]> git.saurik.com Git - wxWidgets.git/blob - include/wx/string.h
take const wxConfig object in wxDocManager::FileHistoryLoad() and wxFileHistory:...
[wxWidgets.git] / include / wx / string.h
1 ///////////////////////////////////////////////////////////////////////////////
2 // Name: wx/string.h
3 // Purpose: wxString class
4 // Author: Vadim Zeitlin
5 // Modified by:
6 // Created: 29/01/98
7 // RCS-ID: $Id$
8 // Copyright: (c) 1998 Vadim Zeitlin <zeitlin@dptmaths.ens-cachan.fr>
9 // Licence: wxWindows licence
10 ///////////////////////////////////////////////////////////////////////////////
11
12 /*
13 Efficient string class [more or less] compatible with MFC CString,
14 wxWidgets version 1 wxString and std::string and some handy functions
15 missing from string.h.
16 */
17
18 #ifndef _WX_WXSTRING_H__
19 #define _WX_WXSTRING_H__
20
21 // ----------------------------------------------------------------------------
22 // headers
23 // ----------------------------------------------------------------------------
24
25 #include "wx/defs.h" // everybody should include this
26
27 #ifndef __WXPALMOS5__
28 #if defined(__WXMAC__) || defined(__VISAGECPP__)
29 #include <ctype.h>
30 #endif
31
32 #if defined(__VISAGECPP__) && __IBMCPP__ >= 400
33 // problem in VACPP V4 with including stdlib.h multiple times
34 // strconv includes it anyway
35 # include <stdio.h>
36 # include <string.h>
37 # include <stdarg.h>
38 # include <limits.h>
39 #else
40 # include <string.h>
41 # include <stdio.h>
42 # include <stdarg.h>
43 # include <limits.h>
44 # include <stdlib.h>
45 #endif
46
47 #ifdef HAVE_STRCASECMP_IN_STRINGS_H
48 #include <strings.h> // for strcasecmp()
49 #endif // HAVE_STRCASECMP_IN_STRINGS_H
50 #endif // ! __WXPALMOS5__
51
52 #include "wx/wxcrtbase.h" // for wxChar, wxStrlen() etc.
53 #include "wx/strvararg.h"
54 #include "wx/buffer.h" // for wxCharBuffer
55 #include "wx/strconv.h" // for wxConvertXXX() macros and wxMBConv classes
56 #include "wx/stringimpl.h"
57 #include "wx/stringops.h"
58 #include "wx/unichar.h"
59
60 class WXDLLIMPEXP_FWD_BASE wxString;
61
62 // unless this symbol is predefined to disable the compatibility functions, do
63 // use them
64 #ifndef WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
65 #define WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER 1
66 #endif
67
68 // ---------------------------------------------------------------------------
69 // macros
70 // ---------------------------------------------------------------------------
71
72 // casts [unfortunately!] needed to call some broken functions which require
73 // "char *" instead of "const char *"
74 #define WXSTRINGCAST (wxChar *)(const wxChar *)
75 #define wxCSTRINGCAST (wxChar *)(const wxChar *)
76 #define wxMBSTRINGCAST (char *)(const char *)
77 #define wxWCSTRINGCAST (wchar_t *)(const wchar_t *)
78
79 // ----------------------------------------------------------------------------
80 // constants
81 // ----------------------------------------------------------------------------
82
83 #if WXWIN_COMPATIBILITY_2_6
84
85 // deprecated in favour of wxString::npos, don't use in new code
86 //
87 // maximum possible length for a string means "take all string" everywhere
88 #define wxSTRING_MAXLEN wxString::npos
89
90 #endif // WXWIN_COMPATIBILITY_2_6
91
92 // ---------------------------------------------------------------------------
93 // global functions complementing standard C string library replacements for
94 // strlen() and portable strcasecmp()
95 //---------------------------------------------------------------------------
96
97 #if WXWIN_COMPATIBILITY_2_8
98 // Use wxXXX() functions from wxcrt.h instead! These functions are for
99 // backwards compatibility only.
100
101 // checks whether the passed in pointer is NULL and if the string is empty
102 wxDEPRECATED( inline bool IsEmpty(const char *p) );
103 inline bool IsEmpty(const char *p) { return (!p || !*p); }
104
105 // safe version of strlen() (returns 0 if passed NULL pointer)
106 wxDEPRECATED( inline size_t Strlen(const char *psz) );
107 inline size_t Strlen(const char *psz)
108 { return psz ? strlen(psz) : 0; }
109
110 // portable strcasecmp/_stricmp
111 wxDEPRECATED( inline int Stricmp(const char *psz1, const char *psz2) );
112 inline int Stricmp(const char *psz1, const char *psz2)
113 {
114 #if defined(__VISUALC__) && defined(__WXWINCE__)
115 register char c1, c2;
116 do {
117 c1 = tolower(*psz1++);
118 c2 = tolower(*psz2++);
119 } while ( c1 && (c1 == c2) );
120
121 return c1 - c2;
122 #elif defined(__VISUALC__) || ( defined(__MWERKS__) && defined(__INTEL__) )
123 return _stricmp(psz1, psz2);
124 #elif defined(__SC__)
125 return _stricmp(psz1, psz2);
126 #elif defined(__BORLANDC__)
127 return stricmp(psz1, psz2);
128 #elif defined(__WATCOMC__)
129 return stricmp(psz1, psz2);
130 #elif defined(__DJGPP__)
131 return stricmp(psz1, psz2);
132 #elif defined(__EMX__)
133 return stricmp(psz1, psz2);
134 #elif defined(__WXPM__)
135 return stricmp(psz1, psz2);
136 #elif defined(__WXPALMOS__) || \
137 defined(HAVE_STRCASECMP_IN_STRING_H) || \
138 defined(HAVE_STRCASECMP_IN_STRINGS_H) || \
139 defined(__GNUWIN32__)
140 return strcasecmp(psz1, psz2);
141 #elif defined(__MWERKS__) && !defined(__INTEL__)
142 register char c1, c2;
143 do {
144 c1 = tolower(*psz1++);
145 c2 = tolower(*psz2++);
146 } while ( c1 && (c1 == c2) );
147
148 return c1 - c2;
149 #else
150 // almost all compilers/libraries provide this function (unfortunately under
151 // different names), that's why we don't implement our own which will surely
152 // be more efficient than this code (uncomment to use):
153 /*
154 register char c1, c2;
155 do {
156 c1 = tolower(*psz1++);
157 c2 = tolower(*psz2++);
158 } while ( c1 && (c1 == c2) );
159
160 return c1 - c2;
161 */
162
163 #error "Please define string case-insensitive compare for your OS/compiler"
164 #endif // OS/compiler
165 }
166
167 #endif // WXWIN_COMPATIBILITY_2_8
168
169 // ----------------------------------------------------------------------------
170 // wxCStrData
171 // ----------------------------------------------------------------------------
172
173 // Lightweight object returned by wxString::c_str() and implicitly convertible
174 // to either const char* or const wchar_t*.
175 class WXDLLIMPEXP_BASE wxCStrData
176 {
177 private:
178 // Ctors; for internal use by wxString and wxCStrData only
179 wxCStrData(const wxString *str, size_t offset = 0, bool owned = false)
180 : m_str(str), m_offset(offset), m_owned(owned) {}
181
182 public:
183 // Ctor constructs the object from char literal; they are needed to make
184 // operator?: compile and they intentionally take char*, not const char*
185 inline wxCStrData(char *buf);
186 inline wxCStrData(wchar_t *buf);
187 inline wxCStrData(const wxCStrData& data);
188
189 inline ~wxCStrData();
190
191 // methods defined inline below must be declared inline or mingw32 3.4.5
192 // warns about "<symbol> defined locally after being referenced with
193 // dllimport linkage"
194 #if wxUSE_UNICODE_WCHAR
195 inline
196 #endif
197 const wchar_t* AsWChar() const;
198 operator const wchar_t*() const { return AsWChar(); }
199
200 #if !wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY
201 inline
202 #endif
203 const char* AsChar() const;
204 const unsigned char* AsUnsignedChar() const
205 { return (const unsigned char *) AsChar(); }
206 operator const char*() const { return AsChar(); }
207 operator const unsigned char*() const { return AsUnsignedChar(); }
208
209 operator const void*() const { return AsChar(); }
210
211 inline const wxCharBuffer AsCharBuf() const;
212 inline const wxWCharBuffer AsWCharBuf() const;
213
214 inline wxString AsString() const;
215
216 // returns the value as C string in internal representation (equivalent
217 // to AsString().wx_str(), but more efficient)
218 const wxStringCharType *AsInternal() const;
219
220 // allow expressions like "c_str()[0]":
221 inline wxUniChar operator[](size_t n) const;
222 wxUniChar operator[](int n) const { return operator[](size_t(n)); }
223 wxUniChar operator[](long n) const { return operator[](size_t(n)); }
224 #ifndef wxSIZE_T_IS_UINT
225 wxUniChar operator[](unsigned int n) const { return operator[](size_t(n)); }
226 #endif // size_t != unsigned int
227
228 // these operators are needed to emulate the pointer semantics of c_str():
229 // expressions like "wxChar *p = str.c_str() + 1;" should continue to work
230 // (we need both versions to resolve ambiguities):
231 wxCStrData operator+(int n) const
232 { return wxCStrData(m_str, m_offset + n, m_owned); }
233 wxCStrData operator+(long n) const
234 { return wxCStrData(m_str, m_offset + n, m_owned); }
235 wxCStrData operator+(size_t n) const
236 { return wxCStrData(m_str, m_offset + n, m_owned); }
237
238 // and these for "str.c_str() + (p2 - p1)" (it also works for any integer
239 // expression but it must be ptrdiff_t and not e.g. int to work in this
240 // example):
241 wxCStrData operator-(ptrdiff_t n) const
242 {
243 wxASSERT_MSG( n <= (ptrdiff_t)m_offset,
244 _T("attempt to construct address before the beginning of the string") );
245 return wxCStrData(m_str, m_offset - n, m_owned);
246 }
247
248 // this operator is needed to make expressions like "*c_str()" or
249 // "*(c_str() + 2)" work
250 inline wxUniChar operator*() const;
251
252 private:
253 const wxString *m_str;
254 size_t m_offset;
255 bool m_owned;
256
257 friend class WXDLLIMPEXP_FWD_BASE wxString;
258 };
259
260 // ----------------------------------------------------------------------------
261 // wxStringPrintfMixin
262 // ---------------------------------------------------------------------------
263
264 // NB: VC6 has a bug that causes linker errors if you have template methods
265 // in a class using __declspec(dllimport). The solution is to split such
266 // class into two classes, one that contains the template methods and does
267 // *not* use WXDLLIMPEXP_BASE and another class that contains the rest
268 // (with DLL linkage).
269 //
270 // We only do this for VC6 here, because the code is less efficient
271 // (Printf() has to use dynamic_cast<>) and because OpenWatcom compiler
272 // cannot compile this code.
273
274 #if defined(__VISUALC__) && __VISUALC__ < 1300
275 #define wxNEEDS_WXSTRING_PRINTF_MIXIN
276 #endif
277
278 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
279 // this class contains implementation of wxString's vararg methods, it's
280 // exported from wxBase DLL
281 class WXDLLIMPEXP_BASE wxStringPrintfMixinBase
282 {
283 protected:
284 wxStringPrintfMixinBase() {}
285
286 #if !wxUSE_UTF8_LOCALE_ONLY
287 int DoPrintfWchar(const wxChar *format, ...);
288 static wxString DoFormatWchar(const wxChar *format, ...);
289 #endif
290 #if wxUSE_UNICODE_UTF8
291 int DoPrintfUtf8(const char *format, ...);
292 static wxString DoFormatUtf8(const char *format, ...);
293 #endif
294 };
295
296 // this class contains template wrappers for wxString's vararg methods, it's
297 // intentionally *not* exported from the DLL in order to fix the VC6 bug
298 // described above
299 class wxStringPrintfMixin : public wxStringPrintfMixinBase
300 {
301 private:
302 // to further complicate things, we can't return wxString from
303 // wxStringPrintfMixin::Format() because wxString is not yet declared at
304 // this point; the solution is to use this fake type trait template - this
305 // way the compiler won't know the return type until Format() is used
306 // (this doesn't compile with Watcom, but VC6 compiles it just fine):
307 template<typename T> struct StringReturnType
308 {
309 typedef wxString type;
310 };
311
312 public:
313 // these are duplicated wxString methods, they're also declared below
314 // if !wxNEEDS_WXSTRING_PRINTF_MIXIN:
315
316 // static wxString Format(const wString& format, ...) ATTRIBUTE_PRINTF_1;
317 WX_DEFINE_VARARG_FUNC_SANS_N0(static typename StringReturnType<T1>::type,
318 Format, 1, (const wxFormatString&),
319 DoFormatWchar, DoFormatUtf8)
320 // We have to implement the version without template arguments manually
321 // because of the StringReturnType<> hack, although WX_DEFINE_VARARG_FUNC
322 // normally does it itself. It has to be a template so that we can use
323 // the hack, even though there's no real template parameter. We can't move
324 // it to wxStrig, because it would shadow these versions of Format() then.
