITERATOR_TAG void /* dummy type */
582#endif
583
584 #define WX_STR_ITERATOR_IMPL(iterator_name, pointer_type, reference_type) \
585 private: \
586 typedef wxStringImpl::iterator_name underlying_iterator; \
587 public: \
588 typedef WX_STR_ITERATOR_TAG iterator_category; \
589 typedef wxUniChar value_type; \
590 typedef int difference_type; \
591 typedef reference_type reference; \
592 typedef pointer_type pointer; \
593 \
594 reference operator[](size_t n) const { return *(*this + n); } \
595 \
596 iterator_name& operator++() \
597 { wxStringOperations::IncIter(m_cur); return *this; } \
598 iterator_name& operator--() \
599 { wxStringOperations::DecIter(m_cur); return *this; } \
600 iterator_name operator++(int) \
601 { \
602 iterator_name tmp = *this; \
603 wxStringOperations::IncIter(m_cur); \
604 return tmp; \
605 } \
606 iterator_name operator--(int) \
607 { \
608 iterator_name tmp = *this; \
609 wxStringOperations::DecIter(m_cur); \
610 return tmp; \
611 } \
612 \
613 iterator_name& operator+=(int n) \
614 { \
615 m_cur = wxStringOperations::AddToIter(m_cur, n); \
616 return *this; \
617 } \
618 iterator_name& operator+=(size_t n) \
619 { \
620 m_cur = wxStringOperations::AddToIter(m_cur, (int)n); \
621 return *this; \
622 } \
623 iterator_name& operator-=(int n) \
624 { \
625 m_cur = wxStringOperations::AddToIter(m_cur, -n); \
626 return *this; \
627 } \
628 iterator_name& operator-=(size_t n) \
629 { \
630 m_cur = wxStringOperations::AddToIter(m_cur, -(int)n); \
631 return *this; \
632 } \
633 \
634 difference_type operator-(const iterator_name& i) const \
635 { return wxStringOperations::DiffIters(m_cur, i.m_cur); } \
636 \
637 bool operator==(const iterator_name& i) const \
638 { return m_cur == i.m_cur; } \
639 bool operator!=(const iterator_name& i) const \
640 { return m_cur != i.m_cur; } \
641 \
642 bool operator<(const iterator_name& i) const \
643 { return m_cur < i.m_cur; } \
644 bool operator>(const iterator_name& i) const \
645 { return m_cur > i.m_cur; } \
646 bool operator<=(const iterator_name& i) const \
647 { return m_cur <= i.m_cur; } \
648 bool operator>=(const iterator_name& i) const \
649 { return m_cur >= i.m_cur; } \
650 \
651 private: \
652 /* for internal wxString use only: */ \
653 underlying_iterator impl() const { return m_cur; } \
654 \
655 friend class wxString; \
656 friend class wxCStrData; \
657 \
658 private: \
659 underlying_iterator m_cur
660
661 class WXDLLIMPEXP_FWD_BASE const_iterator;
662
663#if wxUSE_UNICODE_UTF8
664 // NB: In UTF-8 build, (non-const) iterator needs to keep reference
665 // to the underlying wxStringImpl, because UTF-8 is variable-length
666 // encoding and changing the value pointer to by an iterator (using
667 // its operator*) requires calling wxStringImpl::replace() if the old
668 // and new values differ in their encoding's length.
669 //
670 // Furthermore, the replace() call may invalid all iterators for the
671 // string, so we have to keep track of outstanding iterators and update
672 // them if replace() happens.
673 //
674 // This is implemented by maintaining linked list of iterators for every
675 // string and traversing it in wxUniCharRef::operator=(). Head of the
676 // list is stored in wxString. (FIXME-UTF8)
677
678 class WXDLLIMPEXP_BASE iterator
679 {
680 WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);
681
682 public:
683 iterator() {}
684 iterator(const iterator& i)
685 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
686 iterator& operator=(const iterator& i)
687 { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; }
688
689 reference operator*()
690 { return wxUniCharRef::CreateForString(m_node, m_cur); }
691
692 iterator operator+(int n) const
693 { return iterator(str(), wxStringOperations::AddToIter(m_cur, n)); }
694 iterator operator+(size_t n) const
695 { return iterator(str(), wxStringOperations::AddToIter(m_cur, (int)n)); }
696 iterator operator-(int n) const
697 { return iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); }
698 iterator operator-(size_t n) const
699 { return iterator(str(), wxStringOperations::AddToIter(m_cur, -(int)n)); }
700
701 private:
702 iterator(wxString *str, underlying_iterator ptr)
703 : m_cur(ptr), m_node(str, &m_cur) {}
704
705 wxString* str() const { return wx_const_cast(wxString*, m_node.m_str); }
706
707 wxStringIteratorNode m_node;
708
709 friend class const_iterator;
710 };
711
712 class WXDLLIMPEXP_BASE const_iterator
713 {
714 // NB: reference_type is intentionally value, not reference, the character
715 // may be encoded differently in wxString data:
716 WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar);
717
718 public:
719 const_iterator() {}
720 const_iterator(const const_iterator& i)
721 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
722 const_iterator(const iterator& i)
723 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
724
725 const_iterator& operator=(const const_iterator& i)
726 { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; }
727 const_iterator& operator=(const iterator& i)
728 { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; }
729
730 reference operator*() const
731 { return wxStringOperations::DecodeChar(m_cur); }
732
733 const_iterator operator+(int n) const
734 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, n)); }
735 const_iterator operator+(size_t n) const
736 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, (int)n)); }
737 const_iterator operator-(int n) const
738 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); }
739 const_iterator operator-(size_t n) const
740 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, -(int)n)); }
741
742 private:
743 // for internal wxString use only:
744 const_iterator(const wxString *str, underlying_iterator ptr)
745 : m_cur(ptr), m_node(str, &m_cur) {}
746
747 const wxString* str() const { return m_node.m_str; }
748
749 wxStringIteratorNode m_node;
750 };
751
752 size_t IterToImplPos(wxString::iterator i) const
753 { return wxStringImpl::const_iterator(i.impl()) - m_impl.begin(); }
754
755#else // !wxUSE_UNICODE_UTF8
756
757 class WXDLLIMPEXP_BASE iterator
758 {
759 WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);
760
761 public:
762 iterator() {}
763 iterator(const iterator& i) : m_cur(i.m_cur) {}
764
765 reference operator*()
766 { return wxUniCharRef::CreateForString(m_cur); }
767
768 iterator operator+(int n) const
769 { return iterator(wxStringOperations::AddToIter(m_cur, n)); }
770 iterator operator+(size_t n) const
771 { return iterator(wxStringOperations::AddToIter(m_cur, (int)n)); }
772 iterator operator-(int n) const
773 { return iterator(wxStringOperations::AddToIter(m_cur, -n)); }
774 iterator operator-(size_t n) const
775 { return iterator(wxStringOperations::AddToIter(m_cur, -(int)n)); }
776
777 private:
778 // for internal wxString use only:
779 iterator(underlying_iterator ptr) : m_cur(ptr) {}
780 iterator(wxString *WXUNUSED(str), underlying_iterator ptr) : m_cur(ptr) {}
781
782 friend class const_iterator;
783 };
784
785 class WXDLLIMPEXP_BASE const_iterator
786 {
787 // NB: reference_type is intentionally value, not reference, the character
788 // may be encoded differently in wxString data:
789 WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar);
790
791 public:
792 const_iterator() {}
793 const_iterator(const const_iterator& i) : m_cur(i.m_cur) {}
794 const_iterator(const iterator& i) : m_cur(i.m_cur) {}
795
796 reference operator*() const
797 { return wxStringOperations::DecodeChar(m_cur); }
798
799 const_iterator operator+(int n) const
800 { return const_iterator(wxStringOperations::AddToIter(m_cur, n)); }
801 const_iterator operator+(size_t n) const
802 { return const_iterator(wxStringOperations::AddToIter(m_cur, (int)n)); }
803 const_iterator operator-(int n) const
804 { return const_iterator(wxStringOperations::AddToIter(m_cur, -n)); }
805 const_iterator operator-(size_t n) const
806 { return const_iterator(wxStringOperations::AddToIter(m_cur, -(int)n)); }
807
808 private:
809 // for internal wxString use only:
810 const_iterator(underlying_iterator ptr) : m_cur(ptr) {}
811 const_iterator(const wxString *WXUNUSED(str), underlying_iterator ptr)
812 : m_cur(ptr) {}
813 };
814#endif // wxUSE_UNICODE_UTF8/!wxUSE_UNICODE_UTF8
815
816 #undef WX_STR_ITERATOR_TAG
817 #undef WX_STR_ITERATOR_IMPL
818
819 friend class iterator;
820 friend class const_iterator;
821
822 template <typename T>
823 class reverse_iterator_impl
824 {
825 public:
826 typedef T iterator_type;
827
828 typedef typename T::iterator_category iterator_category;
829 typedef typename T::value_type value_type;
830 typedef typename T::difference_type difference_type;
831 typedef typename T::reference reference;
832 typedef typename T::pointer *pointer;
833
834 reverse_iterator_impl() {}
835 reverse_iterator_impl(iterator_type i) : m_cur(i) {}
836 reverse_iterator_impl(const reverse_iterator_impl& ri)
837 : m_cur(ri.m_cur) {}
838
839 iterator_type base() const { return m_cur; }
840
841 reference operator*() const { return *(m_cur-1); }
842 reference operator[](size_t n) const { return *(*this + n); }
843
844 reverse_iterator_impl& operator++()
845 { --m_cur; return *this; }
846 reverse_iterator_impl operator++(int)
847 { reverse_iterator_impl tmp = *this; --m_cur; return tmp; }
848 reverse_iterator_impl& operator--()
849 { ++m_cur; return *this; }
850 reverse_iterator_impl operator--(int)
851 { reverse_iterator_impl tmp = *this; ++m_cur; return tmp; }
852
853 // NB: explicit <T> in the functions below is to keep BCC 5.5 happy
854 reverse_iterator_impl operator+(int n) const
855 { return reverse_iterator_impl<T>(m_cur - n); }
856 reverse_iterator_impl operator+(size_t n) const
857 { return reverse_iterator_impl<T>(m_cur - n); }
858 reverse_iterator_impl operator-(int n) const
859 { return reverse_iterator_impl<T>(m_cur + n); }
860 reverse_iterator_impl operator-(size_t n) const
861 { return reverse_iterator_impl<T>(m_cur + n); }
862 reverse_iterator_impl operator+=(int n)
863 { m_cur -= n; return *this; }
864 reverse_iterator_impl operator+=(size_t n)
865 { m_cur -= n; return *this; }
866 reverse_iterator_impl operator-=(int n)
867 { m_cur += n; return *this; }
868 reverse_iterator_impl operator-=(size_t n)
869 { m_cur += n; return *this; }
870
871 unsigned operator-(const reverse_iterator_impl& i) const
872 { return i.m_cur - m_cur; }
873
874 bool operator==(const reverse_iterator_impl& ri) const
875 { return m_cur == ri.m_cur; }
876 bool operator!=(const reverse_iterator_impl& ri) const
877 { return !(*this == ri); }
878
879 bool operator<(const reverse_iterator_impl& i) const
880 { return m_cur > i.m_cur; }
881 bool operator>(const reverse_iterator_impl& i) const
882 { return m_cur < i.m_cur; }
883 bool operator<=(const reverse_iterator_impl& i) const
884 { return m_cur >= i.m_cur; }
885 bool operator>=(const reverse_iterator_impl& i) const
886 { return m_cur <= i.m_cur; }
887
888 private:
889 iterator_type m_cur;
890 };
891
892 typedef reverse_iterator_impl<iterator> reverse_iterator;
893 typedef reverse_iterator_impl<const_iterator> const_reverse_iterator;
894
895private:
896 // used to transform an expression built using c_str() (and hence of type
897 // wxCStrData) to an iterator into the string
898 static const_iterator CreateConstIterator(const wxCStrData& data)
899 {
900 return const_iterator(data.m_str,
901 (data.m_str->begin() + data.m_offset).impl());
902 }
903
904 // in UTF-8 STL build, creation from std::string requires conversion under
905 // non-UTF8 locales, so we can't have and use wxString(wxStringImpl) ctor;
906 // instead we define dummy type that lets us have wxString ctor for creation
907 // from wxStringImpl that couldn't be used by user code (in all other builds,
908 // "standard" ctors can be used):
909#if wxUSE_UNICODE_UTF8 && wxUSE_STL_BASED_WXSTRING
910 struct CtorFromStringImplTag {};
911
912 wxString(CtorFromStringImplTag* WXUNUSED(dummy), const wxStringImpl& src)
913 : m_impl(src) {}
914
915 static wxString FromImpl(const wxStringImpl& src)
916 { return wxString((CtorFromStringImplTag*)NULL, src); }
917#else
918 #if !wxUSE_STL_BASED_WXSTRING
919 wxString(const wxStringImpl& src) : m_impl(src) { }
920 // else: already defined as wxString(wxStdString) below
921 #endif
922 static wxString FromImpl(const wxStringImpl& src) { return wxString(src); }
923#endif
924
925public:
926 // constructors and destructor
927 // ctor for an empty string
928 wxString() {}
929
930 // copy ctor
931 wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { }
932
933 // string containing nRepeat copies of ch
934 wxString(wxUniChar ch, size_t nRepeat = 1)
935 { assign(nRepeat, ch); }
936 wxString(size_t nRepeat, wxUniChar ch)
937 { assign(nRepeat, ch); }
938 wxString(wxUniCharRef ch, size_t nRepeat = 1)
939 { assign(nRepeat, ch); }
940 wxString(size_t nRepeat, wxUniCharRef ch)
941 { assign(nRepeat, ch); }
942 wxString(char ch, size_t nRepeat = 1)
943 { assign(nRepeat, ch); }
944 wxString(size_t nRepeat, char ch)
945 { assign(nRepeat, ch); }
946 wxString(wchar_t ch, size_t nRepeat = 1)
947 { assign(nRepeat, ch); }
948 wxString(size_t nRepeat, wchar_t ch)
949 { assign(nRepeat, ch); }
950
951 // ctors from char* strings:
952 wxString(const char *psz)
953 : m_impl(ImplStr(psz)) {}
954 wxString(const char *psz, const wxMBConv& conv)
955 : m_impl(ImplStr(psz, conv)) {}
956 wxString(const char *psz, size_t nLength)
957 { assign(psz, nLength); }
958 wxString(const char *psz, const wxMBConv& conv, size_t nLength)
959 {
960 SubstrBufFromMB str(ImplStr(psz, nLength, conv));
961 m_impl.assign(str.data, str.len);
962 }
963
964 // and unsigned char*:
965 wxString(const unsigned char *psz)
966 : m_impl(ImplStr((const char*)psz)) {}