325 template<typename T>
326 inline static typename StringReturnType<T>::type
327 Format(const T& fmt)
328 {
329 // NB: this doesn't compile if T is not (some form of) a string;
330 // this makes Format's prototype equivalent to
331 // Format(const wxFormatString& fmt)
332 return DoFormatWchar(wxFormatString(fmt));
333 }
334
335 // int Printf(const wxString& format, ...);
336 WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&),
337 DoPrintfWchar, DoPrintfUtf8)
338 // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2;
339 WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&),
340 DoPrintfWchar, DoPrintfUtf8)
341
342 protected:
343 wxStringPrintfMixin() : wxStringPrintfMixinBase() {}
344 };
345 #endif // wxNEEDS_WXSTRING_PRINTF_MIXIN
346
347
348 // ----------------------------------------------------------------------------
349 // wxString: string class trying to be compatible with std::string, MFC
350 // CString and wxWindows 1.x wxString all at once
351 // ---------------------------------------------------------------------------
352
353 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
354 // "non dll-interface class 'wxStringPrintfMixin' used as base interface
355 // for dll-interface class 'wxString'" -- this is OK in our case
356 #pragma warning (disable:4275)
357 #endif
358
359 #if wxUSE_UNICODE_UTF8
360 // see the comment near wxString::iterator for why we need this
361 class WXDLLIMPEXP_BASE wxStringIteratorNode
362 {
363 public:
364 wxStringIteratorNode()
365 : m_str(NULL), m_citer(NULL), m_iter(NULL), m_prev(NULL), m_next(NULL) {}
366 wxStringIteratorNode(const wxString *str,
367 wxStringImpl::const_iterator *citer)
368 { DoSet(str, citer, NULL); }
369 wxStringIteratorNode(const wxString *str, wxStringImpl::iterator *iter)
370 { DoSet(str, NULL, iter); }
371 ~wxStringIteratorNode()
372 { clear(); }
373
374 inline void set(const wxString *str, wxStringImpl::const_iterator *citer)
375 { clear(); DoSet(str, citer, NULL); }
376 inline void set(const wxString *str, wxStringImpl::iterator *iter)
377 { clear(); DoSet(str, NULL, iter); }
378
379 const wxString *m_str;
380 wxStringImpl::const_iterator *m_citer;
381 wxStringImpl::iterator *m_iter;
382 wxStringIteratorNode *m_prev, *m_next;
383
384 private:
385 inline void clear();
386 inline void DoSet(const wxString *str,
387 wxStringImpl::const_iterator *citer,
388 wxStringImpl::iterator *iter);
389
390 // the node belongs to a particular iterator instance, it's not copied
391 // when a copy of the iterator is made
392 DECLARE_NO_COPY_CLASS(wxStringIteratorNode)
393 };
394 #endif // wxUSE_UNICODE_UTF8
395
396 class WXDLLIMPEXP_BASE wxString
397 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
398 : public wxStringPrintfMixin
399 #endif
400 {
401 // NB: special care was taken in arranging the member functions in such order
402 // that all inline functions can be effectively inlined, verify that all
403 // performance critical functions are still inlined if you change order!
404 public:
405 // an 'invalid' value for string index, moved to this place due to a CW bug
406 static const size_t npos;
407
408 private:
409 // if we hadn't made these operators private, it would be possible to
410 // compile "wxString s; s = 17;" without any warnings as 17 is implicitly
411 // converted to char in C and we do have operator=(char)
412 //
413 // NB: we don't need other versions (short/long and unsigned) as attempt
414 // to assign another numeric type to wxString will now result in
415 // ambiguity between operator=(char) and operator=(int)
416 wxString& operator=(int);
417
418 // these methods are not implemented - there is _no_ conversion from int to
419 // string, you're doing something wrong if the compiler wants to call it!
420 //
421 // try `s << i' or `s.Printf("%d", i)' instead
422 wxString(int);
423
424
425 // buffer for holding temporary substring when using any of the methods
426 // that take (char*,size_t) or (wchar_t*,size_t) arguments:
427 template<typename T>
428 struct SubstrBufFromType
429 {
430 T data;
431 size_t len;
432
433 SubstrBufFromType(const T& data_, size_t len_)
434 : data(data_), len(len_)
435 {
436 wxASSERT_MSG( len != npos, "must have real length" );
437 }
438 };
439
440 #if wxUSE_UNICODE_UTF8
441 // even char* -> char* needs conversion, from locale charset to UTF-8
442 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromWC;
443 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromMB;
444 #elif wxUSE_UNICODE_WCHAR
445 typedef SubstrBufFromType<const wchar_t*> SubstrBufFromWC;
446 typedef SubstrBufFromType<wxWCharBuffer> SubstrBufFromMB;
447 #else
448 typedef SubstrBufFromType<const char*> SubstrBufFromMB;
449 typedef SubstrBufFromType<wxCharBuffer> SubstrBufFromWC;
450 #endif
451
452
453 // Functions implementing primitive operations on string data; wxString
454 // methods and iterators are implemented in terms of it. The differences
455 // between UTF-8 and wchar_t* representations of the string are mostly
456 // contained here.
457
458 #if wxUSE_UNICODE_UTF8
459 static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength,
460 const wxMBConv& conv);
461 static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength,
462 const wxMBConv& conv);
463 #elif wxUSE_UNICODE_WCHAR
464 static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength,
465 const wxMBConv& conv);
466 #else
467 static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength,
468 const wxMBConv& conv);
469 #endif
470
471 #if !wxUSE_UNICODE_UTF8 // wxUSE_UNICODE_WCHAR or !wxUSE_UNICODE
472 // returns C string encoded as the implementation expects:
473 #if wxUSE_UNICODE
474 static const wchar_t* ImplStr(const wchar_t* str)
475 { return str ? str : wxT(""); }
476 static const SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
477 { return SubstrBufFromWC(str, (str && n == npos) ? wxWcslen(str) : n); }
478 static wxWCharBuffer ImplStr(const char* str,
479 const wxMBConv& conv = wxConvLibc)
480 { return ConvertStr(str, npos, conv).data; }
481 static SubstrBufFromMB ImplStr(const char* str, size_t n,
482 const wxMBConv& conv = wxConvLibc)
483 { return ConvertStr(str, n, conv); }
484 #else
485 static const char* ImplStr(const char* str,
486 const wxMBConv& WXUNUSED(conv) = wxConvLibc)
487 { return str ? str : ""; }
488 static const SubstrBufFromMB ImplStr(const char* str, size_t n,
489 const wxMBConv& WXUNUSED(conv) = wxConvLibc)
490 { return SubstrBufFromMB(str, (str && n == npos) ? wxStrlen(str) : n); }
491 static wxCharBuffer ImplStr(const wchar_t* str)
492 { return ConvertStr(str, npos, wxConvLibc).data; }
493 static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
494 { return ConvertStr(str, n, wxConvLibc); }
495 #endif
496
497 // translates position index in wxString to/from index in underlying
498 // wxStringImpl:
499 static size_t PosToImpl(size_t pos) { return pos; }
500 static void PosLenToImpl(size_t pos, size_t len,
501 size_t *implPos, size_t *implLen)
502 { *implPos = pos; *implLen = len; }
503 static size_t LenToImpl(size_t len) { return len; }
504 static size_t PosFromImpl(size_t pos) { return pos; }
505
506 #else // wxUSE_UNICODE_UTF8
507
508 static wxCharBuffer ImplStr(const char* str,
509 const wxMBConv& conv = wxConvLibc)
510 { return ConvertStr(str, npos, conv).data; }
511 static SubstrBufFromMB ImplStr(const char* str, size_t n,
512 const wxMBConv& conv = wxConvLibc)
513 { return ConvertStr(str, n, conv); }
514
515 static wxCharBuffer ImplStr(const wchar_t* str)
516 { return ConvertStr(str, npos, wxMBConvUTF8()).data; }
517 static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
518 { return ConvertStr(str, n, wxMBConvUTF8()); }
519
520 size_t PosToImpl(size_t pos) const
521 {
522 if ( pos == 0 || pos == npos )
523 return pos;
524 else
525 return (begin() + pos).impl() - m_impl.begin();
526 }
527
528 void PosLenToImpl(size_t pos, size_t len, size_t *implPos, size_t *implLen) const;
529
530 size_t LenToImpl(size_t len) const
531 {
532 size_t pos, len2;
533 PosLenToImpl(0, len, &pos, &len2);
534 return len2;
535 }
536
537 size_t PosFromImpl(size_t pos) const
538 {
539 if ( pos == 0 || pos == npos )
540 return pos;
541 else
542 return const_iterator(this, m_impl.begin() + pos) - begin();
543 }
544 #endif // !wxUSE_UNICODE_UTF8/wxUSE_UNICODE_UTF8
545
546 public:
547 // standard types
548 typedef wxUniChar value_type;
549 typedef wxUniChar char_type;
550 typedef wxUniCharRef reference;
551 typedef wxChar* pointer;
552 typedef const wxChar* const_pointer;
553
554 typedef size_t size_type;
555 typedef wxUniChar const_reference;
556
557 #if wxUSE_STL
558 #if wxUSE_UNICODE_UTF8
559 // random access is not O(1), as required by Random Access Iterator
560 #define WX_STR_ITERATOR_TAG std::bidirectional_iterator_tag
561 #else
562 #define WX_STR_ITERATOR_TAG std::random_access_iterator_tag
563 #endif
564 #else
565 #define WX_STR_ITERATOR_TAG void /* dummy type */
566 #endif
567
568 #define WX_STR_ITERATOR_IMPL(iterator_name, pointer_type, reference_type) \
569 private: \
570 typedef wxStringImpl::iterator_name underlying_iterator; \
571 public: \
572 typedef WX_STR_ITERATOR_TAG iterator_category; \
573 typedef wxUniChar value_type; \
574 typedef int difference_type; \
575 typedef reference_type reference; \
576 typedef pointer_type pointer; \
577 \
578 reference operator[](size_t n) const { return *(*this + n); } \
579 \
580 iterator_name& operator++() \
581 { wxStringOperations::IncIter(m_cur); return *this; } \
582 iterator_name& operator--() \
583 { wxStringOperations::DecIter(m_cur); return *this; } \
584 iterator_name operator++(int) \
585 { \
586 iterator_name tmp = *this; \
587 wxStringOperations::IncIter(m_cur); \
588 return tmp; \
589 } \
590 iterator_name operator--(int) \
591 { \
592 iterator_name tmp = *this; \
593 wxStringOperations::DecIter(m_cur); \
594 return tmp; \
595 } \
596 \
597 iterator_name& operator+=(ptrdiff_t n) \
598 { \
599 m_cur = wxStringOperations::AddToIter(m_cur, n); \
600 return *this; \
601 } \
602 iterator_name& operator-=(ptrdiff_t n) \
603 { \
604 m_cur = wxStringOperations::AddToIter(m_cur, -n); \
605 return *this; \
606 } \
607 \
608 difference_type operator-(const iterator_name& i) const \
609 { return wxStringOperations::DiffIters(m_cur, i.m_cur); } \
610 \
611 bool operator==(const iterator_name& i) const \
612 { return m_cur == i.m_cur; } \
613 bool operator!=(const iterator_name& i) const \
614 { return m_cur != i.m_cur; } \
615 \
616 bool operator<(const iterator_name& i) const \
617 { return m_cur < i.m_cur; } \
618 bool operator>(const iterator_name& i) const \
619 { return m_cur > i.m_cur; } \
620 bool operator<=(const iterator_name& i) const \
621 { return m_cur <= i.m_cur; } \
622 bool operator>=(const iterator_name& i) const \
623 { return m_cur >= i.m_cur; } \
624 \
625 private: \
626 /* for internal wxString use only: */ \
627 underlying_iterator impl() const { return m_cur; } \
628 \
629 friend class wxString; \
630 friend class wxCStrData; \
631 \
632 private: \
633 underlying_iterator m_cur
634
635 class WXDLLIMPEXP_FWD_BASE const_iterator;
636
637 #if wxUSE_UNICODE_UTF8
638 // NB: In UTF-8 build, (non-const) iterator needs to keep reference
639 // to the underlying wxStringImpl, because UTF-8 is variable-length
640 // encoding and changing the value pointer to by an iterator (using
641 // its operator*) requires calling wxStringImpl::replace() if the old
642 // and new values differ in their encoding's length.
643 //
644 // Furthermore, the replace() call may invalid all iterators for the
645 // string, so we have to keep track of outstanding iterators and update
646 // them if replace() happens.
647 //
648 // This is implemented by maintaining linked list of iterators for every
649 // string and traversing it in wxUniCharRef::operator=(). Head of the
650 // list is stored in wxString. (FIXME-UTF8)
651
652 class WXDLLIMPEXP_BASE iterator
653 {
654 WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);
655
656 public:
657 iterator() {}
658 iterator(const iterator& i)
659 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
660 iterator& operator=(const iterator& i)
661 {
662 if (&i != this)
663 {
664 m_cur = i.m_cur;
665 m_node.set(i.str(), &m_cur);
666 }
667 return *this;
668 }
669
670 reference operator*()
671 { return wxUniCharRef::CreateForString(m_node, m_cur); }
672
673 iterator operator+(ptrdiff_t n) const
674 { return iterator(str(), wxStringOperations::AddToIter(m_cur, n)); }
675 iterator operator-(ptrdiff_t n) const
676 { return iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); }
677
678 private:
679 iterator(wxString *str, underlying_iterator ptr)
680 : m_cur(ptr), m_node(str, &m_cur) {}
681
682 wxString* str() const { return wx_const_cast(wxString*, m_node.m_str); }
683
684 wxStringIteratorNode m_node;
685
686 friend class const_iterator;
687 };
688
689 class WXDLLIMPEXP_BASE const_iterator
690 {
691 // NB: reference_type is intentionally value, not reference, the character
692 // may be encoded differently in wxString data:
693 WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar);
694
695 public:
696 const_iterator() {}
697 const_iterator(const const_iterator& i)
698 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
699 const_iterator(const iterator& i)
700 : m_cur(i.m_cur), m_node(i.str(), &m_cur) {}
701
702 const_iterator& operator=(const const_iterator& i)
703 {
704 if (&i != this)
705 {
706 m_cur = i.m_cur;
707 m_node.set(i.str(), &m_cur);
708 }
709 return *this;
710 }
711 const_iterator& operator=(const iterator& i)
712 { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; }
713
714 reference operator*() const
715 { return wxStringOperations::DecodeChar(m_cur); }
716
717 const_iterator operator+(ptrdiff_t n) const
718 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, n)); }
719 const_iterator operator-(ptrdiff_t n) const
720 { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); }
721
722 private:
723 // for internal wxString use only:
724 const_iterator(const wxString *str, underlying_iterator ptr)
725 : m_cur(ptr), m_node(str, &m_cur) {}
726
727 const wxString* str() const { return m_node.m_str; }
728
729 wxStringIteratorNode m_node;
730 };
731
732 size_t IterToImplPos(wxString::iterator i) const
733 { return wxStringImpl::const_iterator(i.impl()) - m_impl.begin(); }
734
735 #else // !wxUSE_UNICODE_UTF8
736
737 class WXDLLIMPEXP_BASE iterator
738 {
739 WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);
740
741 public:
742 iterator() {}
743 iterator(const iterator& i) : m_cur(i.m_cur) {}
744
745 reference operator*()
746 { return wxUniCharRef::CreateForString(m_cur); }
747
748 iterator operator+(ptrdiff_t n) const
749 { return iterator(wxStringOperations::AddToIter(m_cur, n)); }
750 iterator operator-(ptrdiff_t n) const
751 { return iterator(wxStringOperations::AddToIter(m_cur, -n)); }
752
753 private:
754 // for internal wxString use only:
755 iterator(underlying_iterator ptr) : m_cur(ptr) {}
756 iterator(wxString *WXUNUSED(str), underlying_iterator ptr) : m_cur(ptr) {}
757
758 friend class const_iterator;
759 };
760
761 class WXDLLIMPEXP_BASE const_iterator
762 {
763 // NB: reference_type is intentionally value, not reference, the character
764 // may be encoded differently in wxString data:
765 WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar);
766
767 public:
768 const_iterator() {}
769 const_iterator(const const_iterator& i) : m_cur(i.m_cur) {}
770 const_iterator(const iterator& i) : m_cur(i.m_cur) {}
771
772 reference operator*() const
773 { return wxStringOperations::DecodeChar(m_cur); }
774
775 const_iterator operator+(ptrdiff_t n) const
776 { return const_iterator(wxStringOperations::AddToIter(m_cur, n)); }
777 const_iterator operator-(ptrdiff_t n) const
778 { return const_iterator(wxStringOperations::AddToIter(m_cur, -n)); }
779
780 private:
781 // for internal wxString use only:
782 const_iterator(underlying_iterator ptr) : m_cur(ptr) {}
783 const_iterator(const wxString *WXUNUSED(str), underlying_iterator ptr)
784 : m_cur(ptr) {}
785 };
786 #endif // wxUSE_UNICODE_UTF8/!wxUSE_UNICODE_UTF8
787
788 #undef WX_STR_ITERATOR_TAG
789 #undef WX_STR_ITERATOR_IMPL
790
791 friend class iterator;
792 friend class const_iterator;
793
794 template <typename T>
795 class reverse_iterator_impl
796 {
797 public:
798 typedef T iterator_type;
799
800 typedef typename T::iterator_category iterator_category;
801 typedef typename T::value_type value_type;
802 typedef typename T::difference_type difference_type;
803 typedef typename T::reference reference;
804 typedef typename T::pointer *pointer;
805
806 reverse_iterator_impl() {}
807 reverse_iterator_impl(iterator_type i) : m_cur(i) {}
808 reverse_iterator_impl(const reverse_iterator_impl& ri)
809 : m_cur(ri.m_cur) {}
810
811 iterator_type base() const { return m_cur; }
812
813 reference operator*() const { return *(m_cur-1); }
814 reference operator[](size_t n) const { return *(*this + n); }
815
816 reverse_iterator_impl& operator++()
817 { --m_cur; return *this; }
818 reverse_iterator_impl operator++(int)
819 { reverse_iterator_impl tmp = *this; --m_cur; return tmp; }
820 reverse_iterator_impl& operator--()
821 { ++m_cur; return *this; }
822 reverse_iterator_impl operator--(int)
823 { reverse_iterator_impl tmp = *this; ++m_cur; return tmp; }
824
825 // NB: explicit <T> in the functions below is to keep BCC 5.5 happy
826 reverse_iterator_impl operator+(ptrdiff_t n) const
827 { return reverse_iterator_impl<T>(m_cur - n); }
828 reverse_iterator_impl operator-(ptrdiff_t n) const
829 { return reverse_iterator_impl<T>(m_cur + n); }
830 reverse_iterator_impl operator+=(ptrdiff_t n)
831 { m_cur -= n; return *this; }
832 reverse_iterator_impl operator-=(ptrdiff_t n)
833 { m_cur += n; return *this; }
834
835 unsigned operator-(const reverse_iterator_impl& i) const
836 { return i.m_cur - m_cur; }
837
838 bool operator==(const reverse_iterator_impl& ri) const
839 { return m_cur == ri.m_cur; }
840 bool operator!=(const reverse_iterator_impl& ri) const
841 { return !(*this == ri); }
842
843 bool operator<(const reverse_iterator_impl& i) const
844 { return m_cur > i.m_cur; }
845 bool operator>(const reverse_iterator_impl& i) const
846 { return m_cur < i.m_cur; }
847 bool operator<=(const reverse_iterator_impl& i) const
848 { return m_cur >= i.m_cur; }
849 bool operator>=(const reverse_iterator_impl& i) const
850 { return m_cur <= i.m_cur; }
851
852 private:
853 iterator_type m_cur;
854 };
855
856 typedef reverse_iterator_impl<iterator> reverse_iterator;
857 typedef reverse_iterator_impl<const_iterator> const_reverse_iterator;
858
859 private:
860 // used to transform an expression built using c_str() (and hence of type
861 // wxCStrData) to an iterator into the string
862 static const_iterator CreateConstIterator(const wxCStrData& data)
863 {
864 return const_iterator(data.m_str,
865 (data.m_str->begin() + data.m_offset).impl());
866 }
867
868 // in UTF-8 STL build, creation from std::string requires conversion under
869 // non-UTF8 locales, so we can't have and use wxString(wxStringImpl) ctor;
870 // instead we define dummy type that lets us have wxString ctor for creation
871 // from wxStringImpl that couldn't be used by user code (in all other builds,
872 // "standard" ctors can be used):
873 #if wxUSE_UNICODE_UTF8 && wxUSE_STL_BASED_WXSTRING
874 struct CtorFromStringImplTag {};
875
876 wxString(CtorFromStringImplTag* WXUNUSED(dummy), const wxStringImpl& src)
877 : m_impl(src) {}
878
879 static wxString FromImpl(const wxStringImpl& src)
880 { return wxString((CtorFromStringImplTag*)NULL, src); }
881 #else
882 #if !wxUSE_STL_BASED_WXSTRING
883 wxString(const wxStringImpl& src) : m_impl(src) { }
884 // else: already defined as wxString(wxStdString) below
885 #endif
886 static wxString FromImpl(const wxStringImpl& src) { return wxString(src); }
887 #endif
888
889 public:
890 // constructors and destructor
891 // ctor for an empty string
892 wxString() {}
893
894 // copy ctor
895 wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { }
896
897 // string containing nRepeat copies of ch
898 wxString(wxUniChar ch, size_t nRepeat = 1 )
899 { assign(nRepeat, ch); }
900 wxString(size_t nRepeat, wxUniChar ch)
901 { assign(nRepeat, ch); }
902 wxString(wxUniCharRef ch, size_t nRepeat = 1)
903 { assign(nRepeat, ch); }
904 wxString(size_t nRepeat, wxUniCharRef ch)
905 { assign(nRepeat, ch); }
906 wxString(char ch, size_t nRepeat = 1)
907 { assign(nRepeat, ch); }
908 wxString(size_t nRepeat, char ch)
909 { assign(nRepeat, ch); }
910 wxString(wchar_t ch, size_t nRepeat = 1)
911 { assign(nRepeat, ch); }
912 wxString(size_t nRepeat, wchar_t ch)
913 { assign(nRepeat, ch); }
914
915 // ctors from char* strings:
916 wxString(const char *psz)
917 : m_impl(ImplStr(psz)) {}
918 wxString(const char *psz, const wxMBConv& conv)
919 : m_impl(ImplStr(psz, conv)) {}
920 wxString(const char *psz, size_t nLength)
921 { assign(psz, nLength); }
922 wxString(const char *psz, const wxMBConv& conv, size_t nLength)
923 {
924 SubstrBufFromMB str(ImplStr(psz, nLength, conv));
925 m_impl.assign(str.data, str.len);
926 }
927
928 // and unsigned char*:
929 wxString(const unsigned char *psz)
930 : m_impl(ImplStr((const char*)psz)) {}
931 wxString(const unsigned char *psz, const wxMBConv& conv)
932 : m_impl(ImplStr((const char*)psz, conv)) {}
933 wxString(const unsigned char *psz, size_t nLength)
934 { assign((const char*)psz, nLength); }
935 wxString(const unsigned char *psz, const wxMBConv& conv, size_t nLength)
936 {
937 SubstrBufFromMB str(ImplStr((const char*)psz, nLength, conv));
938 m_impl.assign(str.data, str.len);
939 }
940
941 // ctors from wchar_t* strings:
942 wxString(const wchar_t *pwz)
943 : m_impl(ImplStr(pwz)) {}
944 wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv))
945 : m_impl(ImplStr(pwz)) {}
946 wxString(const wchar_t *pwz, size_t nLength)
947 { assign(pwz, nLength); }
948 wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv), size_t nLength)
949 { assign(pwz, nLength); }
950
951 wxString(const wxCharBuffer& buf)
952 { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length
953 wxString(const wxWCharBuffer& buf)
954 { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length
955
956 // NB: this version uses m_impl.c_str() to force making a copy of the
957 // string, so that "wxString(str.c_str())" idiom for passing strings
958 // between threads works
959 wxString(const wxCStrData& cstr)
960 : m_impl(cstr.AsString().m_impl.c_str()) { }
961
962 // as we provide both ctors with this signature for both char and unsigned
963 // char string, we need to provide one for wxCStrData to resolve ambiguity
964 wxString(const wxCStrData& cstr, size_t nLength)
965 : m_impl(cstr.AsString().Mid(0, nLength).m_impl) {}
966
967 // and because wxString is convertible to wxCStrData and const wxChar *
968 // we also need to provide this one
969 wxString(const wxString& str, size_t nLength)
970 { assign(str, nLength); }
971
972 // even if we're not built with wxUSE_STL == 1 it is very convenient to allow
973 // implicit conversions from std::string to wxString and vice verse as this
974 // allows to use the same strings in non-GUI and GUI code, however we don't
975 // want to unconditionally add this ctor as it would make wx lib dependent on
976 // libstdc++ on some Linux versions which is bad, so instead we ask the
977 // client code to define this wxUSE_STD_STRING symbol if they need it
978 #if wxUSE_STD_STRING
979 #if wxUSE_UNICODE_WCHAR
980 wxString(const wxStdWideString& str) : m_impl(str) {}
981 #else // UTF-8 or ANSI
982 wxString(const wxStdWideString& str)
983 { assign(str.c_str(), str.length()); }
984 #endif
985
986 #if !wxUSE_UNICODE // ANSI build
987 // FIXME-UTF8: do this in UTF8 build #if wxUSE_UTF8_LOCALE_ONLY, too
988 wxString(const std::string& str) : m_impl(str) {}
989 #else // Unicode
990 wxString(const std::string& str)
991 { assign(str.c_str(), str.length()); }
992 #endif
993 #endif // wxUSE_STD_STRING
994
995 // Unlike ctor from std::string, we provide conversion to std::string only
996 // if wxUSE_STL and not merely wxUSE_STD_STRING (which is on by default),
997 // because it conflicts with operator const char/wchar_t*:
998 #if wxUSE_STL
999 #if wxUSE_UNICODE_WCHAR && wxUSE_STL_BASED_WXSTRING
1000 // wxStringImpl is std::string in the encoding we want
1001 operator const wxStdWideString&() const { return m_impl; }
1002 #else
1003 // wxStringImpl is either not std::string or needs conversion
1004 operator wxStdWideString() const
1005 // FIXME-UTF8: broken for embedded NULs
1006 { return wxStdWideString(wc_str()); }
1007 #endif
1008
1009 #if (!wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY) && wxUSE_STL_BASED_WXSTRING
1010 // wxStringImpl is std::string in the encoding we want
1011 operator const std::string&() const { return m_impl; }
1012 #else
1013 // wxStringImpl is either not std::string or needs conversion
1014 operator std::string() const
1015 // FIXME-UTF8: broken for embedded NULs
1016 { return std::string(mb_str()); }
1017 #endif
1018 #endif // wxUSE_STL
1019
1020 wxString Clone() const
1021 {
1022 // make a deep copy of the string, i.e. the returned string will have
1023 // ref count = 1 with refcounted implementation
1024 return wxString::FromImpl(wxStringImpl(m_impl.c_str(), m_impl.length()));
1025 }
1026
1027 // first valid index position
1028 const_iterator begin() const { return const_iterator(this, m_impl.begin()); }
1029 iterator begin() { return iterator(this, m_impl.begin()); }
1030 // position one after the last valid one
1031 const_iterator end() const { return const_iterator(this, m_impl.end()); }
1032 iterator end() { return iterator(this, m_impl.end()); }
1033
1034 // first element of the reversed string
1035 const_reverse_iterator rbegin() const
1036 { return const_reverse_iterator(end()); }
1037 reverse_iterator rbegin()
1038 { return reverse_iterator(end()); }
1039 // one beyond the end of the reversed string
1040 const_reverse_iterator rend() const
1041 { return const_reverse_iterator(begin()); }
1042 reverse_iterator rend()
1043 { return reverse_iterator(begin()); }
1044
1045 // std::string methods:
1046 #if wxUSE_UNICODE_UTF8
1047 size_t length() const { return end() - begin(); } // FIXME-UTF8: optimize!
1048 #else
1049 size_t length() const { return m_impl.length(); }
1050 #endif
1051
1052 size_type size() const { return length(); }
1053 size_type max_size() const { return npos; }
1054
1055 bool empty() const { return m_impl.empty(); }
1056
1057 size_type capacity() const { return m_impl.capacity(); } // FIXME-UTF8
1058 void reserve(size_t sz) { m_impl.reserve(sz); } // FIXME-UTF8
1059
1060 void resize(size_t nSize, wxUniChar ch = wxT('\0'))
1061 {
1062 const size_t len = length();
1063 if ( nSize == len)
1064 return;
1065
1066 #if wxUSE_UNICODE_UTF8
1067 if ( nSize < len )
1068 {
1069 // we can't use wxStringImpl::resize() for truncating the string as it
1070 // counts in bytes, not characters
1071 erase(nSize);
1072 return;
1073 }
1074
1075 // we also can't use (presumably more efficient) resize() if we have to
1076 // append characters taking more than one byte
1077 if ( !ch.IsAscii() )
1078 append(nSize - len, ch);
1079 else
1080 #endif // wxUSE_UNICODE_UTF8
1081 m_impl.resize(nSize, (wxStringCharType)ch);
1082 }
1083
1084 wxString substr(size_t nStart = 0, size_t nLen = npos) const
1085 {
1086 size_t pos, len;
1087 PosLenToImpl(nStart, nLen, &pos, &len);
1088 return FromImpl(m_impl.substr(pos, len));
1089 }
1090
1091 // generic attributes & operations
1092 // as standard strlen()
1093 size_t Len() const { return length(); }
1094 // string contains any characters?
1095 bool IsEmpty() const { return empty(); }
1096 // empty string is "false", so !str will return true
1097 bool operator!() const { return empty(); }
1098 // truncate the string to given length
1099 wxString& Truncate(size_t uiLen);
1100 // empty string contents
1101 void Empty()
1102 {
1103 Truncate(0);
1104
1105 wxASSERT_MSG( empty(), _T("string not empty after call to Empty()?") );
1106 }
1107 // empty the string and free memory
1108 void Clear()
1109 {
1110 wxString tmp(wxEmptyString);
1111 swap(tmp);
1112 }
1113
1114 // contents test
1115 // Is an ascii value
1116 bool IsAscii() const;
1117 // Is a number
1118 bool IsNumber() const;
1119 // Is a word
1120 bool IsWord() const;
1121
1122 // data access (all indexes are 0 based)
1123 // read access
1124 wxUniChar at(size_t n) const
1125 { return *(begin() + n); } // FIXME-UTF8: optimize?