967 wxString(const unsigned char *psz, const wxMBConv& conv)
968 : m_impl(ImplStr((const char*)psz, conv)) {}
969 wxString(const unsigned char *psz, size_t nLength)
970 { assign((const char*)psz, nLength); }
971 wxString(const unsigned char *psz, const wxMBConv& conv, size_t nLength)
972 {
973 SubstrBufFromMB str(ImplStr((const char*)psz, nLength, conv));
974 m_impl.assign(str.data, str.len);
975 }
976
977 // ctors from wchar_t* strings:
978 wxString(const wchar_t *pwz)
979 : m_impl(ImplStr(pwz)) {}
980 wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv))
981 : m_impl(ImplStr(pwz)) {}
982 wxString(const wchar_t *pwz, size_t nLength)
983 { assign(pwz, nLength); }
984 wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv), size_t nLength)
985 { assign(pwz, nLength); }
986
987 wxString(const wxCharBuffer& buf)
988 { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length
989 wxString(const wxWCharBuffer& buf)
990 { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length
991
992 wxString(const wxCStrData& cstr)
993 : m_impl(cstr.AsString().m_impl) { }
994
995 // as we provide both ctors with this signature for both char and unsigned
996 // char string, we need to provide one for wxCStrData to resolve ambiguity
997 wxString(const wxCStrData& cstr, size_t nLength)
998 : m_impl(cstr.AsString().Mid(0, nLength).m_impl) {}
999
1000 // and because wxString is convertible to wxCStrData and const wxChar *
1001 // we also need to provide this one
1002 wxString(const wxString& str, size_t nLength)
1003 { assign(str, nLength); }
1004
1005 // even if we're not built with wxUSE_STL == 1 it is very convenient to allow
1006 // implicit conversions from std::string to wxString and vice verse as this
1007 // allows to use the same strings in non-GUI and GUI code, however we don't
1008 // want to unconditionally add this ctor as it would make wx lib dependent on
1009 // libstdc++ on some Linux versions which is bad, so instead we ask the
1010 // client code to define this wxUSE_STD_STRING symbol if they need it
1011#if wxUSE_STD_STRING
1012 #if wxUSE_UNICODE_WCHAR
1013 wxString(const wxStdWideString& str) : m_impl(str) {}
1014 #else // UTF-8 or ANSI
1015 wxString(const wxStdWideString& str)
1016 { assign(str.c_str(), str.length()); }
1017 #endif
1018
1019 #if !wxUSE_UNICODE // ANSI build
1020 // FIXME-UTF8: do this in UTF8 build #if wxUSE_UTF8_LOCALE_ONLY, too
1021 wxString(const std::string& str) : m_impl(str) {}
1022 #else // Unicode
1023 wxString(const std::string& str)
1024 { assign(str.c_str(), str.length()); }
1025 #endif
1026#endif // wxUSE_STD_STRING
1027
1028 // Unlike ctor from std::string, we provide conversion to std::string only
1029 // if wxUSE_STL and not merely wxUSE_STD_STRING (which is on by default),
1030 // because it conflicts with operator const char/wchar_t*:
1031#if wxUSE_STL
1032 #if wxUSE_UNICODE_WCHAR && wxUSE_STL_BASED_WXSTRING
1033 // wxStringImpl is std::string in the encoding we want
1034 operator const wxStdWideString&() const { return m_impl; }
1035 #else
1036 // wxStringImpl is either not std::string or needs conversion
1037 operator wxStdWideString() const
1038 // FIXME-UTF8: broken for embedded NULs
1039 { return wxStdWideString(wc_str()); }
1040 #endif
1041
1042 #if (!wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY) && wxUSE_STL_BASED_WXSTRING
1043 // wxStringImpl is std::string in the encoding we want
1044 operator const std::string&() const { return m_impl; }
1045 #else
1046 // wxStringImpl is either not std::string or needs conversion
1047 operator std::string() const
1048 // FIXME-UTF8: broken for embedded NULs
1049 { return std::string(mb_str()); }
1050 #endif
1051#endif // wxUSE_STL
1052
1053 // first valid index position
1054 const_iterator begin() const { return const_iterator(this, m_impl.begin()); }
1055 iterator begin() { return iterator(this, m_impl.begin()); }
1056 // position one after the last valid one
1057 const_iterator end() const { return const_iterator(this, m_impl.end()); }
1058 iterator end() { return iterator(this, m_impl.end()); }
1059
1060 // first element of the reversed string
1061 const_reverse_iterator rbegin() const
1062 { return const_reverse_iterator(end()); }
1063 reverse_iterator rbegin()
1064 { return reverse_iterator(end()); }
1065 // one beyond the end of the reversed string
1066 const_reverse_iterator rend() const
1067 { return const_reverse_iterator(begin()); }
1068 reverse_iterator rend()
1069 { return reverse_iterator(begin()); }
1070
1071 // std::string methods:
1072#if wxUSE_UNICODE_UTF8
1073 size_t length() const { return end() - begin(); } // FIXME-UTF8: optimize!
1074#else
1075 size_t length() const { return m_impl.length(); }
1076#endif
1077
1078 size_type size() const { return length(); }
1079 size_type max_size() const { return npos; }
1080
1081 bool empty() const { return m_impl.empty(); }
1082
1083 size_type capacity() const { return m_impl.capacity(); } // FIXME-UTF8
1084 void reserve(size_t sz) { m_impl.reserve(sz); } // FIXME-UTF8
1085
1086 void resize(size_t nSize, wxUniChar ch = wxT('\0'))
1087 {
1088#if wxUSE_UNICODE_UTF8
1089 if ( !ch.IsAscii() )
1090 {
1091 size_t len = length();
1092 if ( nSize == len)
1093 return;
1094 else if ( nSize < len )
1095 erase(nSize);
1096 else
1097 append(nSize - len, ch);
1098 }
1099 else
1100#endif
1101 m_impl.resize(nSize, (wxStringCharType)ch);
1102 }
1103
1104 wxString substr(size_t nStart = 0, size_t nLen = npos) const
1105 {
1106 size_t pos, len;
1107 PosLenToImpl(nStart, nLen, &pos, &len);
1108 return FromImpl(m_impl.substr(pos, len));
1109 }
1110
1111 // generic attributes & operations
1112 // as standard strlen()
1113 size_t Len() const { return length(); }
1114 // string contains any characters?
1115 bool IsEmpty() const { return empty(); }
1116 // empty string is "false", so !str will return true
1117 bool operator!() const { return empty(); }
1118 // truncate the string to given length
1119 wxString& Truncate(size_t uiLen);
1120 // empty string contents
1121 void Empty()
1122 {
1123 Truncate(0);
1124
1125 wxASSERT_MSG( empty(), _T("string not empty after call to Empty()?") );
1126 }
1127 // empty the string and free memory
1128 void Clear()
1129 {
1130 wxString tmp(wxEmptyString);
1131 swap(tmp);
1132 }
1133
1134 // contents test
1135 // Is an ascii value
1136 bool IsAscii() const;
1137 // Is a number
1138 bool IsNumber() const;
1139 // Is a word
1140 bool IsWord() const;
1141
1142 // data access (all indexes are 0 based)
1143 // read access
1144 wxUniChar at(size_t n) const
1145 { return *(begin() + n); } // FIXME-UTF8: optimize?
1146 wxUniChar GetChar(size_t n) const
1147 { return at(n); }
1148 // read/write access
1149 wxUniCharRef at(size_t n)
1150 { return *(begin() + n); } // FIXME-UTF8: optimize?
1151 wxUniCharRef GetWritableChar(size_t n)
1152 { return at(n); }
1153 // write access
1154 void SetChar(size_t n, wxUniChar ch)
1155 { at(n) = ch; }
1156
1157 // get last character
1158 wxUniChar Last() const
1159 {
1160 wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
1161 return *rbegin();
1162 }
1163
1164 // get writable last character
1165 wxUniCharRef Last()
1166 {
1167 wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
1168 return *rbegin();
1169 }
1170
1171 /*
1172 Note that we we must define all of the overloads below to avoid
1173 ambiguity when using str[0].
1174 */
1175 wxUniChar operator[](int n) const
1176 { return at(n); }
1177 wxUniChar operator[](long n) const
1178 { return at(n); }
1179 wxUniChar operator[](size_t n) const
1180 { return at(n); }
1181#ifndef wxSIZE_T_IS_UINT
1182 wxUniChar operator[](unsigned int n) const
1183 { return at(n); }
1184#endif // size_t != unsigned int
1185
1186 // operator versions of GetWriteableChar()
1187 wxUniCharRef operator[](int n)
1188 { return at(n); }
1189 wxUniCharRef operator[](long n)
1190 { return at(n); }
1191 wxUniCharRef operator[](size_t n)
1192 { return at(n); }
1193#ifndef wxSIZE_T_IS_UINT
1194 wxUniCharRef operator[](unsigned int n)
1195 { return at(n); }
1196#endif // size_t != unsigned int
1197
1198 // explicit conversion to C string (use this with printf()!)
1199 wxCStrData c_str() const { return wxCStrData(this); }
1200 wxCStrData data() const { return c_str(); }
1201
1202 // implicit conversion to C string
1203 operator wxCStrData() const { return c_str(); }
1204
1205 // the first two operators conflict with operators for conversion to
1206 // std::string and they must be disabled in STL build; the next one only
1207 // makes sense if conversions to char* are also defined and not defining it
1208 // in STL build also helps us to get more clear error messages for the code
1209 // which relies on implicit conversion to char* in STL build
1210#if !wxUSE_STL
1211 operator const char*() const { return c_str(); }
1212 operator const wchar_t*() const { return c_str(); }
1213
1214 // implicit conversion to untyped pointer for compatibility with previous
1215 // wxWidgets versions: this is the same as conversion to const char * so it
1216 // may fail!
1217 operator const void*() const { return c_str(); }
1218#endif // wxUSE_STL
1219
1220 // identical to c_str(), for MFC compatibility
1221 const wxCStrData GetData() const { return c_str(); }
1222
1223 // explicit conversion to C string in internal representation (char*,
1224 // wchar_t*, UTF-8-encoded char*, depending on the build):
1225 const wxStringCharType *wx_str() const { return m_impl.c_str(); }
1226
1227 // conversion to *non-const* multibyte or widestring buffer; modifying
1228 // returned buffer won't affect the string, these methods are only useful
1229 // for passing values to const-incorrect functions
1230 wxWritableCharBuffer char_str(const wxMBConv& conv = wxConvLibc) const
1231 { return mb_str(conv); }
1232 wxWritableWCharBuffer wchar_str() const { return wc_str(); }
1233
1234 // conversion to/from plain (i.e. 7 bit) ASCII: this is useful for
1235 // converting numbers or strings which are certain not to contain special
1236 // chars (typically system functions, X atoms, environment variables etc.)
1237 //
1238 // the behaviour of these functions with the strings containing anything
1239 // else than 7 bit ASCII characters is undefined, use at your own risk.
1240#if wxUSE_UNICODE
1241 static wxString FromAscii(const char *ascii, size_t len); // string
1242 static wxString FromAscii(const char *ascii); // string
1243 static wxString FromAscii(const char ascii); // char
1244 const wxCharBuffer ToAscii() const;
1245#else // ANSI
1246 static wxString FromAscii(const char *ascii) { return wxString( ascii ); }
1247 static wxString FromAscii(const char *ascii, size_t len)
1248 { return wxString( ascii, len ); }
1249 static wxString FromAscii(const char ascii) { return wxString( ascii ); }
1250 const char *ToAscii() const { return c_str(); }
1251#endif // Unicode/!Unicode
1252
1253 // conversion to/from UTF-8:
1254#if wxUSE_UNICODE_UTF8
1255 static wxString FromUTF8(const char *utf8)
1256 {
1257 if ( !utf8 )
1258 return wxEmptyString;
1259
1260 wxASSERT( wxStringOperations::IsValidUtf8String(utf8) );
1261 return FromImpl(wxStringImpl(utf8));
1262 }
1263 static wxString FromUTF8(const char *utf8, size_t len)
1264 {
1265 if ( !utf8 )
1266 return wxEmptyString;
1267 if ( len == npos )
1268 return FromUTF8(utf8);
1269
1270 wxASSERT( wxStringOperations::IsValidUtf8String(utf8, len) );
1271 return FromImpl(wxStringImpl(utf8, len));
1272 }
1273 const char* utf8_str() const { return wx_str(); }
1274 const char* ToUTF8() const { return wx_str(); }
1275#elif wxUSE_UNICODE_WCHAR
1276 static wxString FromUTF8(const char *utf8)
1277 { return wxString(utf8, wxMBConvUTF8()); }
1278 static wxString FromUTF8(const char *utf8, size_t len)
1279 { return wxString(utf8, wxMBConvUTF8(), len); }
1280 const wxCharBuffer utf8_str() const { return mb_str(wxMBConvUTF8()); }
1281 const wxCharBuffer ToUTF8() const { return utf8_str(); }
1282#else // ANSI
1283 static wxString FromUTF8(const char *utf8)
1284 { return wxString(wxMBConvUTF8().cMB2WC(utf8)); }
1285 static wxString FromUTF8(const char *utf8, size_t len)
1286 {
1287 size_t wlen;
1288 wxWCharBuffer buf(wxMBConvUTF8().cMB2WC(utf8, len == npos ? wxNO_LEN : len, &wlen));
1289 return wxString(buf.data(), wlen);
1290 }
1291 const wxCharBuffer utf8_str() const
1292 { return wxMBConvUTF8().cWC2MB(wc_str()); }
1293 const wxCharBuffer ToUTF8() const { return utf8_str(); }
1294#endif
1295
1296 // functions for storing binary data in wxString:
1297#if wxUSE_UNICODE
1298 static wxString From8BitData(const char *data, size_t len)
1299 { return wxString(data, wxConvISO8859_1, len); }
1300 // version for NUL-terminated data:
1301 static wxString From8BitData(const char *data)
1302 { return wxString(data, wxConvISO8859_1); }
1303 const wxCharBuffer To8BitData() const { return mb_str(wxConvISO8859_1); }
1304#else // ANSI
1305 static wxString From8BitData(const char *data, size_t len)
1306 { return wxString(data, len); }
1307 // version for NUL-terminated data:
1308 static wxString From8BitData(const char *data)
1309 { return wxString(data); }
1310 const char *To8BitData() const { return c_str(); }
1311#endif // Unicode/ANSI
1312
1313 // conversions with (possible) format conversions: have to return a
1314 // buffer with temporary data
1315 //
1316 // the functions defined (in either Unicode or ANSI) mode are mb_str() to
1317 // return an ANSI (multibyte) string, wc_str() to return a wide string and
1318 // fn_str() to return a string which should be used with the OS APIs
1319 // accepting the file names. The return value is always the same, but the
1320 // type differs because a function may either return pointer to the buffer
1321 // directly or have to use intermediate buffer for translation.
1322#if wxUSE_UNICODE
1323
1324#if wxUSE_UTF8_LOCALE_ONLY
1325 const char* mb_str() const { return wx_str(); }
1326 const wxCharBuffer mb_str(const wxMBConv& conv) const;
1327#else
1328 const wxCharBuffer mb_str(const wxMBConv& conv = wxConvLibc) const;
1329#endif
1330
1331 const wxWX2MBbuf mbc_str() const { return mb_str(*wxConvCurrent); }
1332
1333#if wxUSE_UNICODE_WCHAR
1334 const wxChar* wc_str() const { return wx_str(); }
1335#elif wxUSE_UNICODE_UTF8
1336 const wxWCharBuffer wc_str() const;
1337#endif
1338 // for compatibility with !wxUSE_UNICODE version
1339 const wxWX2WCbuf wc_str(const wxMBConv& WXUNUSED(conv)) const
1340 { return wc_str(); }
1341
1342#if wxMBFILES
1343 const wxCharBuffer fn_str() const { return mb_str(wxConvFile); }
1344#else // !wxMBFILES
1345 const wxWX2WCbuf fn_str() const { return wc_str(); }
1346#endif // wxMBFILES/!wxMBFILES
1347
1348#else // ANSI
1349 const wxChar* mb_str() const { return wx_str(); }
1350
1351 // for compatibility with wxUSE_UNICODE version
1352 const wxChar* mb_str(const wxMBConv& WXUNUSED(conv)) const { return wx_str(); }
1353
1354 const wxWX2MBbuf mbc_str() const { return mb_str(); }
1355
1356#if wxUSE_WCHAR_T
1357 const wxWCharBuffer wc_str(const wxMBConv& conv = wxConvLibc) const;
1358#endif // wxUSE_WCHAR_T
1359#ifdef __WXOSX__
1360 const wxCharBuffer fn_str() const { return wxConvFile.cWC2WX( wc_str( wxConvLocal ) ); }
1361#else
1362 const wxChar* fn_str() const { return c_str(); }
1363#endif
1364#endif // Unicode/ANSI
1365
1366 // overloaded assignment
1367 // from another wxString
1368 wxString& operator=(const wxString& stringSrc)
1369 { m_impl = stringSrc.m_impl; return *this; }
1370 wxString& operator=(const wxCStrData& cstr)
1371 { return *this = cstr.AsString(); }
1372 // from a character
1373 wxString& operator=(wxUniChar ch)
1374 { m_impl = wxStringOperations::EncodeChar(ch); return *this; }
1375 wxString& operator=(wxUniCharRef ch)
1376 { return operator=((wxUniChar)ch); }
1377 wxString& operator=(char ch)
1378 { return operator=(wxUniChar(ch)); }
1379 wxString& operator=(unsigned char ch)
1380 { return operator=(wxUniChar(ch)); }
1381 wxString& operator=(wchar_t ch)
1382 { return operator=(wxUniChar(ch)); }
1383 // from a C string - STL probably will crash on NULL,
1384 // so we need to compensate in that case
1385#if wxUSE_STL_BASED_WXSTRING
1386 wxString& operator=(const char *psz)
1387 { if (psz) m_impl = ImplStr(psz); else Clear(); return *this; }
1388 wxString& operator=(const wchar_t *pwz)
1389 { if (pwz) m_impl = ImplStr(pwz); else Clear(); return *this; }
1390#else
1391 wxString& operator=(const char *psz)
1392 { m_impl = ImplStr(psz); return *this; }
1393 wxString& operator=(const wchar_t *pwz)
1394 { m_impl = ImplStr(pwz); return *this; }
1395#endif
1396 wxString& operator=(const unsigned char *psz)
1397 { return operator=((const char*)psz); }
1398
1399 // from wxWCharBuffer
1400 wxString& operator=(const wxWCharBuffer& s)
1401 { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs
1402 // from wxCharBuffer
1403 wxString& operator=(const wxCharBuffer& s)
1404 { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs
1405
1406 // string concatenation
1407 // in place concatenation
1408 /*
1409 Concatenate and return the result. Note that the left to right
1410 associativity of << allows to write things like "str << str1 << str2
1411 << ..." (unlike with +=)
1412 */
1413 // string += string
1414 wxString& operator<<(const wxString& s)
1415 {
1416#if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
1417 wxASSERT_MSG( s.IsValid(),
1418 _T("did you forget to call UngetWriteBuf()?") );
1419#endif
1420
1421 append(s);
1422 return *this;
1423 }
1424 // string += C string
1425 wxString& operator<<(const char *psz)
1426 { append(psz); return *this; }
1427 wxString& operator<<(const wchar_t *pwz)
1428 { append(pwz); return *this; }
1429 wxString& operator<<(const wxCStrData& psz)
1430 { append(psz.AsString()); return *this; }
1431 // string += char
1432 wxString& operator<<(wxUniChar ch) { append(1, ch); return *this; }
1433 wxString& operator<<(wxUniCharRef ch) { append(1, ch); return *this; }
1434 wxString& operator<<(char ch) { append(1, ch); return *this; }
1435 wxString& operator<<(unsigned char ch) { append(1, ch); return *this; }
1436 wxString& operator<<(wchar_t ch) { append(1, ch); return *this; }
1437
1438 // string += buffer (i.e. from wxGetString)
1439 wxString& operator<<(const wxWCharBuffer& s)
1440 { return operator<<((const wchar_t *)s); }
1441 wxString& operator<<(const wxCharBuffer& s)
1442 { return operator<<((const char *)s); }
1443
1444 // string += C string
1445 wxString& Append(const wxString& s)
1446 {
1447 // test for empty() to share the string if possible
1448 if ( empty() )
1449 *this = s;
1450 else
1451 append(s);
1452 return *this;
1453 }
1454 wxString& Append(const char* psz)
1455 { append(psz); return *this; }
1456 wxString& Append(const wchar_t* pwz)
1457 { append(pwz); return *this; }
1458 wxString& Append(const wxCStrData& psz)
1459 { append(psz); return *this; }
1460 wxString& Append(const wxCharBuffer& psz)
1461 { append(psz); return *this; }
1462 wxString& Append(const wxWCharBuffer& psz)
1463 { append(psz); return *this; }
1464 // append count copies of given character
1465 wxString& Append(wxUniChar ch, size_t count = 1u)
1466 { append(count, ch); return *this; }
1467 wxString& Append(wxUniCharRef ch, size_t count = 1u)
1468 { append(count, ch); return *this; }
1469 wxString& Append(char ch, size_t count = 1u)
1470 { append(count, ch); return *this; }
1471 wxString& Append(unsigned char ch, size_t count = 1u)
1472 { append(count, ch); return *this; }
1473 wxString& Append(wchar_t ch, size_t count = 1u)
1474 { append(count, ch); return *this; }
1475 wxString& Append(const char* psz, size_t nLen)
1476 { append(psz, nLen); return *this; }
1477 wxString& Append(const wchar_t* pwz, size_t nLen)
1478 { append(pwz, nLen); return *this; }
1479
1480 // prepend a string, return the string itself
1481 wxString& Prepend(const wxString& str)
1482 { *this = str + *this; return *this; }
1483
1484 // non-destructive concatenation
1485 // two strings
1486 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string1,
1487 const wxString& string2);
1488 // string with a single char
1489 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxUniChar ch);
1490 // char with a string
1491 friend wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string);
1492 // string with C string
1493 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string,
1494 const char *psz);
1495 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string,
1496 const wchar_t *pwz);
1497 // C string with string
1498 friend wxString WXDLLIMPEXP_BASE operator+(const char *psz,
1499 const wxString& string);
1500 friend wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz,
1501 const wxString& string);
1502
1503 // stream-like functions
1504 // insert an int into string
1505 wxString& operator<<(int i)
1506 { return (*this) << Format(_T("%d"), i); }
1507 // insert an unsigned int into string
1508 wxString& operator<<(unsigned int ui)
1509 { return (*this) << Format(_T("%u"), ui); }
1510 // insert a long into string
1511 wxString& operator<<(long l)
1512 { return (*this) << Format(_T("%ld"), l); }
1513 // insert an unsigned long into string
1514 wxString& operator<<(unsigned long ul)
1515 { return (*this) << Format(_T("%lu"), ul); }
1516#if defined wxLongLong_t && !defined wxLongLongIsLong
1517 // insert a long long if they exist and aren't longs
1518 wxString& operator<<(wxLongLong_t ll)
1519 {
1520 const wxChar *fmt = _T("%") wxLongLongFmtSpec _T("d");
1521 return (*this) << Format(fmt, ll);
1522 }
1523 // insert an unsigned long long
1524 wxString& operator<<(wxULongLong_t ull)
1525 {
1526 const wxChar *fmt = _T("%") wxLongLongFmtSpec _T("u");
1527 return (*this) << Format(fmt , ull);
1528 }
1529#endif
1530 // insert a float into string
1531 wxString& operator<<(float f)
1532 { return (*this) << Format(_T("%f"), f); }
1533 // insert a double into string
1534 wxString& operator<<(double d)
1535 { return (*this) << Format(_T("%g"), d); }
1536
1537 // string comparison
1538 // case-sensitive comparison (returns a value < 0, = 0 or > 0)
1539 int Cmp(const char *psz) const
1540 { return compare(psz); }
1541 int Cmp(const wchar_t *pwz) const
1542 { return compare(pwz); }
1543 int Cmp(const wxString& s) const
1544 { return compare(s); }
1545 int Cmp(const wxCStrData& s) const
1546 { return compare(s); }
1547 int Cmp(const wxCharBuffer& s) const
1548 { return compare(s); }
1549 int Cmp(const wxWCharBuffer& s) const
1550 { return compare(s); }
1551 // same as Cmp() but not case-sensitive
1552 int CmpNoCase(const wxString& s) const;
1553 // test for the string equality, either considering case or not
1554 // (if compareWithCase then the case matters)
1555 bool IsSameAs(const wxString& str, bool compareWithCase = true) const
1556 { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
1557 bool IsSameAs(const char *str, bool compareWithCase = true) const
1558 { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
1559 bool IsSameAs(const wchar_t *str, bool compareWithCase = true) const
1560 { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
1561 bool IsSameAs(const wxCStrData& str, bool compareWithCase = true) const
1562 { return IsSameAs(str.AsString(), compareWithCase); }
1563 bool IsSameAs(const wxCharBuffer& str, bool compareWithCase = true) const
1564 { return IsSameAs(str.data(), compareWithCase); }
1565 bool IsSameAs(const wxWCharBuffer& str, bool compareWithCase = true) const
1566 { return IsSameAs(str.data(), compareWithCase); }
1567 // comparison with a single character: returns true if equal
1568 bool IsSameAs(wxUniChar c, bool compareWithCase = true) const;
1569 // FIXME-UTF8: remove these overloads
1570 bool IsSameAs(wxUniCharRef c, bool compareWithCase = true) const
1571 { return IsSameAs(wxUniChar(c), compareWithCase); }
1572 bool IsSameAs(char c, bool compareWithCase = true) const
1573 { return IsSameAs(wxUniChar(c), compareWithCase); }
1574 bool IsSameAs(unsigned char c, bool compareWithCase = true) const
1575 { return IsSameAs(wxUniChar(c), compareWithCase); }
1576 bool IsSameAs(wchar_t c, bool compareWithCase = true) const
1577 { return IsSameAs(wxUniChar(c), compareWithCase); }
1578 bool IsSameAs(int c, bool compareWithCase = true) const
1579 { return IsSameAs(wxUniChar(c), compareWithCase); }
1580
1581 // simple sub-string extraction
1582 // return substring starting at nFirst of length nCount (or till the end
1583 // if nCount = default value)
1584 wxString Mid(size_t nFirst, size_t nCount = npos) const;
1585
1586 // operator version of Mid()
1587 wxString operator()(size_t start, size_t len) const
1588 { return Mid(start, len); }
1589
1590 // check if the string starts with the given prefix and return the rest
1591 // of the string in the provided pointer if it is not NULL; otherwise
1592 // return false
1593 bool StartsWith(const wxString& prefix, wxString *rest = NULL) const;
1594 // check if the string ends with the given suffix and return the
1595 // beginning of the string before the suffix in the provided pointer if
1596 // it is not NULL; otherwise return false
1597 bool EndsWith(const wxString& suffix, wxString *rest = NULL) const;
1598
1599 // get first nCount characters
1600 wxString Left(size_t nCount) const;
1601 // get last nCount characters
1602 wxString Right(size_t nCount) const;
1603 // get all characters before the first occurance of ch
1604 // (returns the whole string if ch not found)
1605 wxString BeforeFirst(wxUniChar ch) const;
1606 // get all characters before the last occurence of ch
1607 // (returns empty string if ch not found)
1608 wxString BeforeLast(wxUniChar ch) const;
1609 // get all characters after the first occurence of ch
1610 // (returns empty string if ch not found)
1611 wxString AfterFirst(wxUniChar ch) const;
1612 // get all characters after the last occurence of ch
1613 // (returns the whole string if ch not found)
1614 wxString AfterLast(wxUniChar ch) const;
1615
1616 // for compatibility only, use more explicitly named functions above
1617 wxString Before(wxUniChar ch) const { return BeforeLast(ch); }
1618 wxString After(wxUniChar ch) const { return AfterFirst(ch); }
1619
1620 // case conversion
1621 // convert to upper case in place, return the string itself
1622 wxString& MakeUpper();
1623 // convert to upper case, return the copy of the string
1624 // Here's something to remember: BC++ doesn't like returns in inlines.