1126 wxUniChar GetChar(size_t n) const
1127 { return at(n); }
1128 // read/write access
1129 wxUniCharRef at(size_t n)
1130 { return *(begin() + n); } // FIXME-UTF8: optimize?
1131 wxUniCharRef GetWritableChar(size_t n)
1132 { return at(n); }
1133 // write access
1134 void SetChar(size_t n, wxUniChar ch)
1135 { at(n) = ch; }
1136
1137 // get last character
1138 wxUniChar Last() const
1139 {
1140 wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
1141 return *rbegin();
1142 }
1143
1144 // get writable last character
1145 wxUniCharRef Last()
1146 {
1147 wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
1148 return *rbegin();
1149 }
1150
1151 /*
1152 Note that we we must define all of the overloads below to avoid
1153 ambiguity when using str[0].
1154 */
1155 wxUniChar operator[](int n) const
1156 { return at(n); }
1157 wxUniChar operator[](long n) const
1158 { return at(n); }
1159 wxUniChar operator[](size_t n) const
1160 { return at(n); }
1161 #ifndef wxSIZE_T_IS_UINT
1162 wxUniChar operator[](unsigned int n) const
1163 { return at(n); }
1164 #endif // size_t != unsigned int
1165
1166 // operator versions of GetWriteableChar()
1167 wxUniCharRef operator[](int n)
1168 { return at(n); }
1169 wxUniCharRef operator[](long n)
1170 { return at(n); }
1171 wxUniCharRef operator[](size_t n)
1172 { return at(n); }
1173 #ifndef wxSIZE_T_IS_UINT
1174 wxUniCharRef operator[](unsigned int n)
1175 { return at(n); }
1176 #endif // size_t != unsigned int
1177
1178 // explicit conversion to C string (use this with printf()!)
1179 wxCStrData c_str() const { return wxCStrData(this); }
1180 wxCStrData data() const { return c_str(); }
1181
1182 // implicit conversion to C string
1183 operator wxCStrData() const { return c_str(); }
1184
1185 // the first two operators conflict with operators for conversion to
1186 // std::string and they must be disabled in STL build; the next one only
1187 // makes sense if conversions to char* are also defined and not defining it
1188 // in STL build also helps us to get more clear error messages for the code
1189 // which relies on implicit conversion to char* in STL build
1190 #if !wxUSE_STL
1191 operator const char*() const { return c_str(); }
1192 operator const wchar_t*() const { return c_str(); }
1193
1194 // implicit conversion to untyped pointer for compatibility with previous
1195 // wxWidgets versions: this is the same as conversion to const char * so it
1196 // may fail!
1197 operator const void*() const { return c_str(); }
1198 #endif // wxUSE_STL
1199
1200 // identical to c_str(), for MFC compatibility
1201 const wxCStrData GetData() const { return c_str(); }
1202
1203 // explicit conversion to C string in internal representation (char*,
1204 // wchar_t*, UTF-8-encoded char*, depending on the build):
1205 const wxStringCharType *wx_str() const { return m_impl.c_str(); }
1206
1207 // conversion to *non-const* multibyte or widestring buffer; modifying
1208 // returned buffer won't affect the string, these methods are only useful
1209 // for passing values to const-incorrect functions
1210 wxWritableCharBuffer char_str(const wxMBConv& conv = wxConvLibc) const
1211 { return mb_str(conv); }
1212 wxWritableWCharBuffer wchar_str() const { return wc_str(); }
1213
1214 // conversion to/from plain (i.e. 7 bit) ASCII: this is useful for
1215 // converting numbers or strings which are certain not to contain special
1216 // chars (typically system functions, X atoms, environment variables etc.)
1217 //
1218 // the behaviour of these functions with the strings containing anything
1219 // else than 7 bit ASCII characters is undefined, use at your own risk.
1220 #if wxUSE_UNICODE
1221 static wxString FromAscii(const char *ascii, size_t len);
1222 static wxString FromAscii(const char *ascii);
1223 static wxString FromAscii(char ascii);
1224 const wxCharBuffer ToAscii() const;
1225 #else // ANSI
1226 static wxString FromAscii(const char *ascii) { return wxString( ascii ); }
1227 static wxString FromAscii(const char *ascii, size_t len)
1228 { return wxString( ascii, len ); }
1229 static wxString FromAscii(char ascii) { return wxString( ascii ); }
1230 const char *ToAscii() const { return c_str(); }
1231 #endif // Unicode/!Unicode
1232
1233 // also provide unsigned char overloads as signed/unsigned doesn't matter
1234 // for 7 bit ASCII characters
1235 static wxString FromAscii(const unsigned char *ascii)
1236 { return FromAscii((const char *)ascii); }
1237 static wxString FromAscii(const unsigned char *ascii, size_t len)
1238 { return FromAscii((const char *)ascii, len); }
1239
1240 // conversion to/from UTF-8:
1241 #if wxUSE_UNICODE_UTF8
1242 static wxString FromUTF8(const char *utf8)
1243 {
1244 if ( !utf8 )
1245 return wxEmptyString;
1246
1247 wxASSERT( wxStringOperations::IsValidUtf8String(utf8) );
1248 return FromImpl(wxStringImpl(utf8));
1249 }
1250 static wxString FromUTF8(const char *utf8, size_t len)
1251 {
1252 if ( !utf8 )
1253 return wxEmptyString;
1254 if ( len == npos )
1255 return FromUTF8(utf8);
1256
1257 wxASSERT( wxStringOperations::IsValidUtf8String(utf8, len) );
1258 return FromImpl(wxStringImpl(utf8, len));
1259 }
1260 const char* utf8_str() const { return wx_str(); }
1261 const char* ToUTF8() const { return wx_str(); }
1262 #elif wxUSE_UNICODE_WCHAR
1263 static wxString FromUTF8(const char *utf8)
1264 { return wxString(utf8, wxMBConvUTF8()); }
1265 static wxString FromUTF8(const char *utf8, size_t len)
1266 { return wxString(utf8, wxMBConvUTF8(), len); }
1267 const wxCharBuffer utf8_str() const { return mb_str(wxMBConvUTF8()); }
1268 const wxCharBuffer ToUTF8() const { return utf8_str(); }
1269 #else // ANSI
1270 static wxString FromUTF8(const char *utf8)
1271 { return wxString(wxMBConvUTF8().cMB2WC(utf8)); }
1272 static wxString FromUTF8(const char *utf8, size_t len)
1273 {
1274 size_t wlen;
1275 wxWCharBuffer buf(wxMBConvUTF8().cMB2WC(utf8, len == npos ? wxNO_LEN : len, &wlen));
1276 return wxString(buf.data(), wlen);
1277 }
1278 const wxCharBuffer utf8_str() const
1279 { return wxMBConvUTF8().cWC2MB(wc_str()); }
1280 const wxCharBuffer ToUTF8() const { return utf8_str(); }
1281 #endif
1282
1283 // functions for storing binary data in wxString:
1284 #if wxUSE_UNICODE
1285 static wxString From8BitData(const char *data, size_t len)
1286 { return wxString(data, wxConvISO8859_1, len); }
1287 // version for NUL-terminated data:
1288 static wxString From8BitData(const char *data)
1289 { return wxString(data, wxConvISO8859_1); }
1290 const wxCharBuffer To8BitData() const { return mb_str(wxConvISO8859_1); }
1291 #else // ANSI
1292 static wxString From8BitData(const char *data, size_t len)
1293 { return wxString(data, len); }
1294 // version for NUL-terminated data:
1295 static wxString From8BitData(const char *data)
1296 { return wxString(data); }
1297 const char *To8BitData() const { return c_str(); }
1298 #endif // Unicode/ANSI
1299
1300 // conversions with (possible) format conversions: have to return a
1301 // buffer with temporary data
1302 //
1303 // the functions defined (in either Unicode or ANSI) mode are mb_str() to
1304 // return an ANSI (multibyte) string, wc_str() to return a wide string and
1305 // fn_str() to return a string which should be used with the OS APIs
1306 // accepting the file names. The return value is always the same, but the
1307 // type differs because a function may either return pointer to the buffer
1308 // directly or have to use intermediate buffer for translation.
1309 #if wxUSE_UNICODE
1310
1311 #if wxUSE_UTF8_LOCALE_ONLY
1312 const char* mb_str() const { return wx_str(); }
1313 const wxCharBuffer mb_str(const wxMBConv& conv) const;
1314 #else
1315 const wxCharBuffer mb_str(const wxMBConv& conv = wxConvLibc) const;
1316 #endif
1317
1318 const wxWX2MBbuf mbc_str() const { return mb_str(*wxConvCurrent); }
1319
1320 #if wxUSE_UNICODE_WCHAR
1321 const wxChar* wc_str() const { return wx_str(); }
1322 #elif wxUSE_UNICODE_UTF8
1323 const wxWCharBuffer wc_str() const;
1324 #endif
1325 // for compatibility with !wxUSE_UNICODE version
1326 const wxWX2WCbuf wc_str(const wxMBConv& WXUNUSED(conv)) const
1327 { return wc_str(); }
1328
1329 #if wxMBFILES
1330 const wxCharBuffer fn_str() const { return mb_str(wxConvFile); }
1331 #else // !wxMBFILES
1332 const wxWX2WCbuf fn_str() const { return wc_str(); }
1333 #endif // wxMBFILES/!wxMBFILES
1334
1335 #else // ANSI
1336 const wxChar* mb_str() const { return wx_str(); }
1337
1338 // for compatibility with wxUSE_UNICODE version
1339 const wxChar* mb_str(const wxMBConv& WXUNUSED(conv)) const { return wx_str(); }
1340
1341 const wxWX2MBbuf mbc_str() const { return mb_str(); }
1342
1343 #if wxUSE_WCHAR_T
1344 const wxWCharBuffer wc_str(const wxMBConv& conv = wxConvLibc) const;
1345 #endif // wxUSE_WCHAR_T
1346 const wxCharBuffer fn_str() const { return wxConvFile.cWC2WX( wc_str( wxConvLibc ) ); }
1347 #endif // Unicode/ANSI
1348
1349 // overloaded assignment
1350 // from another wxString
1351 wxString& operator=(const wxString& stringSrc)
1352 { if (&stringSrc != this) m_impl = stringSrc.m_impl; return *this; }
1353 wxString& operator=(const wxCStrData& cstr)
1354 { return *this = cstr.AsString(); }
1355 // from a character
1356 wxString& operator=(wxUniChar ch)
1357 {
1358 #if wxUSE_UNICODE_UTF8
1359 if ( !ch.IsAscii() )
1360 m_impl = wxStringOperations::EncodeChar(ch);
1361 else
1362 #endif
1363 m_impl = (wxStringCharType)ch;
1364 return *this;
1365 }
1366 wxString& operator=(wxUniCharRef ch)
1367 { return operator=((wxUniChar)ch); }
1368 wxString& operator=(char ch)
1369 { return operator=(wxUniChar(ch)); }
1370 wxString& operator=(unsigned char ch)
1371 { return operator=(wxUniChar(ch)); }
1372 wxString& operator=(wchar_t ch)
1373 { return operator=(wxUniChar(ch)); }
1374 // from a C string - STL probably will crash on NULL,
1375 // so we need to compensate in that case
1376 #if wxUSE_STL_BASED_WXSTRING
1377 wxString& operator=(const char *psz)
1378 { if (psz) m_impl = ImplStr(psz); else Clear(); return *this; }
1379 wxString& operator=(const wchar_t *pwz)
1380 { if (pwz) m_impl = ImplStr(pwz); else Clear(); return *this; }
1381 #else
1382 wxString& operator=(const char *psz)
1383 { m_impl = ImplStr(psz); return *this; }
1384 wxString& operator=(const wchar_t *pwz)
1385 { m_impl = ImplStr(pwz); return *this; }
1386 #endif
1387 wxString& operator=(const unsigned char *psz)
1388 { return operator=((const char*)psz); }
1389
1390 // from wxWCharBuffer
1391 wxString& operator=(const wxWCharBuffer& s)
1392 { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs
1393 // from wxCharBuffer
1394 wxString& operator=(const wxCharBuffer& s)
1395 { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs
1396
1397 // string concatenation
1398 // in place concatenation
1399 /*
1400 Concatenate and return the result. Note that the left to right
1401 associativity of << allows to write things like "str << str1 << str2
1402 << ..." (unlike with +=)
1403 */
1404 // string += string
1405 wxString& operator<<(const wxString& s)
1406 {
1407 #if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
1408 wxASSERT_MSG( s.IsValid(),
1409 _T("did you forget to call UngetWriteBuf()?") );
1410 #endif
1411
1412 append(s);
1413 return *this;
1414 }
1415 // string += C string
1416 wxString& operator<<(const char *psz)
1417 { append(psz); return *this; }
1418 wxString& operator<<(const wchar_t *pwz)
1419 { append(pwz); return *this; }
1420 wxString& operator<<(const wxCStrData& psz)
1421 { append(psz.AsString()); return *this; }
1422 // string += char
1423 wxString& operator<<(wxUniChar ch) { append(1, ch); return *this; }
1424 wxString& operator<<(wxUniCharRef ch) { append(1, ch); return *this; }
1425 wxString& operator<<(char ch) { append(1, ch); return *this; }
1426 wxString& operator<<(unsigned char ch) { append(1, ch); return *this; }
1427 wxString& operator<<(wchar_t ch) { append(1, ch); return *this; }
1428
1429 // string += buffer (i.e. from wxGetString)
1430 wxString& operator<<(const wxWCharBuffer& s)
1431 { return operator<<((const wchar_t *)s); }
1432 wxString& operator<<(const wxCharBuffer& s)
1433 { return operator<<((const char *)s); }
1434
1435 // string += C string
1436 wxString& Append(const wxString& s)
1437 {
1438 // test for empty() to share the string if possible
1439 if ( empty() )
1440 *this = s;
1441 else
1442 append(s);
1443 return *this;
1444 }
1445 wxString& Append(const char* psz)
1446 { append(psz); return *this; }
1447 wxString& Append(const wchar_t* pwz)
1448 { append(pwz); return *this; }
1449 wxString& Append(const wxCStrData& psz)
1450 { append(psz); return *this; }
1451 wxString& Append(const wxCharBuffer& psz)
1452 { append(psz); return *this; }
1453 wxString& Append(const wxWCharBuffer& psz)
1454 { append(psz); return *this; }
1455 wxString& Append(const char* psz, size_t nLen)
1456 { append(psz, nLen); return *this; }
1457 wxString& Append(const wchar_t* pwz, size_t nLen)
1458 { append(pwz, nLen); return *this; }
1459 wxString& Append(const wxCStrData& psz, size_t nLen)
1460 { append(psz, nLen); return *this; }
1461 wxString& Append(const wxCharBuffer& psz, size_t nLen)
1462 { append(psz, nLen); return *this; }
1463 wxString& Append(const wxWCharBuffer& psz, size_t nLen)
1464 { append(psz, nLen); return *this; }
1465 // append count copies of given character
1466 wxString& Append(wxUniChar ch, size_t count = 1u)
1467 { append(count, ch); return *this; }
1468 wxString& Append(wxUniCharRef ch, size_t count = 1u)
1469 { append(count, ch); return *this; }
1470 wxString& Append(char ch, size_t count = 1u)
1471 { append(count, ch); return *this; }
1472 wxString& Append(unsigned char ch, size_t count = 1u)
1473 { append(count, ch); return *this; }
1474 wxString& Append(wchar_t ch, size_t count = 1u)
1475 { append(count, ch); return *this; }
1476
1477 // prepend a string, return the string itself
1478 wxString& Prepend(const wxString& str)
1479 { *this = str + *this; return *this; }
1480
1481 // non-destructive concatenation
1482 // two strings
1483 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string1,
1484 const wxString& string2);
1485 // string with a single char
1486 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxUniChar ch);
1487 // char with a string
1488 friend wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string);
1489 // string with C string
1490 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string,
1491 const char *psz);
1492 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string,
1493 const wchar_t *pwz);
1494 // C string with string
1495 friend wxString WXDLLIMPEXP_BASE operator+(const char *psz,
1496 const wxString& string);
1497 friend wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz,
1498 const wxString& string);
1499
1500 // stream-like functions
1501 // insert an int into string
1502 wxString& operator<<(int i)
1503 { return (*this) << Format(_T("%d"), i); }
1504 // insert an unsigned int into string
1505 wxString& operator<<(unsigned int ui)
1506 { return (*this) << Format(_T("%u"), ui); }
1507 // insert a long into string
1508 wxString& operator<<(long l)
1509 { return (*this) << Format(_T("%ld"), l); }
1510 // insert an unsigned long into string
1511 wxString& operator<<(unsigned long ul)
1512 { return (*this) << Format(_T("%lu"), ul); }
1513 #if defined wxLongLong_t && !defined wxLongLongIsLong
1514 // insert a long long if they exist and aren't longs
1515 wxString& operator<<(wxLongLong_t ll)
1516 {
1517 const wxChar *fmt = _T("%") wxLongLongFmtSpec _T("d");
1518 return (*this) << Format(fmt, ll);
1519 }
1520 // insert an unsigned long long
1521 wxString& operator<<(wxULongLong_t ull)
1522 {
1523 const wxChar *fmt = _T("%") wxLongLongFmtSpec _T("u");
1524 return (*this) << Format(fmt , ull);
1525 }
1526 #endif
1527 // insert a float into string
1528 wxString& operator<<(float f)
1529 { return (*this) << Format(_T("%f"), f); }
1530 // insert a double into string
1531 wxString& operator<<(double d)
1532 { return (*this) << Format(_T("%g"), d); }
1533
1534 // string comparison
1535 // case-sensitive comparison (returns a value < 0, = 0 or > 0)
1536 int Cmp(const char *psz) const
1537 { return compare(psz); }
1538 int Cmp(const wchar_t *pwz) const
1539 { return compare(pwz); }
1540 int Cmp(const wxString& s) const
1541 { return compare(s); }
1542 int Cmp(const wxCStrData& s) const
1543 { return compare(s); }
1544 int Cmp(const wxCharBuffer& s) const
1545 { return compare(s); }
1546 int Cmp(const wxWCharBuffer& s) const
1547 { return compare(s); }
1548 // same as Cmp() but not case-sensitive
1549 int CmpNoCase(const wxString& s) const;
1550
1551 // test for the string equality, either considering case or not
1552 // (if compareWithCase then the case matters)
1553 bool IsSameAs(const wxString& str, bool compareWithCase = true) const
1554 {
1555 #if !wxUSE_UNICODE_UTF8
1556 // in UTF-8 build, length() is O(n) and doing this would be _slower_
1557 if ( length() != str.length() )
1558 return false;
1559 #endif
1560 return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0;
1561 }
1562 bool IsSameAs(const char *str, bool compareWithCase = true) const
1563 { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
1564 bool IsSameAs(const wchar_t *str, bool compareWithCase = true) const
1565 { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
1566
1567 bool IsSameAs(const wxCStrData& str, bool compareWithCase = true) const
1568 { return IsSameAs(str.AsString(), compareWithCase); }
1569 bool IsSameAs(const wxCharBuffer& str, bool compareWithCase = true) const
1570 { return IsSameAs(str.data(), compareWithCase); }
1571 bool IsSameAs(const wxWCharBuffer& str, bool compareWithCase = true) const
1572 { return IsSameAs(str.data(), compareWithCase); }
1573 // comparison with a single character: returns true if equal
1574 bool IsSameAs(wxUniChar c, bool compareWithCase = true) const;
1575 // FIXME-UTF8: remove these overloads
1576 bool IsSameAs(wxUniCharRef c, bool compareWithCase = true) const
1577 { return IsSameAs(wxUniChar(c), compareWithCase); }
1578 bool IsSameAs(char c, bool compareWithCase = true) const
1579 { return IsSameAs(wxUniChar(c), compareWithCase); }
1580 bool IsSameAs(unsigned char c, bool compareWithCase = true) const
1581 { return IsSameAs(wxUniChar(c), compareWithCase); }
1582 bool IsSameAs(wchar_t c, bool compareWithCase = true) const
1583 { return IsSameAs(wxUniChar(c), compareWithCase); }
1584 bool IsSameAs(int c, bool compareWithCase = true) const
1585 { return IsSameAs(wxUniChar(c), compareWithCase); }
1586
1587 // simple sub-string extraction
1588 // return substring starting at nFirst of length nCount (or till the end
1589 // if nCount = default value)
1590 wxString Mid(size_t nFirst, size_t nCount = npos) const;
1591
1592 // operator version of Mid()
1593 wxString operator()(size_t start, size_t len) const
1594 { return Mid(start, len); }
1595
1596 // check if the string starts with the given prefix and return the rest
1597 // of the string in the provided pointer if it is not NULL; otherwise
1598 // return false
1599 bool StartsWith(const wxString& prefix, wxString *rest = NULL) const;
1600 // check if the string ends with the given suffix and return the
1601 // beginning of the string before the suffix in the provided pointer if
1602 // it is not NULL; otherwise return false
1603 bool EndsWith(const wxString& suffix, wxString *rest = NULL) const;
1604
1605 // get first nCount characters
1606 wxString Left(size_t nCount) const;
1607 // get last nCount characters
1608 wxString Right(size_t nCount) const;
1609 // get all characters before the first occurance of ch
1610 // (returns the whole string if ch not found)
1611 wxString BeforeFirst(wxUniChar ch) const;
1612 // get all characters before the last occurence of ch
1613 // (returns empty string if ch not found)
1614 wxString BeforeLast(wxUniChar ch) const;
1615 // get all characters after the first occurence of ch
1616 // (returns empty string if ch not found)
1617 wxString AfterFirst(wxUniChar ch) const;
1618 // get all characters after the last occurence of ch
1619 // (returns the whole string if ch not found)
1620 wxString AfterLast(wxUniChar ch) const;
1621
1622 // for compatibility only, use more explicitly named functions above
1623 wxString Before(wxUniChar ch) const { return BeforeLast(ch); }
1624 wxString After(wxUniChar ch) const { return AfterFirst(ch); }
1625
1626 // case conversion
1627 // convert to upper case in place, return the string itself
1628 wxString& MakeUpper();
1629 // convert to upper case, return the copy of the string
1630 // Here's something to remember: BC++ doesn't like returns in inlines.
1631 wxString Upper() const ;
1632 // convert to lower case in place, return the string itself
1633 wxString& MakeLower();
1634 // convert to lower case, return the copy of the string
1635 wxString Lower() const ;
1636
1637 // trimming/padding whitespace (either side) and truncating
1638 // remove spaces from left or from right (default) side
1639 wxString& Trim(bool bFromRight = true);
1640 // add nCount copies chPad in the beginning or at the end (default)
1641 wxString& Pad(size_t nCount, wxUniChar chPad = wxT(' '), bool bFromRight = true);
1642
1643 // searching and replacing
1644 // searching (return starting index, or -1 if not found)
1645 int Find(wxUniChar ch, bool bFromEnd = false) const; // like strchr/strrchr
1646 int Find(wxUniCharRef ch, bool bFromEnd = false) const
1647 { return Find(wxUniChar(ch), bFromEnd); }
1648 int Find(char ch, bool bFromEnd = false) const
1649 { return Find(wxUniChar(ch), bFromEnd); }
1650 int Find(unsigned char ch, bool bFromEnd = false) const
1651 { return Find(wxUniChar(ch), bFromEnd); }
1652 int Find(wchar_t ch, bool bFromEnd = false) const
1653 { return Find(wxUniChar(ch), bFromEnd); }
1654 // searching (return starting index, or -1 if not found)
1655 int Find(const wxString& sub) const // like strstr
1656 {
1657 size_type idx = find(sub);
1658 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1659 }
1660 int Find(const char *sub) const // like strstr
1661 {
1662 size_type idx = find(sub);
1663 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1664 }
1665 int Find(const wchar_t *sub) const // like strstr
1666 {
1667 size_type idx = find(sub);
1668 return (idx == npos) ? wxNOT_FOUND : (int)idx;
1669 }
1670
1671 int Find(const wxCStrData& sub) const
1672 { return Find(sub.AsString()); }
1673 int Find(const wxCharBuffer& sub) const
1674 { return Find(sub.data()); }
1675 int Find(const wxWCharBuffer& sub) const
1676 { return Find(sub.data()); }
1677
1678 // replace first (or all of bReplaceAll) occurences of substring with
1679 // another string, returns the number of replacements made
1680 size_t Replace(const wxString& strOld,
1681 const wxString& strNew,
1682 bool bReplaceAll = true);
1683
1684 // check if the string contents matches a mask containing '*' and '?'
1685 bool Matches(const wxString& mask) const;
1686
1687 // conversion to numbers: all functions return true only if the whole
1688 // string is a number and put the value of this number into the pointer
1689 // provided, the base is the numeric base in which the conversion should be
1690 // done and must be comprised between 2 and 36 or be 0 in which case the
1691 // standard C rules apply (leading '0' => octal, "0x" => hex)
1692 // convert to a signed integer
1693 bool ToLong(long *val, int base = 10) const;
1694 // convert to an unsigned integer
1695 bool ToULong(unsigned long *val, int base = 10) const;
1696 // convert to wxLongLong
1697 #if defined(wxLongLong_t)
1698 bool ToLongLong(wxLongLong_t *val, int base = 10) const;
1699 // convert to wxULongLong
1700 bool ToULongLong(wxULongLong_t *val, int base = 10) const;
1701 #endif // wxLongLong_t
1702 // convert to a double
1703 bool ToDouble(double *val) const;
1704
1705
1706 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1707 // formatted input/output
1708 // as sprintf(), returns the number of characters written or < 0 on error
1709 // (take 'this' into account in attribute parameter count)
1710 // int Printf(const wxString& format, ...);
1711 WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&),
1712 DoPrintfWchar, DoPrintfUtf8)
1713 #ifdef __WATCOMC__
1714 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1715 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxString&),
1716 (wxFormatString(f1)));
1717 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxCStrData&),
1718 (wxFormatString(f1)));
1719 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const char*),
1720 (wxFormatString(f1)));
1721 WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wchar_t*),
1722 (wxFormatString(f1)));
1723 #endif
1724 #endif // !wxNEEDS_WXSTRING_PRINTF_MIXIN
1725 // as vprintf(), returns the number of characters written or < 0 on error
1726 int PrintfV(const wxString& format, va_list argptr);
1727
1728 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1729 // returns the string containing the result of Printf() to it
1730 // static wxString Format(const wxString& format, ...) ATTRIBUTE_PRINTF_1;
1731 WX_DEFINE_VARARG_FUNC(static wxString, Format, 1, (const wxFormatString&),
1732 DoFormatWchar, DoFormatUtf8)
1733 #ifdef __WATCOMC__
1734 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1735 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxString&),
1736 (wxFormatString(f1)));
1737 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxCStrData&),
1738 (wxFormatString(f1)));
1739 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const char*),
1740 (wxFormatString(f1)));
1741 WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wchar_t*),
1742 (wxFormatString(f1)));
1743 #endif
1744 #endif
1745 // the same as above, but takes a va_list
1746 static wxString FormatV(const wxString& format, va_list argptr);
1747
1748 // raw access to string memory
1749 // ensure that string has space for at least nLen characters
1750 // only works if the data of this string is not shared
1751 bool Alloc(size_t nLen) { reserve(nLen); return capacity() >= nLen; }
1752 // minimize the string's memory
1753 // only works if the data of this string is not shared
1754 bool Shrink();
1755 #if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
1756 // These are deprecated, use wxStringBuffer or wxStringBufferLength instead
1757 //
1758 // get writable buffer of at least nLen bytes. Unget() *must* be called
1759 // a.s.a.p. to put string back in a reasonable state!