1625 wxString Upper() const ;
1626 // convert to lower case in place, return the string itself
1627 wxString& MakeLower();
1628 // convert to lower case, return the copy of the string
1629 wxString Lower() const ;
1630
1631 // trimming/padding whitespace (either side) and truncating
1632 // remove spaces from left or from right (default) side
1633 wxString& Trim(bool bFromRight = true);
1634 // add nCount copies chPad in the beginning or at the end (default)
1635 wxString& Pad(size_t nCount, wxUniChar chPad = wxT(' '), bool bFromRight = true);
1636
1637 // searching and replacing
1638 // searching (return starting index, or -1 if not found)
1639 int Find(wxUniChar ch, bool bFromEnd = false) const; // like strchr/strrchr
1640 int Find(wxUniCharRef ch, bool bFromEnd = false) const
1641 { return Find(wxUniChar(ch), bFromEnd); }
1642 int Find(char ch, bool bFromEnd = false) const
1643 { return Find(wxUniChar(ch), bFromEnd); }
1644 int Find(unsigned char ch, bool bFromEnd = false) const
1645 { return Find(wxUniChar(ch), bFromEnd); }
1646 int Find(wchar_t ch, bool bFromEnd = false) const
1647 { return Find(wxUniChar(ch), bFromEnd); }
1648 // searching (return starting index, or -1 if not found)
1649 int Find(const wxString& sub) const // like strstr
1650 {
1651 size_type idx = find(sub);
1652 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1653 }
1654 int Find(const char *sub) const // like strstr
1655 {
1656 size_type idx = find(sub);
1657 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1658 }
1659 int Find(const wchar_t *sub) const // like strstr
1660 {
1661 size_type idx = find(sub);
1662 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1663 }
1664
1665 int Find(const wxCStrData& sub) const
1666 { return Find(sub.AsString()); }
1667 int Find(const wxCharBuffer& sub) const
1668 { return Find(sub.data()); }
1669 int Find(const wxWCharBuffer& sub) const
1670 { return Find(sub.data()); }
1671
1672 // replace first (or all of bReplaceAll) occurences of substring with
1673 // another string, returns the number of replacements made
1674 size_t Replace(const wxString& strOld,
1675 const wxString& strNew,
1676 bool bReplaceAll = true);
1677
1678 // check if the string contents matches a mask containing '*' and '?'
1679 bool Matches(const wxString& mask) const;
1680
1681 // conversion to numbers: all functions return true only if the whole
1682 // string is a number and put the value of this number into the pointer
1683 // provided, the base is the numeric base in which the conversion should be
1684 // done and must be comprised between 2 and 36 or be 0 in which case the
1685 // standard C rules apply (leading '0' => octal, "0x" => hex)
1686 // convert to a signed integer
1687 bool ToLong(long *val, int base = 10) const;
1688 // convert to an unsigned integer
1689 bool ToULong(unsigned long *val, int base = 10) const;
1690 // convert to wxLongLong
1691#if defined(wxLongLong_t)
1692 bool ToLongLong(wxLongLong_t *val, int base = 10) const;
1693 // convert to wxULongLong
1694 bool ToULongLong(wxULongLong_t *val, int base = 10) const;
1695#endif // wxLongLong_t
1696 // convert to a double
1697 bool ToDouble(double *val) const;
1698
1699
1700#ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1701 // formatted input/output
1702 // as sprintf(), returns the number of characters written or < 0 on error
1703 // (take 'this' into account in attribute parameter count)
1704 // int Printf(const wxString& format, ...);
1705 WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&),
1706 DoPrintfWchar, DoPrintfUtf8)
1707#ifdef __WATCOMC__
1708 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1709 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxString&),
1710 (wxFormatString(f1)));
1711 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxCStrData&),
1712 (wxFormatString(f1)));
1713 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const char*),
1714 (wxFormatString(f1)));
1715 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wchar_t*),
1716 (wxFormatString(f1)));
1717#endif
1718#endif // !wxNEEDS_WXSTRING_PRINTF_MIXIN
1719 // as vprintf(), returns the number of characters written or < 0 on error
1720 int PrintfV(const wxString& format, va_list argptr);
1721
1722#ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1723 // returns the string containing the result of Printf() to it
1724 // static wxString Format(const wxString& format, ...) ATTRIBUTE_PRINTF_1;
1725 WX_DEFINE_VARARG_FUNC(static wxString, Format, 1, (const wxFormatString&),
1726 DoFormatWchar, DoFormatUtf8)
1727#ifdef __WATCOMC__
1728 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1729 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxString&),
1730 (wxFormatString(f1)));
1731 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxCStrData&),
1732 (wxFormatString(f1)));
1733 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const char*),
1734 (wxFormatString(f1)));
1735 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wchar_t*),
1736 (wxFormatString(f1)));
1737#endif
1738#endif
1739 // the same as above, but takes a va_list
1740 static wxString FormatV(const wxString& format, va_list argptr);
1741
1742 // raw access to string memory
1743 // ensure that string has space for at least nLen characters
1744 // only works if the data of this string is not shared
1745 bool Alloc(size_t nLen) { reserve(nLen); /*return capacity() >= nLen;*/ return true; }
1746 // minimize the string's memory
1747 // only works if the data of this string is not shared
1748 bool Shrink();
1749#if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
1750 // These are deprecated, use wxStringBuffer or wxStringBufferLength instead
1751 //
1752 // get writable buffer of at least nLen bytes. Unget() *must* be called
1753 // a.s.a.p. to put string back in a reasonable state!
1754 wxDEPRECATED( wxStringCharType *GetWriteBuf(size_t nLen) );
1755 // call this immediately after GetWriteBuf() has been used
1756 wxDEPRECATED( void UngetWriteBuf() );
1757 wxDEPRECATED( void UngetWriteBuf(size_t nLen) );
1758#endif // WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && wxUSE_UNICODE_UTF8
1759
1760 // wxWidgets version 1 compatibility functions
1761
1762 // use Mid()
1763 wxString SubString(size_t from, size_t to) const
1764 { return Mid(from, (to - from + 1)); }
1765 // values for second parameter of CompareTo function
1766 enum caseCompare {exact, ignoreCase};
1767 // values for first parameter of Strip function
1768 enum stripType {leading = 0x1, trailing = 0x2, both = 0x3};
1769
1770#ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1771 // use Printf()
1772 // (take 'this' into account in attribute parameter count)
1773 // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2;
1774 WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&),
1775 DoPrintfWchar, DoPrintfUtf8)
1776#ifdef __WATCOMC__
1777 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1778 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxString&),
1779 (wxFormatString(f1)));
1780 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxCStrData&),
1781 (wxFormatString(f1)));
1782 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const char*),
1783 (wxFormatString(f1)));
1784 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wchar_t*),
1785 (wxFormatString(f1)));
1786#endif
1787#endif // wxNEEDS_WXSTRING_PRINTF_MIXIN
1788
1789 // use Cmp()
1790 inline int CompareTo(const wxChar* psz, caseCompare cmp = exact) const
1791 { return cmp == exact ? Cmp(psz) : CmpNoCase(psz); }
1792
1793 // use length()
1794 size_t Length() const { return length(); }
1795 // Count the number of characters
1796 int Freq(wxUniChar ch) const;
1797 // use MakeLower
1798 void LowerCase() { MakeLower(); }
1799 // use MakeUpper
1800 void UpperCase() { MakeUpper(); }
1801 // use Trim except that it doesn't change this string
1802 wxString Strip(stripType w = trailing) const;
1803
1804 // use Find (more general variants not yet supported)
1805 size_t Index(const wxChar* psz) const { return Find(psz); }
1806 size_t Index(wxUniChar ch) const { return Find(ch); }
1807 // use Truncate
1808 wxString& Remove(size_t pos) { return Truncate(pos); }
1809 wxString& RemoveLast(size_t n = 1) { return Truncate(length() - n); }
1810
1811 wxString& Remove(size_t nStart, size_t nLen)
1812 { return (wxString&)erase( nStart, nLen ); }
1813
1814 // use Find()
1815 int First( wxUniChar ch ) const { return Find(ch); }
1816 int First( wxUniCharRef ch ) const { return Find(ch); }
1817 int First( char ch ) const { return Find(ch); }
1818 int First( unsigned char ch ) const { return Find(ch); }
1819 int First( wchar_t ch ) const { return Find(ch); }
1820 int First( const wxString& str ) const { return Find(str); }
1821 int Last( wxUniChar ch ) const { return Find(ch, true); }
1822 bool Contains(const wxString& str) const { return Find(str) != wxNOT_FOUND; }
1823
1824 // use empty()
1825 bool IsNull() const { return empty(); }
1826
1827 // std::string compatibility functions
1828
1829 // take nLen chars starting at nPos
1830 wxString(const wxString& str, size_t nPos, size_t nLen)
1831 { assign(str, nPos, nLen); }
1832 // take all characters from first to last
1833 wxString(const_iterator first, const_iterator last)
1834 : m_impl(first.impl(), last.impl()) { }
1835#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1836 // the 2 overloads below are for compatibility with the existing code using
1837 // pointers instead of iterators
1838 wxString(const char *first, const char *last)
1839 {
1840 SubstrBufFromMB str(ImplStr(first, last - first));
1841 m_impl.assign(str.data, str.len);
1842 }
1843 wxString(const wchar_t *first, const wchar_t *last)
1844 {
1845 SubstrBufFromWC str(ImplStr(first, last - first));
1846 m_impl.assign(str.data, str.len);
1847 }
1848 // and this one is needed to compile code adding offsets to c_str() result
1849 wxString(const wxCStrData& first, const wxCStrData& last)
1850 : m_impl(CreateConstIterator(first).impl(),
1851 CreateConstIterator(last).impl())
1852 {
1853 wxASSERT_MSG( first.m_str == last.m_str,
1854 _T("pointers must be into the same string") );
1855 }
1856#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1857
1858 // lib.string.modifiers
1859 // append elements str[pos], ..., str[pos+n]
1860 wxString& append(const wxString& str, size_t pos, size_t n)
1861 {
1862 size_t from, len;
1863 str.PosLenToImpl(pos, n, &from, &len);
1864 m_impl.append(str.m_impl, from, len);
1865 return *this;
1866 }
1867 // append a string
1868 wxString& append(const wxString& str)
1869 { m_impl.append(str.m_impl); return *this; }
1870 // append first n (or all if n == npos) characters of sz
1871 wxString& append(const char *sz)
1872 { m_impl.append(ImplStr(sz)); return *this; }
1873 wxString& append(const wchar_t *sz)
1874 { m_impl.append(ImplStr(sz)); return *this; }
1875 wxString& append(const char *sz, size_t n)
1876 {
1877 SubstrBufFromMB str(ImplStr(sz, n));
1878 m_impl.append(str.data, str.len);
1879 return *this;
1880 }
1881 wxString& append(const wchar_t *sz, size_t n)
1882 {
1883 SubstrBufFromWC str(ImplStr(sz, n));
1884 m_impl.append(str.data, str.len);
1885 return *this;
1886 }
1887
1888 wxString& append(const wxCStrData& str)
1889 { return append(str.AsString()); }
1890 wxString& append(const wxCharBuffer& str)
1891 { return append(str.data()); }
1892 wxString& append(const wxWCharBuffer& str)
1893 { return append(str.data()); }
1894
1895 // append n copies of ch
1896 wxString& append(size_t n, wxUniChar ch)
1897 {
1898#if wxUSE_UNICODE_UTF8
1899 if ( !ch.IsAscii() )
1900 m_impl.append(wxStringOperations::EncodeNChars(n, ch));
1901 else
1902#endif
1903 m_impl.append(n, (wxStringCharType)ch);
1904 return *this;
1905 }
1906 // append from first to last
1907 wxString& append(const_iterator first, const_iterator last)
1908 { m_impl.append(first.impl(), last.impl()); return *this; }
1909#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1910 wxString& append(const char *first, const char *last)
1911 { return append(first, last - first); }
1912 wxString& append(const wchar_t *first, const wchar_t *last)
1913 { return append(first, last - first); }
1914 wxString& append(const wxCStrData& first, const wxCStrData& last)
1915 { return append(CreateConstIterator(first), CreateConstIterator(last)); }
1916#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1917
1918 // same as `this_string = str'
1919 wxString& assign(const wxString& str)
1920 { m_impl = str.m_impl; return *this; }
1921 wxString& assign(const wxString& str, size_t len)
1922 {
1923 m_impl.assign(str.m_impl, 0, str.LenToImpl(len));
1924 return *this;
1925 }
1926 // same as ` = str[pos..pos + n]
1927 wxString& assign(const wxString& str, size_t pos, size_t n)
1928 {
1929 size_t from, len;
1930 str.PosLenToImpl(pos, n, &from, &len);
1931 m_impl.assign(str.m_impl, from, len);
1932 return *this;
1933 }
1934 // same as `= first n (or all if n == npos) characters of sz'
1935 wxString& assign(const char *sz)
1936 { m_impl.assign(ImplStr(sz)); return *this; }
1937 wxString& assign(const wchar_t *sz)
1938 { m_impl.assign(ImplStr(sz)); return *this; }
1939 wxString& assign(const char *sz, size_t n)
1940 {
1941 SubstrBufFromMB str(ImplStr(sz, n));
1942 m_impl.assign(str.data, str.len);
1943 return *this;
1944 }
1945 wxString& assign(const wchar_t *sz, size_t n)
1946 {
1947 SubstrBufFromWC str(ImplStr(sz, n));
1948 m_impl.assign(str.data, str.len);
1949 return *this;
1950 }
1951
1952 wxString& assign(const wxCStrData& str)
1953 { return assign(str.AsString()); }
1954 wxString& assign(const wxCharBuffer& str)
1955 { return assign(str.data()); }
1956 wxString& assign(const wxWCharBuffer& str)
1957 { return assign(str.data()); }
1958 wxString& assign(const wxCStrData& str, size_t len)
1959 { return assign(str.AsString(), len); }
1960 wxString& assign(const wxCharBuffer& str, size_t len)