1760 wxDEPRECATED( wxStringCharType *GetWriteBuf(size_t nLen) );
1761 // call this immediately after GetWriteBuf() has been used
1762 wxDEPRECATED( void UngetWriteBuf() );
1763 wxDEPRECATED( void UngetWriteBuf(size_t nLen) );
1764 #endif // WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && wxUSE_UNICODE_UTF8
1765
1766 // wxWidgets version 1 compatibility functions
1767
1768 // use Mid()
1769 wxString SubString(size_t from, size_t to) const
1770 { return Mid(from, (to - from + 1)); }
1771 // values for second parameter of CompareTo function
1772 enum caseCompare {exact, ignoreCase};
1773 // values for first parameter of Strip function
1774 enum stripType {leading = 0x1, trailing = 0x2, both = 0x3};
1775
1776 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
1777 // use Printf()
1778 // (take 'this' into account in attribute parameter count)
1779 // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2;
1780 WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&),
1781 DoPrintfWchar, DoPrintfUtf8)
1782 #ifdef __WATCOMC__
1783 // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351
1784 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxString&),
1785 (wxFormatString(f1)));
1786 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxCStrData&),
1787 (wxFormatString(f1)));
1788 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const char*),
1789 (wxFormatString(f1)));
1790 WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wchar_t*),
1791 (wxFormatString(f1)));
1792 #endif
1793 #endif // wxNEEDS_WXSTRING_PRINTF_MIXIN
1794
1795 // use Cmp()
1796 inline int CompareTo(const wxChar* psz, caseCompare cmp = exact) const
1797 { return cmp == exact ? Cmp(psz) : CmpNoCase(psz); }
1798
1799 // use length()
1800 size_t Length() const { return length(); }
1801 // Count the number of characters
1802 int Freq(wxUniChar ch) const;
1803 // use MakeLower
1804 void LowerCase() { MakeLower(); }
1805 // use MakeUpper
1806 void UpperCase() { MakeUpper(); }
1807 // use Trim except that it doesn't change this string
1808 wxString Strip(stripType w = trailing) const;
1809
1810 // use Find (more general variants not yet supported)
1811 size_t Index(const wxChar* psz) const { return Find(psz); }
1812 size_t Index(wxUniChar ch) const { return Find(ch); }
1813 // use Truncate
1814 wxString& Remove(size_t pos) { return Truncate(pos); }
1815 wxString& RemoveLast(size_t n = 1) { return Truncate(length() - n); }
1816
1817 wxString& Remove(size_t nStart, size_t nLen)
1818 { return (wxString&)erase( nStart, nLen ); }
1819
1820 // use Find()
1821 int First( wxUniChar ch ) const { return Find(ch); }
1822 int First( wxUniCharRef ch ) const { return Find(ch); }
1823 int First( char ch ) const { return Find(ch); }
1824 int First( unsigned char ch ) const { return Find(ch); }
1825 int First( wchar_t ch ) const { return Find(ch); }
1826 int First( const wxString& str ) const { return Find(str); }
1827 int Last( wxUniChar ch ) const { return Find(ch, true); }
1828 bool Contains(const wxString& str) const { return Find(str) != wxNOT_FOUND; }
1829
1830 // use empty()
1831 bool IsNull() const { return empty(); }
1832
1833 // std::string compatibility functions
1834
1835 // take nLen chars starting at nPos
1836 wxString(const wxString& str, size_t nPos, size_t nLen)
1837 { assign(str, nPos, nLen); }
1838 // take all characters from first to last
1839 wxString(const_iterator first, const_iterator last)
1840 : m_impl(first.impl(), last.impl()) { }
1841 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1842 // the 2 overloads below are for compatibility with the existing code using
1843 // pointers instead of iterators
1844 wxString(const char *first, const char *last)
1845 {
1846 SubstrBufFromMB str(ImplStr(first, last - first));
1847 m_impl.assign(str.data, str.len);
1848 }
1849 wxString(const wchar_t *first, const wchar_t *last)
1850 {
1851 SubstrBufFromWC str(ImplStr(first, last - first));
1852 m_impl.assign(str.data, str.len);
1853 }
1854 // and this one is needed to compile code adding offsets to c_str() result
1855 wxString(const wxCStrData& first, const wxCStrData& last)
1856 : m_impl(CreateConstIterator(first).impl(),
1857 CreateConstIterator(last).impl())
1858 {
1859 wxASSERT_MSG( first.m_str == last.m_str,
1860 _T("pointers must be into the same string") );
1861 }
1862 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1863
1864 // lib.string.modifiers
1865 // append elements str[pos], ..., str[pos+n]
1866 wxString& append(const wxString& str, size_t pos, size_t n)
1867 {
1868 size_t from, len;
1869 str.PosLenToImpl(pos, n, &from, &len);
1870 m_impl.append(str.m_impl, from, len);
1871 return *this;
1872 }
1873 // append a string
1874 wxString& append(const wxString& str)
1875 { m_impl.append(str.m_impl); return *this; }
1876 // append first n (or all if n == npos) characters of sz
1877 wxString& append(const char *sz)
1878 { m_impl.append(ImplStr(sz)); return *this; }
1879 wxString& append(const wchar_t *sz)
1880 { m_impl.append(ImplStr(sz)); return *this; }
1881 wxString& append(const char *sz, size_t n)
1882 {
1883 SubstrBufFromMB str(ImplStr(sz, n));
1884 m_impl.append(str.data, str.len);
1885 return *this;
1886 }
1887 wxString& append(const wchar_t *sz, size_t n)
1888 {
1889 SubstrBufFromWC str(ImplStr(sz, n));
1890 m_impl.append(str.data, str.len);
1891 return *this;
1892 }
1893
1894 wxString& append(const wxCStrData& str)
1895 { return append(str.AsString()); }
1896 wxString& append(const wxCharBuffer& str)
1897 { return append(str.data()); }
1898 wxString& append(const wxWCharBuffer& str)
1899 { return append(str.data()); }
1900 wxString& append(const wxCStrData& str, size_t n)
1901 { return append(str.AsString(), 0, n); }
1902 wxString& append(const wxCharBuffer& str, size_t n)
1903 { return append(str.data(), n); }
1904 wxString& append(const wxWCharBuffer& str, size_t n)
1905 { return append(str.data(), n); }
1906
1907 // append n copies of ch
1908 wxString& append(size_t n, wxUniChar ch)
1909 {
1910 #if wxUSE_UNICODE_UTF8
1911 if ( !ch.IsAscii() )
1912 m_impl.append(wxStringOperations::EncodeNChars(n, ch));
1913 else
1914 #endif
1915 m_impl.append(n, (wxStringCharType)ch);
1916 return *this;
1917 }
1918 wxString& append(size_t n, wxUniCharRef ch)
1919 { return append(n, wxUniChar(ch)); }
1920 wxString& append(size_t n, char ch)
1921 { return append(n, wxUniChar(ch)); }
1922 wxString& append(size_t n, unsigned char ch)
1923 { return append(n, wxUniChar(ch)); }
1924 wxString& append(size_t n, wchar_t ch)
1925 { return append(n, wxUniChar(ch)); }
1926
1927 // append from first to last
1928 wxString& append(const_iterator first, const_iterator last)
1929 { m_impl.append(first.impl(), last.impl()); return *this; }
1930 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1931 wxString& append(const char *first, const char *last)
1932 { return append(first, last - first); }
1933 wxString& append(const wchar_t *first, const wchar_t *last)
1934 { return append(first, last - first); }
1935 wxString& append(const wxCStrData& first, const wxCStrData& last)
1936 { return append(CreateConstIterator(first), CreateConstIterator(last)); }
1937 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
1938
1939 // same as `this_string = str'
1940 wxString& assign(const wxString& str)
1941 { m_impl = str.m_impl; return *this; }
1942 wxString& assign(const wxString& str, size_t len)
1943 {
1944 m_impl.assign(str.m_impl, 0, str.LenToImpl(len));
1945 return *this;
1946 }
1947 // same as ` = str[pos..pos + n]
1948 wxString& assign(const wxString& str, size_t pos, size_t n)
1949 {
1950 size_t from, len;
1951 str.PosLenToImpl(pos, n, &from, &len);
1952 m_impl.assign(str.m_impl, from, len);
1953 return *this;
1954 }
1955 // same as `= first n (or all if n == npos) characters of sz'
1956 wxString& assign(const char *sz)
1957 { m_impl.assign(ImplStr(sz)); return *this; }
1958 wxString& assign(const wchar_t *sz)
1959 { m_impl.assign(ImplStr(sz)); return *this; }
1960 wxString& assign(const char *sz, size_t n)
1961 {
1962 SubstrBufFromMB str(ImplStr(sz, n));
1963 m_impl.assign(str.data, str.len);
1964 return *this;
1965 }
1966 wxString& assign(const wchar_t *sz, size_t n)
1967 {
1968 SubstrBufFromWC str(ImplStr(sz, n));
1969 m_impl.assign(str.data, str.len);
1970 return *this;
1971 }
1972
1973 wxString& assign(const wxCStrData& str)
1974 { return assign(str.AsString()); }
1975 wxString& assign(const wxCharBuffer& str)
1976 { return assign(str.data()); }
1977 wxString& assign(const wxWCharBuffer& str)
1978 { return assign(str.data()); }
1979 wxString& assign(const wxCStrData& str, size_t len)
1980 { return assign(str.AsString(), len); }
1981 wxString& assign(const wxCharBuffer& str, size_t len)
1982 { return assign(str.data(), len); }
1983 wxString& assign(const wxWCharBuffer& str, size_t len)
1984 { return assign(str.data(), len); }
1985
1986 // same as `= n copies of ch'
1987 wxString& assign(size_t n, wxUniChar ch)
1988 {
1989 #if wxUSE_UNICODE_UTF8
1990 if ( !ch.IsAscii() )
1991 m_impl.assign(wxStringOperations::EncodeNChars(n, ch));
1992 else
1993 #endif
1994 m_impl.assign(n, (wxStringCharType)ch);
1995 return *this;
1996 }
1997
1998 wxString& assign(size_t n, wxUniCharRef ch)
1999 { return assign(n, wxUniChar(ch)); }
2000 wxString& assign(size_t n, char ch)
2001 { return assign(n, wxUniChar(ch)); }
2002 wxString& assign(size_t n, unsigned char ch)
2003 { return assign(n, wxUniChar(ch)); }
2004 wxString& assign(size_t n, wchar_t ch)
2005 { return assign(n, wxUniChar(ch)); }
2006
2007 // assign from first to last
2008 wxString& assign(const_iterator first, const_iterator last)
2009 { m_impl.assign(first.impl(), last.impl()); return *this; }
2010 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2011 wxString& assign(const char *first, const char *last)
2012 { return assign(first, last - first); }
2013 wxString& assign(const wchar_t *first, const wchar_t *last)
2014 { return assign(first, last - first); }
2015 wxString& assign(const wxCStrData& first, const wxCStrData& last)
2016 { return assign(CreateConstIterator(first), CreateConstIterator(last)); }
2017 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2018
2019 // string comparison
2020 int compare(const wxString& str) const;
2021 int compare(const char* sz) const;
2022 int compare(const wchar_t* sz) const;
2023 int compare(const wxCStrData& str) const
2024 { return compare(str.AsString()); }
2025 int compare(const wxCharBuffer& str) const
2026 { return compare(str.data()); }
2027 int compare(const wxWCharBuffer& str) const
2028 { return compare(str.data()); }
2029 // comparison with a substring
2030 int compare(size_t nStart, size_t nLen, const wxString& str) const;
2031 // comparison of 2 substrings
2032 int compare(size_t nStart, size_t nLen,
2033 const wxString& str, size_t nStart2, size_t nLen2) const;
2034 // substring comparison with first nCount characters of sz
2035 int compare(size_t nStart, size_t nLen,
2036 const char* sz, size_t nCount = npos) const;
2037 int compare(size_t nStart, size_t nLen,
2038 const wchar_t* sz, size_t nCount = npos) const;
2039
2040 // insert another string
2041 wxString& insert(size_t nPos, const wxString& str)
2042 { insert(begin() + nPos, str.begin(), str.end()); return *this; }
2043 // insert n chars of str starting at nStart (in str)
2044 wxString& insert(size_t nPos, const wxString& str, size_t nStart, size_t n)
2045 {
2046 size_t from, len;
2047 str.PosLenToImpl(nStart, n, &from, &len);
2048 m_impl.insert(PosToImpl(nPos), str.m_impl, from, len);
2049 return *this;
2050 }
2051 // insert first n (or all if n == npos) characters of sz
2052 wxString& insert(size_t nPos, const char *sz)
2053 { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; }
2054 wxString& insert(size_t nPos, const wchar_t *sz)
2055 { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; }
2056 wxString& insert(size_t nPos, const char *sz, size_t n)
2057 {
2058 SubstrBufFromMB str(ImplStr(sz, n));
2059 m_impl.insert(PosToImpl(nPos), str.data, str.len);
2060 return *this;
2061 }
2062 wxString& insert(size_t nPos, const wchar_t *sz, size_t n)
2063 {
2064 SubstrBufFromWC str(ImplStr(sz, n));
2065 m_impl.insert(PosToImpl(nPos), str.data, str.len);
2066 return *this;
2067 }
2068 // insert n copies of ch
2069 wxString& insert(size_t nPos, size_t n, wxUniChar ch)
2070 {
2071 #if wxUSE_UNICODE_UTF8
2072 if ( !ch.IsAscii() )
2073 m_impl.insert(PosToImpl(nPos), wxStringOperations::EncodeNChars(n, ch));
2074 else
2075 #endif
2076 m_impl.insert(PosToImpl(nPos), n, (wxStringCharType)ch);
2077 return *this;
2078 }
2079 iterator insert(iterator it, wxUniChar ch)
2080 {
2081 #if wxUSE_UNICODE_UTF8
2082 if ( !ch.IsAscii() )
2083 {
2084 size_t pos = IterToImplPos(it);
2085 m_impl.insert(pos, wxStringOperations::EncodeChar(ch));
2086 return iterator(this, m_impl.begin() + pos);
2087 }
2088 else
2089 #endif
2090 return iterator(this, m_impl.insert(it.impl(), (wxStringCharType)ch));
2091 }
2092 void insert(iterator it, const_iterator first, const_iterator last)
2093 { m_impl.insert(it.impl(), first.impl(), last.impl()); }
2094 #if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2095 void insert(iterator it, const char *first, const char *last)
2096 { insert(it - begin(), first, last - first); }
2097 void insert(iterator it, const wchar_t *first, const wchar_t *last)
2098 { insert(it - begin(), first, last - first); }
2099 void insert(iterator it, const wxCStrData& first, const wxCStrData& last)
2100 { insert(it, CreateConstIterator(first), CreateConstIterator(last)); }
2101 #endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER
2102
2103 void insert(iterator it, size_type n, wxUniChar ch)
2104 {
2105 #if wxUSE_UNICODE_UTF8
2106 if ( !ch.IsAscii() )
2107 m_impl.insert(IterToImplPos(it), wxStringOperations::EncodeNChars(n, ch));
2108 else
2109 #endif
2110 m_impl.insert(it.impl(), n, (wxStringCharType)ch);
2111 }
2112
2113 // delete characters from nStart to nStart + nLen
2114 wxString& erase(size_type pos = 0, size_type n = npos)
2115 {
2116 size_t from, len;
2117 PosLenToImpl(pos, n, &from, &len);
2118 m_impl.erase(from, len);
2119 return *this;
2120 }
2121 // delete characters from first up to last
2122 iterator erase(iterator first, iterator last)
2123 { return iterator(this, m_impl.erase(first.impl(), last.impl())); }
2124 iterator erase(iterator first)
2125 { return iterator(this, m_impl.erase(first.impl())); }
2126
2127 #ifdef wxSTRING_BASE_HASNT_CLEAR
2128 void clear() { erase(); }
2129 #else
2130 void clear() { m_impl.clear(); }
2131 #endif
2132
2133 // replaces the substring of length nLen starting at nStart
2134 wxString& replace(size_t nStart, size_t nLen, const char* sz)
2135 {
2136 size_t from, len;
2137 PosLenToImpl(nStart, nLen, &from, &len);
2138 m_impl.replace(from, len, ImplStr(sz));
2139 return *this;
2140 }
2141 wxString& replace(size_t nStart, size_t nLen, const wchar_t* sz)
2142 {
2143 size_t from, len;
2144 PosLenToImpl(nStart, nLen, &from, &len);
2145 m_impl.replace(from, len, ImplStr(sz));
2146 return *this;
2147 }
2148 // replaces the substring of length nLen starting at nStart
2149 wxString& replace(size_t nStart, size_t nLen, const wxString& str)
2150 {
2151 size_t from, len;
2152 PosLenToImpl(nStart, nLen, &from, &len);
2153 m_impl.replace(from, len, str.m_impl);
2154 return *this;
2155 }
2156 // replaces the substring with nCount copies of ch
2157 wxString& replace(size_t nStart, size_t nLen, size_t nCount, wxUniChar ch)
2158 {
2159 size_t from, len;
2160 PosLenToImpl(nStart, nLen, &from, &len);
2161 #if wxUSE_UNICODE_UTF8
2162 if ( !ch.IsAscii() )
2163 m_impl.replace(from, len, wxStringOperations::EncodeNChars(nCount, ch));
2164 else
2165 #endif
2166 m_impl.replace(from, len, nCount, (wxStringCharType)ch);
2167 return *this;
2168 }
2169 // replaces a substring with another substring
2170 wxString& replace(size_t nStart, size_t nLen,
2171 const wxString& str, size_t nStart2, size_t nLen2)
2172 {
2173 size_t from, len;
2174 PosLenToImpl(nStart, nLen, &from, &len);
2175
2176 size_t from2, len2;
2177 str.PosLenToImpl(nStart2, nLen2, &from2, &len2);
2178
2179 m_impl.replace(from, len, str.m_impl, from2, len2);
2180 return *this;
2181 }
2182 // replaces the substring with first nCount chars of sz
2183 wxString& replace(size_t nStart, size_t nLen,
2184 const char* sz, size_t nCount)
2185 {
2186 size_t from, len;
2187 PosLenToImpl(nStart, nLen, &from, &len);
2188
2189 SubstrBufFromMB str(ImplStr(sz, nCount));
2190
2191 m_impl.replace(from, len, str.data, str.len);
2192 return *this;
2193 }
2194 wxString& replace(size_t nStart, size_t nLen,
2195 const wchar_t* sz, size_t nCount)
2196 {
2197 size_t from, len;
2198 PosLenToImpl(nStart, nLen, &from, &len);
2199
2200 SubstrBufFromWC str(ImplStr(sz, nCount));
2201
2202 m_impl.replace(from, len, str.data, str.len);
2203 return *this;
2204 }
2205 wxString& replace(size_t nStart, size_t nLen,
2206 const wxString& s, size_t nCount)
2207 {
2208 size_t from, len;
2209 PosLenToImpl(nStart, nLen, &from, &len);
2210 m_impl.replace(from, len, s.m_impl.c_str(), s.LenToImpl(nCount));
2211 return *this;
2212 }
2213
2214 wxString& replace(iterator first, iterator last, const char* s)
2215 { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; }
2216 wxString& replace(iterator first, iterator last, const wchar_t* s)
2217 { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; }
2218 wxString& replace(iterator first, iterator last, const char* s, size_type n)
2219 {
2220 SubstrBufFromMB str(ImplStr(s, n));
2221 m_impl.replace(first.impl(), last.impl(), str.data, str.len);
2222 return *this;
2223 }
2224 wxString& replace(iterator first, iterator last, const wchar_t* s, size_type n)
2225 {
2226 SubstrBufFromWC str(ImplStr(s, n));
2227 m_impl.replace(first.impl(), last.impl(), str.data, str.len);
2228 return *this;
2229 }
2230 wxString& replace(iterator first, iterator last, const wxString& s)
2231 { m_impl.replace(first.impl(), last.impl(), s.m_impl); return *this; }
2232 wxString& replace(iterator first, iterator last, size_type n, wxUniChar ch)
2233 {
2234 #if wxUSE_UNICODE_UTF8
2235 if ( !ch.IsAscii() )
2236 m_impl.replace(first.impl(), last.impl(),
2237 wxStringOperations::EncodeNChars(n, ch));
2238 else
2239 #endif
2240 m_impl.replace(first.impl(), last.impl(), n, (wxStringCharType)ch);
2241 return *this;
2242 }
2243 wxString& replace(iterator first, iterator last,
2244 const_iterator first1, const_iterator last1)
2245 {
2246 m_impl.replace(first.impl(), last.impl(), first1.impl(), last1.impl());
2247 return *this;
2248 }
2249 wxString& replace(iterator first, iterator last,
2250 const char *first1, const char *last1)
2251 { replace(first, last, first1, last1 - first1); return *this; }
2252 wxString& replace(iterator first, iterator last,
2253 const wchar_t *first1, const wchar_t *last1)
2254 { replace(first, last, first1, last1 - first1); return *this; }
2255
2256 // swap two strings
2257 void swap(wxString& str)
2258 { m_impl.swap(str.m_impl); }
2259
2260 // find a substring
2261 size_t find(const wxString& str, size_t nStart = 0) const
2262 { return PosFromImpl(m_impl.find(str.m_impl, PosToImpl(nStart))); }
2263
2264 // find first n characters of sz
2265 size_t find(const char* sz, size_t nStart = 0, size_t n = npos) const
2266 {
2267 SubstrBufFromMB str(ImplStr(sz, n));
2268 return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len));
2269 }
2270 size_t find(const wchar_t* sz, size_t nStart = 0, size_t n = npos) const
2271 {
2272 SubstrBufFromWC str(ImplStr(sz, n));
2273 return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len));
2274 }
2275 size_t find(const wxCharBuffer& s, size_t nStart = 0, size_t n = npos) const
2276 { return find(s.data(), nStart, n); }
2277 size_t find(const wxWCharBuffer& s, size_t nStart = 0, size_t n = npos) const
2278 { return find(s.data(), nStart, n); }
2279 size_t find(const wxCStrData& s, size_t nStart = 0, size_t n = npos) const
2280 { return find(s.AsWChar(), nStart, n); }
2281
2282 // find the first occurence of character ch after nStart
2283 size_t find(wxUniChar ch, size_t nStart = 0) const
2284 {
2285 #if wxUSE_UNICODE_UTF8
2286 if ( !ch.IsAscii() )
2287 return PosFromImpl(m_impl.find(wxStringOperations::EncodeChar(ch),
2288 PosToImpl(nStart)));
2289 else
2290 #endif
2291 return PosFromImpl(m_impl.find((wxStringCharType)ch,
2292 PosToImpl(nStart)));
2293
2294 }
2295 size_t find(wxUniCharRef ch, size_t nStart = 0) const
2296 { return find(wxUniChar(ch), nStart); }
2297 size_t find(char ch, size_t nStart = 0) const
2298 { return find(wxUniChar(ch), nStart); }
2299 size_t find(unsigned char ch, size_t nStart = 0) const
2300 { return find(wxUniChar(ch), nStart); }
2301 size_t find(wchar_t ch, size_t nStart = 0) const
2302 { return find(wxUniChar(ch), nStart); }
2303
2304 // rfind() family is exactly like find() but works right to left
2305
2306 // as find, but from the end
2307 size_t rfind(const wxString& str, size_t nStart = npos) const
2308 { return PosFromImpl(m_impl.rfind(str.m_impl, PosToImpl(nStart))); }
2309
2310 // as find, but from the end
2311 size_t rfind(const char* sz, size_t nStart = npos, size_t n = npos) const
2312 {
2313 SubstrBufFromMB str(ImplStr(sz, n));
2314 return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len));
2315 }
2316 size_t rfind(const wchar_t* sz, size_t nStart = npos, size_t n = npos) const
2317 {
2318 SubstrBufFromWC str(ImplStr(sz, n));
2319 return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len));
2320 }
2321 size_t rfind(const wxCharBuffer& s, size_t nStart = npos, size_t n = npos) const
2322 { return rfind(s.data(), nStart, n); }
2323 size_t rfind(const wxWCharBuffer& s, size_t nStart = npos, size_t n = npos) const
2324 { return rfind(s.data(), nStart, n); }
2325 size_t rfind(const wxCStrData& s, size_t nStart = npos, size_t n = npos) const
2326 { return rfind(s.AsWChar(), nStart, n); }
2327 // as find, but from the end
2328 size_t rfind(wxUniChar ch, size_t nStart = npos) const
2329 {
2330 #if wxUSE_UNICODE_UTF8
2331 if ( !ch.IsAscii() )
2332 return PosFromImpl(m_impl.rfind(wxStringOperations::EncodeChar(ch),
2333 PosToImpl(nStart)));
2334 else
2335 #endif
2336 return PosFromImpl(m_impl.rfind((wxStringCharType)ch,
2337 PosToImpl(nStart)));
2338 }
2339 size_t rfind(wxUniCharRef ch, size_t nStart = npos) const
2340 { return rfind(wxUniChar(ch), nStart); }
2341 size_t rfind(char ch, size_t nStart = npos) const
2342 { return rfind(wxUniChar(ch), nStart); }
2343 size_t rfind(unsigned char ch, size_t nStart = npos) const
2344 { return rfind(wxUniChar(ch), nStart); }
2345 size_t rfind(wchar_t ch, size_t nStart = npos) const
2346 { return rfind(wxUniChar(ch), nStart); }
2347
2348 // find first/last occurence of any character (not) in the set:
2349 #if wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2350 // FIXME-UTF8: this is not entirely correct, because it doesn't work if
2351 // sizeof(wchar_t)==2 and surrogates are present in the string;
2352 // should we care? Probably not.
2353 size_t find_first_of(const wxString& str, size_t nStart = 0) const
2354 { return m_impl.find_first_of(str.m_impl, nStart); }
2355 size_t find_first_of(const char* sz, size_t nStart = 0) const
2356 { return m_impl.find_first_of(ImplStr(sz), nStart); }
2357 size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const
2358 { return m_impl.find_first_of(ImplStr(sz), nStart); }
2359 size_t find_first_of(const char* sz, size_t nStart, size_t n) const
2360 { return m_impl.find_first_of(ImplStr(sz), nStart, n); }
2361 size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const
2362 { return m_impl.find_first_of(ImplStr(sz), nStart, n); }
2363 size_t find_first_of(wxUniChar c, size_t nStart = 0) const
2364 { return m_impl.find_first_of((wxChar)c, nStart); }
2365
2366 size_t find_last_of(const wxString& str, size_t nStart = npos) const
2367 { return m_impl.find_last_of(str.m_impl, nStart); }
2368 size_t find_last_of(const char* sz, size_t nStart = npos) const
2369 { return m_impl.find_last_of(ImplStr(sz), nStart); }
2370 size_t find_last_of(const wchar_t* sz, size_t nStart = npos) const
2371 { return m_impl.find_last_of(ImplStr(sz), nStart); }
2372 size_t find_last_of(const char* sz, size_t nStart, size_t n) const
2373 { return m_impl.find_last_of(ImplStr(sz), nStart, n); }
2374 size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const
2375 { return m_impl.find_last_of(ImplStr(sz), nStart, n); }
2376 size_t find_last_of(wxUniChar c, size_t nStart = npos) const
2377 { return m_impl.find_last_of((wxChar)c, nStart); }
2378
2379 size_t find_first_not_of(const wxString& str, size_t nStart = 0) const
2380 { return m_impl.find_first_not_of(str.m_impl, nStart); }
2381 size_t find_first_not_of(const char* sz, size_t nStart = 0) const
2382 { return m_impl.find_first_not_of(ImplStr(sz), nStart); }
2383 size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const
2384 { return m_impl.find_first_not_of(ImplStr(sz), nStart); }
2385 size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const
2386 { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); }
2387 size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const
2388 { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); }
2389 size_t find_first_not_of(wxUniChar c, size_t nStart = 0) const
2390 { return m_impl.find_first_not_of((wxChar)c, nStart); }
2391
2392 size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
2393 { return m_impl.find_last_not_of(str.m_impl, nStart); }
2394 size_t find_last_not_of(const char* sz, size_t nStart = npos) const
2395 { return m_impl.find_last_not_of(ImplStr(sz), nStart); }
2396 size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const
2397 { return m_impl.find_last_not_of(ImplStr(sz), nStart); }
2398 size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const
2399 { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); }
2400 size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const
2401 { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); }
2402 size_t find_last_not_of(wxUniChar c, size_t nStart = npos) const
2403 { return m_impl.find_last_not_of((wxChar)c, nStart); }
2404 #else
2405 // we can't use std::string implementation in UTF-8 build, because the
2406 // character sets would be interpreted wrongly:
2407
2408 // as strpbrk() but starts at nStart, returns npos if not found
2409 size_t find_first_of(const wxString& str, size_t nStart = 0) const
2410 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2411 { return find_first_of(str.wc_str(), nStart); }
2412 #else
2413 { return find_first_of(str.mb_str(), nStart); }
2414 #endif
2415 // same as above
2416 size_t find_first_of(const char* sz, size_t nStart = 0) const;
2417 size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const;
2418 size_t find_first_of(const char* sz, size_t nStart, size_t n) const;
2419 size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const;
2420 // same as find(char, size_t)
2421 size_t find_first_of(wxUniChar c, size_t nStart = 0) const
2422 { return find(c, nStart); }
2423 // find the last (starting from nStart) char from str in this string
2424 size_t find_last_of (const wxString& str, size_t nStart = npos) const
2425 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2426 { return find_last_of(str.wc_str(), nStart); }
2427 #else
2428 { return find_last_of(str.mb_str(), nStart); }
2429 #endif
2430 // same as above
2431 size_t find_last_of (const char* sz, size_t nStart = npos) const;
2432 size_t find_last_of (const wchar_t* sz, size_t nStart = npos) const;
2433 size_t find_last_of(const char* sz, size_t nStart, size_t n) const;
2434 size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const;
2435 // same as above
2436 size_t find_last_of(wxUniChar c, size_t nStart = npos) const
2437 { return rfind(c, nStart); }
2438
2439 // find first/last occurence of any character not in the set
2440
2441 // as strspn() (starting from nStart), returns npos on failure