1961 { return assign(str.data(), len); }
1962 wxString& assign(const wxWCharBuffer& str, size_t len)
1963 { return assign(str.data(), len); }
1964
1965 // same as `= n copies of ch'
1966 wxString& assign(size_t n, wxUniChar ch)
1967 {
1968#if wxUSE_UNICODE_UTF8
1969 if ( !ch.IsAscii() )
1970 m_impl.assign(wxStringOperations::EncodeNChars(n, ch));
1971 else
1972#endif
1973 m_impl.assign(n, (wxStringCharType)ch);
1974 return *this;
1975 }
1976
1977 wxString& assign(size_t n, wxUniCharRef ch)
1978 { return assign(n, wxUniChar(ch)); }
1979 wxString& assign(size_t n, char ch)
1980 { return assign(n, wxUniChar(ch)); }
1981 wxString& assign(size_t n, unsigned char ch)
1982 { return assign(n, wxUniChar(ch)); }
1983 wxString& assign(size_t n, wchar_t ch)
1984 { return assign(n, wxUniChar(ch)); }
1985
1986 // assign from first to last
1987 wxString& assign(const_iterator first, const_iterator last)
1988 { m_impl.assign(first.impl(), last.impl()); return *this; }
1989#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1990 wxString& assign(const char *first, const char *last)
1991 { return assign(first, last - first); }
1992 wxString& assign(const wchar_t *first, const wchar_t *last)
1993 { return assign(first, last - first); }
1994 wxString& assign(const wxCStrData& first, const wxCStrData& last)
1995 { return assign(CreateConstIterator(first), CreateConstIterator(last)); }
1996#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1997
1998 // string comparison
1999 int compare(const wxString& str) const;
2000 int compare(const char* sz) const;
2001 int compare(const wchar_t* sz) const;
2002 int compare(const wxCStrData& str) const
2003 { return compare(str.AsString()); }
2004 int compare(const wxCharBuffer& str) const
2005 { return compare(str.data()); }
2006 int compare(const wxWCharBuffer& str) const
2007 { return compare(str.data()); }
2008 // comparison with a substring
2009 int compare(size_t nStart, size_t nLen, const wxString& str) const;
2010 // comparison of 2 substrings
2011 int compare(size_t nStart, size_t nLen,
2012 const wxString& str, size_t nStart2, size_t nLen2) const;
2013 // substring comparison with first nCount characters of sz
2014 int compare(size_t nStart, size_t nLen,
2015 const char* sz, size_t nCount = npos) const;
2016 int compare(size_t nStart, size_t nLen,
2017 const wchar_t* sz, size_t nCount = npos) const;
2018
2019 // insert another string
2020 wxString& insert(size_t nPos, const wxString& str)
2021 { insert(begin() + nPos, str.begin(), str.end()); return *this; }
2022 // insert n chars of str starting at nStart (in str)
2023 wxString& insert(size_t nPos, const wxString& str, size_t nStart, size_t n)
2024 {
2025 size_t from, len;
2026 str.PosLenToImpl(nStart, n, &from, &len);
2027 m_impl.insert(PosToImpl(nPos), str.m_impl, from, len);
2028 return *this;
2029 }
2030 // insert first n (or all if n == npos) characters of sz
2031 wxString& insert(size_t nPos, const char *sz)
2032 { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; }
2033 wxString& insert(size_t nPos, const wchar_t *sz)
2034 { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; }
2035 wxString& insert(size_t nPos, const char *sz, size_t n)
2036 {
2037 SubstrBufFromMB str(ImplStr(sz, n));
2038 m_impl.insert(PosToImpl(nPos), str.data, str.len);
2039 return *this;
2040 }
2041 wxString& insert(size_t nPos, const wchar_t *sz, size_t n)
2042 {
2043 SubstrBufFromWC str(ImplStr(sz, n));
2044 m_impl.insert(PosToImpl(nPos), str.data, str.len);
2045 return *this;
2046 }
2047 // insert n copies of ch
2048 wxString& insert(size_t nPos, size_t n, wxUniChar ch)
2049 {
2050#if wxUSE_UNICODE_UTF8
2051 if ( !ch.IsAscii() )
2052 m_impl.insert(PosToImpl(nPos), wxStringOperations::EncodeNChars(n, ch));
2053 else
2054#endif
2055 m_impl.insert(PosToImpl(nPos), n, (wxStringCharType)ch);
2056 return *this;
2057 }
2058 iterator insert(iterator it, wxUniChar ch)
2059 {
2060#if wxUSE_UNICODE_UTF8
2061 if ( !ch.IsAscii() )
2062 {
2063 size_t pos = IterToImplPos(it);
2064 m_impl.insert(pos, wxStringOperations::EncodeChar(ch));
2065 return iterator(this, m_impl.begin() + pos);
2066 }
2067 else
2068#endif
2069 return iterator(this, m_impl.insert(it.impl(), (wxStringCharType)ch));
2070 }
2071 void insert(iterator it, const_iterator first, const_iterator last)
2072 { m_impl.insert(it.impl(), first.impl(), last.impl()); }
2073#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2074 void insert(iterator it, const char *first, const char *last)
2075 { insert(it - begin(), first, last - first); }
2076 void insert(iterator it, const wchar_t *first, const wchar_t *last)
2077 { insert(it - begin(), first, last - first); }
2078 void insert(iterator it, const wxCStrData& first, const wxCStrData& last)
2079 { insert(it, CreateConstIterator(first), CreateConstIterator(last)); }
2080#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2081
2082 void insert(iterator it, size_type n, wxUniChar ch)
2083 {
2084#if wxUSE_UNICODE_UTF8
2085 if ( !ch.IsAscii() )
2086 m_impl.insert(IterToImplPos(it), wxStringOperations::EncodeNChars(n, ch));
2087 else
2088#endif
2089 m_impl.insert(it.impl(), n, (wxStringCharType)ch);
2090 }
2091
2092 // delete characters from nStart to nStart + nLen
2093 wxString& erase(size_type pos = 0, size_type n = npos)
2094 {
2095 size_t from, len;
2096 PosLenToImpl(pos, n, &from, &len);
2097 m_impl.erase(from, len);
2098 return *this;
2099 }
2100 // delete characters from first up to last
2101 iterator erase(iterator first, iterator last)
2102 { return iterator(this, m_impl.erase(first.impl(), last.impl())); }
2103 iterator erase(iterator first)
2104 { return iterator(this, m_impl.erase(first.impl())); }
2105
2106#ifdef wxSTRING_BASE_HASNT_CLEAR
2107 void clear() { erase(); }
2108#else
2109 void clear() { m_impl.clear(); }
2110#endif
2111
2112 // replaces the substring of length nLen starting at nStart
2113 wxString& replace(size_t nStart, size_t nLen, const char* sz)
2114 {
2115 size_t from, len;
2116 PosLenToImpl(nStart, nLen, &from, &len);
2117 m_impl.replace(from, len, ImplStr(sz));
2118 return *this;
2119 }
2120 wxString& replace(size_t nStart, size_t nLen, const wchar_t* sz)
2121 {
2122 size_t from, len;
2123 PosLenToImpl(nStart, nLen, &from, &len);
2124 m_impl.replace(from, len, ImplStr(sz));
2125 return *this;
2126 }
2127 // replaces the substring of length nLen starting at nStart
2128 wxString& replace(size_t nStart, size_t nLen, const wxString& str)
2129 {
2130 size_t from, len;
2131 PosLenToImpl(nStart, nLen, &from, &len);
2132 m_impl.replace(from, len, str.m_impl);
2133 return *this;
2134 }
2135 // replaces the substring with nCount copies of ch
2136 wxString& replace(size_t nStart, size_t nLen, size_t nCount, wxUniChar ch)
2137 {
2138 size_t from, len;
2139 PosLenToImpl(nStart, nLen, &from, &len);
2140#if wxUSE_UNICODE_UTF8
2141 if ( !ch.IsAscii() )
2142 m_impl.replace(from, len, wxStringOperations::EncodeNChars(nCount, ch));
2143 else
2144#endif
2145 m_impl.replace(from, len, nCount, (wxStringCharType)ch);
2146 return *this;
2147 }
2148 // replaces a substring with another substring
2149 wxString& replace(size_t nStart, size_t nLen,
2150 const wxString& str, size_t nStart2, size_t nLen2)
2151 {
2152 size_t from, len;
2153 PosLenToImpl(nStart, nLen, &from, &len);
2154
2155 size_t from2, len2;
2156 str.PosLenToImpl(nStart2, nLen2, &from2, &len2);
2157
2158 m_impl.replace(from, len, str.m_impl, from2, len2);
2159 return *this;
2160 }
2161 // replaces the substring with first nCount chars of sz
2162 wxString& replace(size_t nStart, size_t nLen,
2163 const char* sz, size_t nCount)
2164 {
2165 size_t from, len;
2166 PosLenToImpl(nStart, nLen, &from, &len);
2167
2168 SubstrBufFromMB str(ImplStr(sz, nCount));
2169
2170 m_impl.replace(from, len, str.data, str.len);
2171 return *this;
2172 }
2173 wxString& replace(size_t nStart, size_t nLen,
2174 const wchar_t* sz, size_t nCount)
2175 {
2176 size_t from, len;
2177 PosLenToImpl(nStart, nLen, &from, &len);
2178
2179 SubstrBufFromWC str(ImplStr(sz, nCount));
2180
2181 m_impl.replace(from, len, str.data, str.len);
2182 return *this;
2183 }
2184 wxString& replace(size_t nStart, size_t nLen,
2185 const wxString& s, size_t nCount)
2186 {
2187 size_t from, len;
2188 PosLenToImpl(nStart, nLen, &from, &len);
2189 m_impl.replace(from, len, s.m_impl.c_str(), s.LenToImpl(nCount));
2190 return *this;
2191 }
2192
2193 wxString& replace(iterator first, iterator last, const char* s)
2194 { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; }
2195 wxString& replace(iterator first, iterator last, const wchar_t* s)
2196 { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; }
2197 wxString& replace(iterator first, iterator last, const char* s, size_type n)
2198 {
2199 SubstrBufFromMB str(ImplStr(s, n));
2200 m_impl.replace(first.impl(), last.impl(), str.data, str.len);
2201 return *this;
2202 }
2203 wxString& replace(iterator first, iterator last, const wchar_t* s, size_type n)
2204 {
2205 SubstrBufFromWC str(ImplStr(s, n));
2206 m_impl.replace(first.impl(), last.impl(), str.data, str.len);
2207 return *this;
2208 }
2209 wxString& replace(iterator first, iterator last, const wxString& s)
2210 { m_impl.replace(first.impl(), last.impl(), s.m_impl); return *this; }
2211 wxString& replace(iterator first, iterator last, size_type n, wxUniChar ch)
2212 {
2213#if wxUSE_UNICODE_UTF8
2214 if ( !ch.IsAscii() )
2215 m_impl.replace(first.impl(), last.impl(),
2216 wxStringOperations::EncodeNChars(n, ch));
2217 else
2218#endif
2219 m_impl.replace(first.impl(), last.impl(), n, (wxStringCharType)ch);
2220 return *this;
2221 }
2222 wxString& replace(iterator first, iterator last,
2223 const_iterator first1, const_iterator last1)
2224 {
2225 m_impl.replace(first.impl(), last.impl(), first1.impl(), last1.impl());
2226 return *this;
2227 }
2228 wxString& replace(iterator first, iterator last,
2229 const char *first1, const char *last1)
2230 { replace(first, last, first1, last1 - first1); return *this; }
2231 wxString& replace(iterator first, iterator last,
2232 const wchar_t *first1, const wchar_t *last1)
2233 { replace(first, last, first1, last1 - first1); return *this; }
2234
2235 // swap two strings
2236 void swap(wxString& str)
2237 { m_impl.swap(str.m_impl); }
2238
2239 // find a substring
2240 size_t find(const wxString& str, size_t nStart = 0) const
2241 { return PosFromImpl(m_impl.find(str.m_impl, PosToImpl(nStart))); }
2242
2243 // find first n characters of sz
2244 size_t find(const char* sz, size_t nStart = 0, size_t n = npos) const
2245 {
2246 SubstrBufFromMB str(ImplStr(sz, n));
2247 return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len));
2248 }
2249 size_t find(const wchar_t* sz, size_t nStart = 0, size_t n = npos) const
2250 {
2251 SubstrBufFromWC str(ImplStr(sz, n));
2252 return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len));
2253 }
2254 size_t find(const wxCharBuffer& s, size_t nStart = 0, size_t n = npos) const
2255 { return find(s.data(), nStart, n); }
2256 size_t find(const wxWCharBuffer& s, size_t nStart = 0, size_t n = npos) const
2257 { return find(s.data(), nStart, n); }
2258 size_t find(const wxCStrData& s, size_t nStart = 0, size_t n = npos) const
2259 { return find(s.AsWChar(), nStart, n); }
2260
2261 // find the first occurence of character ch after nStart
2262 size_t find(wxUniChar ch, size_t nStart = 0) const
2263 {
2264 return PosFromImpl(m_impl.find(wxStringOperations::EncodeChar(ch),
2265 PosToImpl(nStart)));
2266 }
2267 size_t find(wxUniCharRef ch, size_t nStart = 0) const
2268 { return find(wxUniChar(ch), nStart); }
2269 size_t find(char ch, size_t nStart = 0) const
2270 { return find(wxUniChar(ch), nStart); }
2271 size_t find(unsigned char ch, size_t nStart = 0) const
2272 { return find(wxUniChar(ch), nStart); }
2273 size_t find(wchar_t ch, size_t nStart = 0) const
2274 { return find(wxUniChar(ch), nStart); }
2275
2276 // rfind() family is exactly like find() but works right to left
2277
2278 // as find, but from the end
2279 size_t rfind(const wxString& str, size_t nStart = npos) const
2280 { return PosFromImpl(m_impl.rfind(str.m_impl, PosToImpl(nStart))); }
2281
2282 // as find, but from the end
2283 size_t rfind(const char* sz, size_t nStart = npos, size_t n = npos) const
2284 {
2285 SubstrBufFromMB str(ImplStr(sz, n));
2286 return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len));
2287 }
2288 size_t rfind(const wchar_t* sz, size_t nStart = npos, size_t n = npos) const
2289 {
2290 SubstrBufFromWC str(ImplStr(sz, n));
2291 return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len));
2292 }
2293 size_t rfind(const wxCharBuffer& s, size_t nStart = npos, size_t n = npos) const
2294 { return rfind(s.data(), nStart, n); }
2295 size_t rfind(const wxWCharBuffer& s, size_t nStart = npos, size_t n = npos) const
2296 { return rfind(s.data(), nStart, n); }
2297 size_t rfind(const wxCStrData& s, size_t nStart = npos, size_t n = npos) const
2298 { return rfind(s.AsWChar(), nStart, n); }
2299 // as find, but from the end
2300 size_t rfind(wxUniChar ch, size_t nStart = npos) const
2301 {
2302 return PosFromImpl(m_impl.rfind(wxStringOperations::EncodeChar(ch),
2303 PosToImpl(nStart)));
2304 }
2305 size_t rfind(wxUniCharRef ch, size_t nStart = npos) const
2306 { return rfind(wxUniChar(ch), nStart); }
2307 size_t rfind(char ch, size_t nStart = npos) const
2308 { return rfind(wxUniChar(ch), nStart); }
2309 size_t rfind(unsigned char ch, size_t nStart = npos) const
2310 { return rfind(wxUniChar(ch), nStart); }
2311 size_t rfind(wchar_t ch, size_t nStart = npos) const
2312 { return rfind(wxUniChar(ch), nStart); }
2313
2314 // find first/last occurence of any character (not) in the set:
2315#if wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2316 // FIXME-UTF8: this is not entirely correct, because it doesn't work if
2317 // sizeof(wchar_t)==2 and surrogates are present in the string;
2318 // should we care? Probably not.
2319 size_t find_first_of(const wxString& str, size_t nStart = 0) const
2320 { return m_impl.find_first_of(str.m_impl, nStart); }
2321 size_t find_first_of(const char* sz, size_t nStart = 0) const
2322 { return m_impl.find_first_of(ImplStr(sz), nStart); }
2323 size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const
2324 { return m_impl.find_first_of(ImplStr(sz), nStart); }
2325 size_t find_first_of(const char* sz, size_t nStart, size_t n) const
2326 { return m_impl.find_first_of(ImplStr(sz), nStart, n); }
2327 size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const
2328 { return m_impl.find_first_of(ImplStr(sz), nStart, n); }
2329 size_t find_first_of(wxUniChar c, size_t nStart = 0) const
2330 { return m_impl.find_first_of((wxChar)c, nStart); }
2331
2332 size_t find_last_of(const wxString& str, size_t nStart = npos) const
2333 { return m_impl.find_last_of(str.m_impl, nStart); }
2334 size_t find_last_of(const char* sz, size_t nStart = npos) const
2335 { return m_impl.find_last_of(ImplStr(sz), nStart); }
2336 size_t find_last_of(const wchar_t* sz, size_t nStart = npos) const
2337 { return m_impl.find_last_of(ImplStr(sz), nStart); }
2338 size_t find_last_of(const char* sz, size_t nStart, size_t n) const
2339 { return m_impl.find_last_of(ImplStr(sz), nStart, n); }
2340 size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const
2341 { return m_impl.find_last_of(ImplStr(sz), nStart, n); }
2342 size_t find_last_of(wxUniChar c, size_t nStart = npos) const
2343 { return m_impl.find_last_of((wxChar)c, nStart); }
2344
2345 size_t find_first_not_of(const wxString& str, size_t nStart = 0) const
2346 { return m_impl.find_first_not_of(str.m_impl, nStart); }
2347 size_t find_first_not_of(const char* sz, size_t nStart = 0) const
2348 { return m_impl.find_first_not_of(ImplStr(sz), nStart); }
2349 size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const
2350 { return m_impl.find_first_not_of(ImplStr(sz), nStart); }
2351 size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const
2352 { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); }
2353 size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const
2354 { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); }
2355 size_t find_first_not_of(wxUniChar c, size_t nStart = 0) const
2356 { return m_impl.find_first_not_of((wxChar)c, nStart); }
2357
2358 size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
2359 { return m_impl.find_last_not_of(str.m_impl, nStart); }
2360 size_t find_last_not_of(const char* sz, size_t nStart = npos) const
2361 { return m_impl.find_last_not_of(ImplStr(sz), nStart); }
2362 size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const
2363 { return m_impl.find_last_not_of(ImplStr(sz), nStart); }
2364 size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const
2365 { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); }
2366 size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const
2367 { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); }
2368 size_t find_last_not_of(wxUniChar c, size_t nStart = npos) const
2369 { return m_impl.find_last_not_of((wxChar)c, nStart); }
2370#else
2371 // we can't use std::string implementation in UTF-8 build, because the
2372 // character sets would be interpreted wrongly:
2373
2374 // as strpbrk() but starts at nStart, returns npos if not found
2375 size_t find_first_of(const wxString& str, size_t nStart = 0) const
2376#if wxUSE_UNICODE // FIXME-UTF8: temporary
2377 { return find_first_of(str.wc_str(), nStart); }
2378#else
2379 { return find_first_of(str.mb_str(), nStart); }
2380#endif
2381 // same as above
2382 size_t find_first_of(const char* sz, size_t nStart = 0) const;
2383 size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const;
2384 size_t find_first_of(const char* sz, size_t nStart, size_t n) const;
2385 size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const;
2386 // same as find(char, size_t)
2387 size_t find_first_of(wxUniChar c, size_t nStart = 0) const
2388 { return find(c, nStart); }
2389 // find the last (starting from nStart) char from str in this string
2390 size_t find_last_of (const wxString& str, size_t nStart = npos) const
2391#if wxUSE_UNICODE // FIXME-UTF8: temporary
2392 { return find_last_of(str.wc_str(), nStart); }
2393#else
2394 { return find_last_of(str.mb_str(), nStart); }
2395#endif
2396 // same as above
2397 size_t find_last_of (const char* sz, size_t nStart = npos) const;
2398 size_t find_last_of (const wchar_t* sz, size_t nStart = npos) const;
2399 size_t find_last_of(const char* sz, size_t nStart, size_t n) const;
2400 size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const;
2401 // same as above
2402 size_t find_last_of(wxUniChar c, size_t nStart = npos) const
2403 { return rfind(c, nStart); }
2404
2405 // find first/last occurence of any character not in the set
2406
2407 // as strspn() (starting from nStart), returns npos on failure
2408 size_t find_first_not_of(const wxString& str, size_t nStart = 0) const
2409#if wxUSE_UNICODE // FIXME-UTF8: temporary
2410 { return find_first_not_of(str.wc_str(), nStart); }
2411#else
2412 { return find_first_not_of(str.mb_str(), nStart); }
2413#endif
2414 // same as above
2415 size_t find_first_not_of(const char* sz, size_t nStart = 0) const;
2416 size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const;
2417 size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const;
2418 size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const;
2419 // same as above
2420 size_t find_first_not_of(wxUniChar ch, size_t nStart = 0) const;
2421 // as strcspn()
2422 size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
2423#if wxUSE_UNICODE // FIXME-UTF8: temporary
2424 { return find_last_not_of(str.wc_str(), nStart); }
2425#else
2426 { return find_last_not_of(str.mb_str(), nStart); }
2427#endif
2428 // same as above
2429 size_t find_last_not_of(const char* sz, size_t nStart = npos) const;
2430 size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const;
2431 size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const;
2432 size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const;
2433 // same as above
2434 size_t find_last_not_of(wxUniChar ch, size_t nStart = npos) const;
2435#endif // wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 or not
2436
2437 // provide char/wchar_t/wxUniCharRef overloads for char-finding functions
2438 // above to resolve ambiguities:
2439 size_t find_first_of(wxUniCharRef ch, size_t nStart = 0) const
2440 { return find_first_of(wxUniChar(ch), nStart); }
2441 size_t find_first_of(char ch, size_t nStart = 0) const
2442 { return find_first_of(wxUniChar(ch), nStart); }
2443 size_t find_first_of(unsigned char ch, size_t nStart = 0) const
2444 { return find_first_of(wxUniChar(ch), nStart); }
2445 size_t find_first_of(wchar_t ch, size_t nStart = 0) const
2446 { return find_first_of(wxUniChar(ch), nStart); }
2447 size_t find_last_of(wxUniCharRef ch, size_t nStart = npos) const
2448 { return find_last_of(wxUniChar(ch), nStart); }
2449 size_t find_last_of(char ch, size_t nStart = npos) const
2450 { return find_last_of(wxUniChar(ch), nStart); }
2451 size_t find_last_of(unsigned char ch, size_t nStart = npos) const
2452 { return find_last_of(wxUniChar(ch), nStart); }
2453 size_t find_last_of(wchar_t ch, size_t nStart = npos) const
2454 { return find_last_of(wxUniChar(ch), nStart); }
2455 size_t find_first_not_of(wxUniCharRef ch, size_t nStart = 0) const
2456 { return find_first_not_of(wxUniChar(ch), nStart); }
2457 size_t find_first_not_of(char ch, size_t nStart = 0) const
2458 { return find_first_not_of(wxUniChar(ch), nStart); }
2459 size_t find_first_not_of(unsigned char ch, size_t nStart = 0) const
2460 { return find_first_not_of(wxUniChar(ch), nStart); }
2461 size_t find_first_not_of(wchar_t ch, size_t nStart = 0) const
2462 { return find_first_not_of(wxUniChar(ch), nStart); }
2463 size_t find_last_not_of(wxUniCharRef ch, size_t nStart = npos) const
2464 { return find_last_not_of(wxUniChar(ch), nStart); }
2465 size_t find_last_not_of(char ch, size_t nStart = npos) const
2466 { return find_last_not_of(wxUniChar(ch), nStart); }
2467 size_t find_last_not_of(unsigned char ch, size_t nStart = npos) const
2468 { return find_last_not_of(wxUniChar(ch), nStart); }
2469 size_t find_last_not_of(wchar_t ch, size_t nStart = npos) const
2470 { return find_last_not_of(wxUniChar(ch), nStart); }
2471
2472 // and additional overloads for the versions taking strings:
2473 size_t find_first_of(const wxCStrData& sz, size_t nStart = 0) const
2474 { return find_first_of(sz.AsString(), nStart); }
2475 size_t find_first_of(const wxCharBuffer& sz, size_t nStart = 0) const
2476 { return find_first_of(sz.data(), nStart); }
2477 size_t find_first_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2478 { return find_first_of(sz.data(), nStart); }
2479 size_t find_first_of(const wxCStrData& sz, size_t nStart, size_t n) const
2480 { return find_first_of(sz.AsWChar(), nStart, n); }
2481 size_t find_first_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2482 { return find_first_of(sz.data(), nStart, n); }
2483 size_t find_first_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2484 { return find_first_of(sz.data(), nStart, n); }
2485
2486 size_t find_last_of(const wxCStrData& sz, size_t nStart = 0) const
2487 { return find_last_of(sz.AsString(), nStart); }
2488 size_t find_last_of(const wxCharBuffer& sz, size_t nStart = 0) const
2489 { return find_last_of(sz.data(), nStart); }
2490 size_t find_last_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2491 { return find_last_of(sz.data(), nStart); }
2492 size_t find_last_of(const wxCStrData& sz, size_t nStart, size_t n) const
2493 { return find_last_of(sz.AsWChar(), nStart, n); }
2494 size_t find_last_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2495 { return find_last_of(sz.data(), nStart, n); }
2496 size_t find_last_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2497 { return find_last_of(sz.data(), nStart, n); }
2498
2499 size_t find_first_not_of(const wxCStrData& sz, size_t nStart = 0) const
2500 { return find_first_not_of(sz.AsString(), nStart); }
2501 size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart = 0) const
2502 { return find_first_not_of(sz.data(), nStart); }
2503 size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2504 { return find_first_not_of(sz.data(), nStart); }
2505 size_t find_first_not_of(const wxCStrData& sz, size_t nStart, size_t n) const
2506 { return find_first_not_of(sz.AsWChar(), nStart, n); }
2507 size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2508 { return find_first_not_of(sz.data(), nStart, n); }
2509 size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2510 { return find_first_not_of(sz.data(), nStart, n); }
2511
2512 size_t find_last_not_of(const wxCStrData& sz, size_t nStart = 0) const
2513 { return find_last_not_of(sz.AsString(), nStart); }
2514 size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart = 0) const
2515 { return find_last_not_of(sz.data(), nStart); }
2516 size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2517 { return find_last_not_of(sz.data(), nStart); }
2518 size_t find_last_not_of(const wxCStrData& sz, size_t nStart, size_t n) const
2519 { return find_last_not_of(sz.AsWChar(), nStart, n); }
2520 size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2521 { return find_last_not_of(sz.data(), nStart, n); }
2522 size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2523 { return find_last_not_of(sz.data(), nStart, n); }
2524
2525 // string += string
2526 wxString& operator+=(const wxString& s)
2527 { m_impl += s.m_impl; return *this; }
2528 // string += C string
2529 wxString& operator+=(const char *psz)
2530 { m_impl += ImplStr(psz); return *this; }
2531 wxString& operator+=(const wchar_t *pwz)
2532 { m_impl += ImplStr(pwz); return *this; }
2533 wxString& operator+=(const wxCStrData& s)
2534 { m_impl += s.AsString().m_impl; return *this; }
2535 wxString& operator+=(const wxCharBuffer& s)
2536 { return operator+=(s.data()); }
2537 wxString& operator+=(const wxWCharBuffer& s)
2538 { return operator+=(s.data()); }
2539 // string += char
2540 wxString& operator+=(wxUniChar ch)
2541 { m_impl += wxStringOperations::EncodeChar(ch); return *this; }
2542 wxString& operator+=(wxUniCharRef ch) { return *this += wxUniChar(ch); }
2543 wxString& operator+=(int ch) { return *this += wxUniChar(ch); }
2544 wxString& operator+=(char ch) { return *this += wxUniChar(ch); }
2545 wxString& operator+=(unsigned char ch) { return *this += wxUniChar(ch); }
2546 wxString& operator+=(wchar_t ch) { return *this += wxUniChar(ch); }
2547
2548private:
2549#if !wxUSE_STL_BASED_WXSTRING
2550 // helpers for wxStringBuffer and wxStringBufferLength
2551 wxStringCharType *DoGetWriteBuf(size_t nLen)
2552 { return m_impl.DoGetWriteBuf(nLen); }
2553 void DoUngetWriteBuf()
2554 { m_impl.DoUngetWriteBuf(); }
2555 void DoUngetWriteBuf(size_t nLen)
2556 { m_impl.DoUngetWriteBuf(nLen); }
2557#endif // !wxUSE_STL_BASED_WXSTRING
2558
2559#ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
2560 #if !wxUSE_UTF8_LOCALE_ONLY
2561 int DoPrintfWchar(const wxChar *format, ...);
2562 static wxString DoFormatWchar(const wxChar *format, ...);
2563 #endif
2564 #if wxUSE_UNICODE_UTF8
2565 int DoPrintfUtf8(const char *format, ...);
2566 static wxString DoFormatUtf8(const char *format, ...);
2567 #endif
2568#endif
2569
2570#if !wxUSE_STL_BASED_WXSTRING
2571 // check string's data validity
2572 bool IsValid() const { return m_impl.GetStringData()->IsValid(); }
2573#endif
2574
2575private:
2576 wxStringImpl m_impl;
2577
2578#ifdef __VISUALC__
2579 // "struct 'ConvertedBuffer<T>' needs to have dll-interface to be used by
2580 // clients of class 'wxString'" - this is private, we don't care
2581 #pragma warning (disable:4251)
2582#endif
2583
2584 // buffers for compatibility conversion from (char*)c_str() and
2585 // (wchar_t*)c_str():
2586 // FIXME-UTF8: bechmark various approaches to keeping compatibility buffers
2587 template<typename T>
2588 struct ConvertedBuffer
2589 {
2590 ConvertedBuffer() : m_buf(NULL) {}
2591 ~ConvertedBuffer()
2592 { free(m_buf); }
2593
2594 operator T*() const { return m_buf; }
2595
2596 ConvertedBuffer& operator=(T *str)
2597 {
2598 free(m_buf);
2599 m_buf = str;
2600 return *this;
2601 }
2602
2603 T *m_buf;
2604 };
2605#if wxUSE_UNICODE && !wxUSE_UTF8_LOCALE_ONLY
2606 ConvertedBuffer<char> m_convertedToChar;
2607#endif
2608#if !wxUSE_UNICODE_WCHAR
2609 ConvertedBuffer<wchar_t> m_convertedToWChar;
2610#endif
2611
2612#ifdef __VISUALC__
2613 #pragma warning (default:4251)
2614#endif
2615
2616#if wxUSE_UNICODE_UTF8
2617 // FIXME-UTF8: (try to) move this elsewhere (TLS) or solve differently
2618 // assigning to character pointer to by wxString::interator may
2619 // change the underlying wxStringImpl iterator, so we have to
2620 // keep track of all iterators and update them as necessary:
2621 struct wxStringIteratorNodeHead
2622 {
2623 wxStringIteratorNodeHead() : ptr(NULL) {}
2624 wxStringIteratorNode *ptr;
2625
2626 // copying is disallowed as it would result in more than one pointer into
2627 // the same linked list
2628 DECLARE_NO_COPY_CLASS(wxStringIteratorNodeHead)
2629 };
2630
2631 wxStringIteratorNodeHead m_iterators;
2632
2633 friend class WXDLLIMPEXP_FWD_BASE wxStringIteratorNode;
2634 friend class WXDLLIMPEXP_FWD_BASE wxUniCharRef;
2635#endif // wxUSE_UNICODE_UTF8
2636
2637 friend class WXDLLIMPEXP_FWD_BASE wxCStrData;
2638 friend class wxImplStringBuffer;
2639 friend class wxImplStringBufferLength;
2640};
2641
2642#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
2643 #pragma warning (default:4275)
2644#endif
2645
2646// string iterator operators that satisfy STL Random Access Iterator
2647// requirements:
2648inline wxString::iterator operator+(int n, wxString::iterator i)
2649 { return i + n; }
2650inline wxString::iterator operator+(size_t n, wxString::iterator i)
2651 { return i + n; }
2652inline wxString::const_iterator operator+(int n, wxString::const_iterator i)
2653 { return i + n; }
2654inline wxString::const_iterator operator+(size_t n, wxString::const_iterator i)
2655 { return i + n; }
2656inline wxString::reverse_iterator operator+(int n, wxString::reverse_iterator i)
2657 { return i + n; }
2658inline wxString::reverse_iterator operator+(size_t n, wxString::reverse_iterator i)
2659 { return i + n; }
2660inline wxString::const_reverse_iterator operator+(int n, wxString::const_reverse_iterator i)
2661 { return i + n; }
2662inline wxString::const_reverse_iterator operator+(size_t n, wxString::const_reverse_iterator i)
2663 { return i + n; }
2664
2665// notice that even though for many compilers the friend declarations above are
2666// enough, from the point of view of C++ standard we must have the declarations
2667// here as friend ones are not injected in the enclosing namespace and without
2668// them the code fails to compile with conforming compilers such as xlC or g++4
2669wxString WXDLLIMPEXP_BASE operator+(const wxString& string1, const wxString& string2);
2670wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const char *psz);
2671wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const wchar_t *pwz);
2672wxString WXDLLIMPEXP_BASE operator+(const char *psz, const wxString& string);
2673wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz, const wxString& string);
2674
2675wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxUniChar ch);
2676wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string);
2677
2678inline wxString operator+(const wxString& string, wxUniCharRef ch)
2679 { return string + (wxUniChar)ch; }
2680inline wxString operator+(const wxString& string, char ch)
2681 { return string + wxUniChar(ch); }
2682inline wxString operator+(const wxString& string, wchar_t ch)
2683 { return string + wxUniChar(ch); }
2684inline wxString operator+(wxUniCharRef ch, const wxString& string)
2685 { return (wxUniChar)ch + string; }
2686inline wxString operator+(char ch, const wxString& string)
2687 { return wxUniChar(ch) + string; }
2688inline wxString operator+(wchar_t ch, const wxString& string)
2689 { return wxUniChar(ch) + string; }
2690
2691
2692#if wxUSE_STL_BASED_WXSTRING
2693 // return an empty wxString (not very useful with wxUSE_STL == 1)
2694 inline const wxString wxGetEmptyString() { return wxString(); }
2695#else // !wxUSE_STL_BASED_WXSTRING
2696 // return an empty wxString (more efficient than wxString() here)
2697 inline const wxString& wxGetEmptyString()
2698 {
2699 return *(wxString *)&wxEmptyString;
2700 }
2701#endif // wxUSE_STL_BASED_WXSTRING/!wxUSE_STL_BASED_WXSTRING
2702
2703// ----------------------------------------------------------------------------
2704// wxStringBuffer: a tiny class allowing to get a writable pointer into string
2705// ----------------------------------------------------------------------------
2706
2707#if !wxUSE_STL_BASED_WXSTRING
2708// string buffer for direct access to string data in their native
2709// representation:
2710class wxImplStringBuffer
2711{
2712public:
2713 typedef wxStringCharType CharType;
2714
2715 wxImplStringBuffer(wxString& str, size_t lenWanted = 1024)
2716 : m_str(str), m_buf(NULL)
2717 { m_buf = m_str.DoGetWriteBuf(lenWanted); }
2718
2719 ~wxImplStringBuffer() { m_str.DoUngetWriteBuf(); }
2720
2721 operator wxStringCharType*() const { return m_buf; }
2722
2723private:
2724 wxString& m_str;
2725 wxStringCharType *m_buf;
2726
2727 DECLARE_NO_COPY_CLASS(wxImplStringBuffer)
2728};
2729
2730class wxImplStringBufferLength
2731{
2732public:
2733 typedef wxStringCharType CharType;
2734
2735 wxImplStringBufferLength(wxString& str, size_t lenWanted = 1024)
2736 : m_str(str), m_buf(NULL), m_len(0), m_lenSet(false)
2737 {
2738 m_buf = m_str.DoGetWriteBuf(lenWanted);
2739 wxASSERT(m_buf != NULL);
2740 }
2741
2742 ~wxImplStringBufferLength()
2743 {
2744 wxASSERT(m_lenSet);
2745 m_str.DoUngetWriteBuf(m_len);
2746 }
2747
2748 operator wxStringCharType*() const { return m_buf; }
2749 void SetLength(size_t length) { m_len = length; m_lenSet = true; }
2750
2751private:
2752 wxString& m_str;
2753 wxStringCharType *m_buf;
2754 size_t m_len;
2755 bool m_lenSet;
2756
2757 DECLARE_NO_COPY_CLASS(wxImplStringBufferLength)
2758};
2759
2760#endif // !wxUSE_STL_BASED_WXSTRING
2761
2762template<typename T>
2763class wxStringTypeBufferBase
2764{
2765public:
2766 typedef T CharType;
2767
2768 wxStringTypeBufferBase(wxString& str, size_t lenWanted = 1024)
2769 : m_str(str), m_buf(lenWanted)
2770 { }
2771
2772
2773 operator CharType*() { return m_buf.data(); }
2774
2775protected:
2776 wxString& m_str;
2777 wxCharTypeBuffer<CharType> m_buf;
2778};
2779
2780template<typename T>
2781class wxStringTypeBufferLengthBase
2782{
2783public:
2784 typedef T CharType;
2785
2786 wxStringTypeBufferLengthBase(wxString& str, size_t lenWanted = 1024)
2787 : m_str(str), m_buf(lenWanted), m_len(0), m_lenSet(false)
2788 { }
2789
2790 ~wxStringTypeBufferLengthBase()
2791 {
2792 wxASSERT(m_lenSet);
2793 m_str.assign(m_buf.data(), m_len);
2794 }
2795
2796 operator CharType*() { return m_buf.data(); }
2797 void SetLength(size_t length) { m_len = length; m_lenSet = true; }
2798
2799protected:
2800 wxString& m_str;
2801 wxCharTypeBuffer<CharType> m_buf;
2802 size_t m_len;
2803 bool m_lenSet;
2804};
2805
2806template<typename T>
2807class wxStringTypeBuffer : public wxStringTypeBufferBase<T>
2808{
2809public:
2810 wxStringTypeBuffer(wxString& str, size_t lenWanted = 1024)
2811 : wxStringTypeBufferBase<T>(str, lenWanted) {}
2812 ~wxStringTypeBuffer()
2813 {
2814 this->m_str.assign(this->m_buf.data());
2815 }
2816
2817 DECLARE_NO_COPY_CLASS(wxStringTypeBuffer)
2818};
2819
2820template<typename T>
2821class wxStringTypeBufferLength : public wxStringTypeBufferLengthBase<T>
2822{
2823public:
2824 wxStringTypeBufferLength(wxString& str, size_t lenWanted = 1024)
2825 : wxStringTypeBufferLengthBase<T>(str, lenWanted) {}
2826
2827 ~wxStringTypeBufferLength()
2828 {
2829 wxASSERT(this->m_lenSet);
2830 this->m_str.assign(this->m_buf.data(), this->m_len);
2831 }
2832
2833 DECLARE_NO_COPY_CLASS(wxStringTypeBufferLength)
2834};
2835
2836#if wxUSE_STL_BASED_WXSTRING
2837class wxImplStringBuffer : public wxStringTypeBufferBase<wxStringCharType>
2838{
2839public:
2840 wxImplStringBuffer(wxString& str, size_t lenWanted = 1024)
2841 : wxStringTypeBufferBase<wxStringCharType>(str, lenWanted) {}
2842 ~wxImplStringBuffer()
2843 { m_str.m_impl.assign(m_buf.data()); }
2844
2845 DECLARE_NO_COPY_CLASS(wxImplStringBuffer)
2846};
2847
2848class wxImplStringBufferLength : public wxStringTypeBufferLengthBase<wxStringCharType>
2849{
2850public:
2851 wxImplStringBufferLength(wxString& str, size_t lenWanted = 1024)
2852 : wxStringTypeBufferLengthBase<wxStringCharType>(str, lenWanted) {}
2853
2854 ~wxImplStringBufferLength()
2855 {
2856 wxASSERT(m_lenSet);
2857 m_str.m_impl.assign(m_buf.data(), m_len);
2858 }
2859
2860 DECLARE_NO_COPY_CLASS(wxImplStringBufferLength)
2861};
2862#endif // wxUSE_STL_BASED_WXSTRING
2863
2864
2865#if wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8
2866typedef wxStringTypeBuffer<wxChar> wxStringBuffer;
2867typedef wxStringTypeBufferLength<wxChar> wxStringBufferLength;
2868#else // if !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2869typedef wxImplStringBuffer wxStringBuffer;
2870typedef wxImplStringBufferLength wxStringBufferLength;
2871#endif // !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2872
2873// ---------------------------------------------------------------------------
2874// wxString comparison functions: operator versions are always case sensitive
2875// ---------------------------------------------------------------------------
2876
2877#define wxCMP_WXCHAR_STRING(p, s, op) 0 op s.Cmp(p)
2878
2879wxDEFINE_ALL_COMPARISONS(const wxChar *, const wxString&, wxCMP_WXCHAR_STRING)
2880
2881#undef wxCMP_WXCHAR_STRING
2882
2883// note that there is an optimization in operator==() and !=(): we (quickly)
2884// checks the strings length first, before comparing their data
2885inline bool operator==(const wxString& s1, const wxString& s2)
2886 { return (s1.Len() == s2.Len()) && (s1.Cmp(s2) == 0); }
2887inline bool operator!=(const wxString& s1, const wxString& s2)
2888 { return (s1.Len() != s2.Len()) || (s1.Cmp(s2) != 0); }
2889inline bool operator< (const wxString& s1, const wxString& s2)
2890 { return s1.Cmp(s2) < 0; }
2891inline bool operator> (const wxString& s1, const wxString& s2)
2892 { return s1.Cmp(s2) > 0; }
2893inline bool operator<=(const wxString& s1, const wxString& s2)
2894 { return s1.Cmp(s2) <= 0; }
2895inline bool operator>=(const wxString& s1, const wxString& s2)
2896 { return s1.Cmp(s2) >= 0; }
2897
2898inline bool operator==(const wxString& s1, const wxCStrData& s2)
2899 { return s1 == s2.AsString(); }
2900inline bool operator==(const wxCStrData& s1, const wxString& s2)
2901 { return s1.AsString() == s2; }
2902inline bool operator!=(const wxString& s1, const wxCStrData& s2)
2903 { return s1 != s2.AsString(); }
2904inline bool operator!=(const wxCStrData& s1, const wxString& s2)
2905 { return s1.AsString() != s2; }
2906
2907inline bool operator==(const wxString& s1, const wxWCharBuffer& s2)
2908 { return (s1.Cmp((const wchar_t *)s2) == 0); }
2909inline bool operator==(const wxWCharBuffer& s1, const wxString& s2)
2910 { return (s2.Cmp((const wchar_t *)s1) == 0); }
2911inline bool operator!=(const wxString& s1, const wxWCharBuffer& s2)
2912 { return (s1.Cmp((const wchar_t *)s2) != 0); }
2913inline bool operator!=(const wxWCharBuffer& s1, const wxString& s2)
2914 { return (s2.Cmp((const wchar_t *)s1) != 0); }
2915
2916inline bool operator==(const wxString& s1, const wxCharBuffer& s2)
2917 { return (s1.Cmp((const char *)s2) == 0); }
2918inline bool operator==(const wxCharBuffer& s1, const wxString& s2)
2919 { return (s2.Cmp((const char *)s1) == 0); }
2920inline bool operator!=(const wxString& s1, const wxCharBuffer& s2)
2921 { return (s1.Cmp((const char *)s2) != 0); }
2922inline bool operator!=(const wxCharBuffer& s1, const wxString& s2)
2923 { return (s2.Cmp((const char *)s1) != 0); }
2924
2925inline wxString operator+(const wxString& string, const wxWCharBuffer& buf)
2926 { return string + (const wchar_t *)buf; }
2927inline wxString operator+(const wxWCharBuffer& buf, const wxString& string)
2928 { return (const wchar_t *)buf + string; }
2929
2930inline wxString operator+(const wxString& string, const wxCharBuffer& buf)
2931 { return string + (const char *)buf; }
2932inline wxString operator+(const wxCharBuffer& buf, const wxString& string)
2933 { return (const char *)buf + string; }
2934
2935// comparison with char
2936inline bool operator==(const wxUniChar& c, const wxString& s) { return s.IsSameAs(c); }
2937inline bool operator==(const wxUniCharRef& c, const wxString& s) { return s.IsSameAs(c); }
2938inline bool operator==(char c, const wxString& s) { return s.IsSameAs(c); }
2939inline bool operator==(wchar_t c, const wxString& s) { return s.IsSameAs(c); }
2940inline bool operator==(int c, const wxString& s) { return s.IsSameAs(c); }
2941inline bool operator==(const wxString& s, const wxUniChar& c) { return s.IsSameAs(c); }
2942inline bool operator==(const wxString& s, const wxUniCharRef& c) { return s.IsSameAs(c); }
2943inline bool operator==(const wxString& s, char c) { return s.IsSameAs(c); }
2944inline bool operator==(const wxString& s, wchar_t c) { return s.IsSameAs(c); }
2945inline bool operator!=(const wxUniChar& c, const wxString& s) { return !s.IsSameAs(c); }
2946inline bool operator!=(const wxUniCharRef& c, const wxString& s) { return !s.IsSameAs(c); }
2947inline bool operator!=(char c, const wxString& s) { return !s.IsSameAs(c); }
2948inline bool operator!=(wchar_t c, const wxString& s) { return !s.IsSameAs(c); }
2949inline bool operator!=(int c, const wxString& s) { return !s.IsSameAs(c); }
2950inline bool operator!=(const wxString& s, const wxUniChar& c) { return !s.IsSameAs(c); }
2951inline bool operator!=(const wxString& s, const wxUniCharRef& c) { return !s.IsSameAs(c); }
2952inline bool operator!=(const wxString& s, char c) { return !s.IsSameAs(c); }
2953inline bool operator!=(const wxString& s, wchar_t c) { return !s.IsSameAs(c); }
2954
2955// comparison with C string in Unicode build
2956#if wxUSE_UNICODE
2957
2958#define wxCMP_CHAR_STRING(p, s, op) wxString(p) op s
2959
2960wxDEFINE_ALL_COMPARISONS(const char *, const wxString&, wxCMP_CHAR_STRING)
2961
2962#undef wxCMP_CHAR_STRING
2963
2964#endif // wxUSE_UNICODE
2965
2966// we also need to provide the operators for comparison with wxCStrData to
2967// resolve ambiguity between operator(const wxChar *,const wxString &) and
2968// operator(const wxChar *, const wxChar *) for "p == s.c_str()"
2969//
2970// notice that these are (shallow) pointer comparisons, not (deep) string ones
2971#define wxCMP_CHAR_CSTRDATA(p, s, op) p op s.AsChar()
2972#define wxCMP_WCHAR_CSTRDATA(p, s, op) p op s.AsWChar()
2973
2974wxDEFINE_ALL_COMPARISONS(const wchar_t *, const wxCStrData&, wxCMP_WCHAR_CSTRDATA)
2975wxDEFINE_ALL_COMPARISONS(const char *, const wxCStrData&, wxCMP_CHAR_CSTRDATA)
2976
2977#undef wxCMP_CHAR_CSTRDATA
2978#undef wxCMP_WCHAR_CSTRDATA
2979
2980// ---------------------------------------------------------------------------
2981// Implementation only from here until the end of file
2982// ---------------------------------------------------------------------------
2983
2984#if wxUSE_STD_IOSTREAM
2985
2986#include "wx/iosfwrap.h"
2987
2988WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxString&);
2989WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCStrData&);
2990WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCharBuffer&);
2991#ifndef __BORLANDC__
2992WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxWCharBuffer&);
2993#endif
2994
2995#endif // wxSTD_STRING_COMPATIBILITY
2996
2997// ---------------------------------------------------------------------------
2998// wxCStrData implementation
2999// ---------------------------------------------------------------------------
3000
3001inline wxCStrData::wxCStrData(char *buf)
3002 : m_str(new wxString(buf)), m_offset(0), m_owned(true) {}
3003inline wxCStrData::wxCStrData(wchar_t *buf)
3004 : m_str(new wxString(buf)), m_offset(0), m_owned(true) {}
3005
3006inline wxCStrData::wxCStrData(const wxCStrData& data)
3007 : m_str(data.m_owned ? new wxString(*data.m_str) : data.m_str),
3008 m_offset(data.m_offset),
3009 m_owned(data.m_owned)
3010{
3011}
3012
3013inline wxCStrData::~wxCStrData()
3014{
3015 if ( m_owned )
3016 delete wx_const_cast(wxString*, m_str); // cast to silence warnings
3017}
3018
3019// simple cases for AsChar() and AsWChar(), the complicated ones are
3020// in string.cpp
3021#if wxUSE_UNICODE_WCHAR
3022inline const wchar_t* wxCStrData::AsWChar() const
3023{
3024 return m_str->wx_str() + m_offset;
3025}
3026#endif // wxUSE_UNICODE_WCHAR
3027
3028#if !wxUSE_UNICODE
3029inline const char* wxCStrData::AsChar() const
3030{
3031 return m_str->wx_str() + m_offset;
3032}
3033#endif // !wxUSE_UNICODE
3034
3035#if wxUSE_UTF8_LOCALE_ONLY
3036inline const char* wxCStrData::AsChar() const
3037{
3038 return wxStringOperations::AddToIter(m_str->wx_str(), m_offset);
3039}
3040#endif // wxUSE_UTF8_LOCALE_ONLY
3041
3042inline const wxCharBuffer wxCStrData::AsCharBuf() const
3043{
3044#if !wxUSE_UNICODE
3045 return wxCharBuffer::CreateNonOwned(AsChar());
3046#else
3047 return AsString().mb_str();
3048#endif
3049}
3050
3051inline const wxWCharBuffer wxCStrData::AsWCharBuf() const
3052{
3053#if wxUSE_UNICODE_WCHAR
3054 return wxWCharBuffer::CreateNonOwned(AsWChar());
3055#else
3056 return AsString().wc_str();
3057#endif
3058}
3059
3060inline wxString wxCStrData::AsString() const
3061{
3062 if ( m_offset == 0 )
3063 return *m_str;
3064 else
3065 return m_str->Mid(m_offset);
3066}
3067
3068inline const wxStringCharType *wxCStrData::AsInternal() const
3069{
3070#if wxUSE_UNICODE_UTF8
3071 return wxStringOperations::AddToIter(m_str->wx_str(), m_offset);
3072#else
3073 return m_str->wx_str() + m_offset;
3074#endif
3075}
3076
3077inline wxUniChar wxCStrData::operator*() const
3078{
3079 if ( m_str->empty() )
3080 return wxUniChar(_T('\0'));
3081 else
3082 return (*m_str)[m_offset];
3083}
3084
3085inline wxUniChar wxCStrData::operator[](size_t n) const
3086{
3087 // NB: we intentionally use operator[] and not at() here because the former
3088 // works for the terminating NUL while the latter does not
3089 return (*m_str)[m_offset + n];
3090}
3091
3092// ----------------------------------------------------------------------------
3093// more wxCStrData operators
3094// ----------------------------------------------------------------------------
3095
3096// we need to define those to allow "size_t pos = p - s.c_str()" where p is
3097// some pointer into the string
3098inline size_t operator-(const char *p, const wxCStrData& cs)
3099{
3100 return p - cs.AsChar();
3101}
3102
3103inline size_t operator-(const wchar_t *p, const wxCStrData& cs)
3104{
3105 return p - cs.AsWChar();
3106}
3107
3108// ----------------------------------------------------------------------------
3109// implementation of wx[W]CharBuffer inline methods using wxCStrData
3110// ----------------------------------------------------------------------------
3111
3112// FIXME-UTF8: move this to buffer.h
3113inline wxCharBuffer::wxCharBuffer(const wxCStrData& cstr)
3114 : wxCharTypeBufferBase(cstr.AsCharBuf())
3115{
3116}
3117
3118inline wxWCharBuffer::wxWCharBuffer(const wxCStrData& cstr)
3119 : wxCharTypeBufferBase(cstr.AsWCharBuf())
3120{
3121}
3122
3123#if wxUSE_UNICODE_UTF8
3124// ----------------------------------------------------------------------------
3125// implementation of wxStringIteratorNode inline methods
3126// ----------------------------------------------------------------------------
3127
3128void wxStringIteratorNode::DoSet(const wxString *str,
3129 wxStringImpl::const_iterator *citer,
3130 wxStringImpl::iterator *iter)
3131{
3132 m_prev = NULL;
3133 m_iter = iter;
3134 m_citer = citer;
3135 m_str = str;
3136 if ( str )
3137 {
3138 m_next = str->m_iterators.ptr;
3139 wx_const_cast(wxString*, m_str)->m_iterators.ptr = this;
3140 if ( m_next )
3141 m_next->m_prev = this;
3142 }
3143 else
3144 {
3145 m_next = NULL;
3146 }
3147}
3148
3149void wxStringIteratorNode::clear()
3150{
3151 if ( m_next )
3152 m_next->m_prev = m_prev;
3153 if ( m_prev )
3154 m_prev->m_next = m_next;
3155 else if ( m_str ) // first in the list
3156 wx_const_cast(wxString*, m_str)->m_iterators.ptr = m_next;
3157
3158 m_next = m_prev = NULL;
3159 m_citer = NULL;
3160 m_iter = NULL;
3161 m_str = NULL;
3162}
3163#endif // wxUSE_UNICODE_UTF8
3164
3165#if WXWIN_COMPATIBILITY_2_8
3166 // lot of code out there doesn't explicitly include wx/crt.h, but uses
3167 // CRT wrappers that are now declared in wx/wxcrt.h and wx/wxcrtvararg.h,
3168 // so let's include this header now that wxString is defined and it's safe
3169 // to do it:
3170 #include "wx/crt.h"
3171#endif
3172
3173#endif // _WX_WXSTRING_H_