2442 size_t find_first_not_of(const wxString& str, size_t nStart = 0) const
2443 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2444 { return find_first_not_of(str.wc_str(), nStart); }
2445 #else
2446 { return find_first_not_of(str.mb_str(), nStart); }
2447 #endif
2448 // same as above
2449 size_t find_first_not_of(const char* sz, size_t nStart = 0) const;
2450 size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const;
2451 size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const;
2452 size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const;
2453 // same as above
2454 size_t find_first_not_of(wxUniChar ch, size_t nStart = 0) const;
2455 // as strcspn()
2456 size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
2457 #if wxUSE_UNICODE // FIXME-UTF8: temporary
2458 { return find_last_not_of(str.wc_str(), nStart); }
2459 #else
2460 { return find_last_not_of(str.mb_str(), nStart); }
2461 #endif
2462 // same as above
2463 size_t find_last_not_of(const char* sz, size_t nStart = npos) const;
2464 size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const;
2465 size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const;
2466 size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const;
2467 // same as above
2468 size_t find_last_not_of(wxUniChar ch, size_t nStart = npos) const;
2469 #endif // wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 or not
2470
2471 // provide char/wchar_t/wxUniCharRef overloads for char-finding functions
2472 // above to resolve ambiguities:
2473 size_t find_first_of(wxUniCharRef ch, size_t nStart = 0) const
2474 { return find_first_of(wxUniChar(ch), nStart); }
2475 size_t find_first_of(char ch, size_t nStart = 0) const
2476 { return find_first_of(wxUniChar(ch), nStart); }
2477 size_t find_first_of(unsigned char ch, size_t nStart = 0) const
2478 { return find_first_of(wxUniChar(ch), nStart); }
2479 size_t find_first_of(wchar_t ch, size_t nStart = 0) const
2480 { return find_first_of(wxUniChar(ch), nStart); }
2481 size_t find_last_of(wxUniCharRef ch, size_t nStart = npos) const
2482 { return find_last_of(wxUniChar(ch), nStart); }
2483 size_t find_last_of(char ch, size_t nStart = npos) const
2484 { return find_last_of(wxUniChar(ch), nStart); }
2485 size_t find_last_of(unsigned char ch, size_t nStart = npos) const
2486 { return find_last_of(wxUniChar(ch), nStart); }
2487 size_t find_last_of(wchar_t ch, size_t nStart = npos) const
2488 { return find_last_of(wxUniChar(ch), nStart); }
2489 size_t find_first_not_of(wxUniCharRef ch, size_t nStart = 0) const
2490 { return find_first_not_of(wxUniChar(ch), nStart); }
2491 size_t find_first_not_of(char ch, size_t nStart = 0) const
2492 { return find_first_not_of(wxUniChar(ch), nStart); }
2493 size_t find_first_not_of(unsigned char ch, size_t nStart = 0) const
2494 { return find_first_not_of(wxUniChar(ch), nStart); }
2495 size_t find_first_not_of(wchar_t ch, size_t nStart = 0) const
2496 { return find_first_not_of(wxUniChar(ch), nStart); }
2497 size_t find_last_not_of(wxUniCharRef ch, size_t nStart = npos) const
2498 { return find_last_not_of(wxUniChar(ch), nStart); }
2499 size_t find_last_not_of(char ch, size_t nStart = npos) const
2500 { return find_last_not_of(wxUniChar(ch), nStart); }
2501 size_t find_last_not_of(unsigned char ch, size_t nStart = npos) const
2502 { return find_last_not_of(wxUniChar(ch), nStart); }
2503 size_t find_last_not_of(wchar_t ch, size_t nStart = npos) const
2504 { return find_last_not_of(wxUniChar(ch), nStart); }
2505
2506 // and additional overloads for the versions taking strings:
2507 size_t find_first_of(const wxCStrData& sz, size_t nStart = 0) const
2508 { return find_first_of(sz.AsString(), nStart); }
2509 size_t find_first_of(const wxCharBuffer& sz, size_t nStart = 0) const
2510 { return find_first_of(sz.data(), nStart); }
2511 size_t find_first_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2512 { return find_first_of(sz.data(), nStart); }
2513 size_t find_first_of(const wxCStrData& sz, size_t nStart, size_t n) const
2514 { return find_first_of(sz.AsWChar(), nStart, n); }
2515 size_t find_first_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2516 { return find_first_of(sz.data(), nStart, n); }
2517 size_t find_first_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2518 { return find_first_of(sz.data(), nStart, n); }
2519
2520 size_t find_last_of(const wxCStrData& sz, size_t nStart = 0) const
2521 { return find_last_of(sz.AsString(), nStart); }
2522 size_t find_last_of(const wxCharBuffer& sz, size_t nStart = 0) const
2523 { return find_last_of(sz.data(), nStart); }
2524 size_t find_last_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2525 { return find_last_of(sz.data(), nStart); }
2526 size_t find_last_of(const wxCStrData& sz, size_t nStart, size_t n) const
2527 { return find_last_of(sz.AsWChar(), nStart, n); }
2528 size_t find_last_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2529 { return find_last_of(sz.data(), nStart, n); }
2530 size_t find_last_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2531 { return find_last_of(sz.data(), nStart, n); }
2532
2533 size_t find_first_not_of(const wxCStrData& sz, size_t nStart = 0) const
2534 { return find_first_not_of(sz.AsString(), nStart); }
2535 size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart = 0) const
2536 { return find_first_not_of(sz.data(), nStart); }
2537 size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2538 { return find_first_not_of(sz.data(), nStart); }
2539 size_t find_first_not_of(const wxCStrData& sz, size_t nStart, size_t n) const
2540 { return find_first_not_of(sz.AsWChar(), nStart, n); }
2541 size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2542 { return find_first_not_of(sz.data(), nStart, n); }
2543 size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2544 { return find_first_not_of(sz.data(), nStart, n); }
2545
2546 size_t find_last_not_of(const wxCStrData& sz, size_t nStart = 0) const
2547 { return find_last_not_of(sz.AsString(), nStart); }
2548 size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart = 0) const
2549 { return find_last_not_of(sz.data(), nStart); }
2550 size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const
2551 { return find_last_not_of(sz.data(), nStart); }
2552 size_t find_last_not_of(const wxCStrData& sz, size_t nStart, size_t n) const
2553 { return find_last_not_of(sz.AsWChar(), nStart, n); }
2554 size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const
2555 { return find_last_not_of(sz.data(), nStart, n); }
2556 size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const
2557 { return find_last_not_of(sz.data(), nStart, n); }
2558
2559 // string += string
2560 wxString& operator+=(const wxString& s)
2561 { m_impl += s.m_impl; return *this; }
2562 // string += C string
2563 wxString& operator+=(const char *psz)
2564 { m_impl += ImplStr(psz); return *this; }
2565 wxString& operator+=(const wchar_t *pwz)
2566 { m_impl += ImplStr(pwz); return *this; }
2567 wxString& operator+=(const wxCStrData& s)
2568 { m_impl += s.AsString().m_impl; return *this; }
2569 wxString& operator+=(const wxCharBuffer& s)
2570 { return operator+=(s.data()); }
2571 wxString& operator+=(const wxWCharBuffer& s)
2572 { return operator+=(s.data()); }
2573 // string += char
2574 wxString& operator+=(wxUniChar ch)
2575 {
2576 #if wxUSE_UNICODE_UTF8
2577 if ( !ch.IsAscii() )
2578 m_impl += wxStringOperations::EncodeChar(ch);
2579 else
2580 #endif
2581 m_impl += (wxStringCharType)ch;
2582 return *this;
2583 }
2584 wxString& operator+=(wxUniCharRef ch) { return *this += wxUniChar(ch); }
2585 wxString& operator+=(int ch) { return *this += wxUniChar(ch); }
2586 wxString& operator+=(char ch) { return *this += wxUniChar(ch); }
2587 wxString& operator+=(unsigned char ch) { return *this += wxUniChar(ch); }
2588 wxString& operator+=(wchar_t ch) { return *this += wxUniChar(ch); }
2589
2590 private:
2591 #if !wxUSE_STL_BASED_WXSTRING
2592 // helpers for wxStringBuffer and wxStringBufferLength
2593 wxStringCharType *DoGetWriteBuf(size_t nLen)
2594 { return m_impl.DoGetWriteBuf(nLen); }
2595 void DoUngetWriteBuf()
2596 { m_impl.DoUngetWriteBuf(); }
2597 void DoUngetWriteBuf(size_t nLen)
2598 { m_impl.DoUngetWriteBuf(nLen); }
2599 #endif // !wxUSE_STL_BASED_WXSTRING
2600
2601 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
2602 #if !wxUSE_UTF8_LOCALE_ONLY
2603 int DoPrintfWchar(const wxChar *format, ...);
2604 static wxString DoFormatWchar(const wxChar *format, ...);
2605 #endif
2606 #if wxUSE_UNICODE_UTF8
2607 int DoPrintfUtf8(const char *format, ...);
2608 static wxString DoFormatUtf8(const char *format, ...);
2609 #endif
2610 #endif
2611
2612 #if !wxUSE_STL_BASED_WXSTRING
2613 // check string's data validity
2614 bool IsValid() const { return m_impl.GetStringData()->IsValid(); }
2615 #endif
2616
2617 private:
2618 wxStringImpl m_impl;
2619
2620 // buffers for compatibility conversion from (char*)c_str() and
2621 // (wchar_t*)c_str():
2622 // FIXME-UTF8: bechmark various approaches to keeping compatibility buffers
2623 template<typename T>
2624 struct ConvertedBuffer
2625 {
2626 ConvertedBuffer() : m_buf(NULL) {}
2627 ~ConvertedBuffer()
2628 { free(m_buf); }
2629
2630 operator T*() const { return m_buf; }
2631
2632 ConvertedBuffer& operator=(T *str)
2633 {
2634 free(m_buf);
2635 m_buf = str;
2636 return *this;
2637 }
2638
2639 T *m_buf;
2640 };
2641 #if wxUSE_UNICODE && !wxUSE_UTF8_LOCALE_ONLY
2642 ConvertedBuffer<char> m_convertedToChar;
2643 #endif
2644 #if !wxUSE_UNICODE_WCHAR
2645 ConvertedBuffer<wchar_t> m_convertedToWChar;
2646 #endif
2647
2648 #if wxUSE_UNICODE_UTF8
2649 // FIXME-UTF8: (try to) move this elsewhere (TLS) or solve differently
2650 // assigning to character pointer to by wxString::interator may
2651 // change the underlying wxStringImpl iterator, so we have to
2652 // keep track of all iterators and update them as necessary:
2653 struct wxStringIteratorNodeHead
2654 {
2655 wxStringIteratorNodeHead() : ptr(NULL) {}
2656 wxStringIteratorNode *ptr;
2657
2658 // copying is disallowed as it would result in more than one pointer into
2659 // the same linked list
2660 DECLARE_NO_COPY_CLASS(wxStringIteratorNodeHead)
2661 };
2662
2663 wxStringIteratorNodeHead m_iterators;
2664
2665 friend class WXDLLIMPEXP_FWD_BASE wxStringIteratorNode;
2666 friend class WXDLLIMPEXP_FWD_BASE wxUniCharRef;
2667 #endif // wxUSE_UNICODE_UTF8
2668
2669 friend class WXDLLIMPEXP_FWD_BASE wxCStrData;
2670 friend class wxStringInternalBuffer;
2671 friend class wxStringInternalBufferLength;
2672 };
2673
2674 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
2675 #pragma warning (default:4275)
2676 #endif
2677
2678 // string iterator operators that satisfy STL Random Access Iterator
2679 // requirements:
2680 inline wxString::iterator operator+(ptrdiff_t n, wxString::iterator i)
2681 { return i + n; }
2682 inline wxString::const_iterator operator+(ptrdiff_t n, wxString::const_iterator i)
2683 { return i + n; }
2684 inline wxString::reverse_iterator operator+(ptrdiff_t n, wxString::reverse_iterator i)
2685 { return i + n; }
2686 inline wxString::const_reverse_iterator operator+(ptrdiff_t n, wxString::const_reverse_iterator i)
2687 { return i + n; }
2688
2689 // notice that even though for many compilers the friend declarations above are
2690 // enough, from the point of view of C++ standard we must have the declarations
2691 // here as friend ones are not injected in the enclosing namespace and without
2692 // them the code fails to compile with conforming compilers such as xlC or g++4
2693 wxString WXDLLIMPEXP_BASE operator+(const wxString& string1, const wxString& string2);
2694 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const char *psz);
2695 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const wchar_t *pwz);
2696 wxString WXDLLIMPEXP_BASE operator+(const char *psz, const wxString& string);
2697 wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz, const wxString& string);
2698
2699 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxUniChar ch);
2700 wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string);
2701
2702 inline wxString operator+(const wxString& string, wxUniCharRef ch)
2703 { return string + (wxUniChar)ch; }
2704 inline wxString operator+(const wxString& string, char ch)
2705 { return string + wxUniChar(ch); }
2706 inline wxString operator+(const wxString& string, wchar_t ch)
2707 { return string + wxUniChar(ch); }
2708 inline wxString operator+(wxUniCharRef ch, const wxString& string)
2709 { return (wxUniChar)ch + string; }
2710 inline wxString operator+(char ch, const wxString& string)
2711 { return wxUniChar(ch) + string; }
2712 inline wxString operator+(wchar_t ch, const wxString& string)
2713 { return wxUniChar(ch) + string; }
2714
2715
2716 #define wxGetEmptyString() wxString()
2717
2718 // ----------------------------------------------------------------------------
2719 // wxStringBuffer: a tiny class allowing to get a writable pointer into string
2720 // ----------------------------------------------------------------------------
2721
2722 #if !wxUSE_STL_BASED_WXSTRING
2723 // string buffer for direct access to string data in their native
2724 // representation:
2725 class wxStringInternalBuffer
2726 {
2727 public:
2728 typedef wxStringCharType CharType;
2729
2730 wxStringInternalBuffer(wxString& str, size_t lenWanted = 1024)
2731 : m_str(str), m_buf(NULL)
2732 { m_buf = m_str.DoGetWriteBuf(lenWanted); }
2733
2734 ~wxStringInternalBuffer() { m_str.DoUngetWriteBuf(); }
2735
2736 operator wxStringCharType*() const { return m_buf; }
2737
2738 private:
2739 wxString& m_str;
2740 wxStringCharType *m_buf;
2741
2742 DECLARE_NO_COPY_CLASS(wxStringInternalBuffer)
2743 };
2744
2745 class wxStringInternalBufferLength
2746 {
2747 public:
2748 typedef wxStringCharType CharType;
2749
2750 wxStringInternalBufferLength(wxString& str, size_t lenWanted = 1024)
2751 : m_str(str), m_buf(NULL), m_len(0), m_lenSet(false)
2752 {
2753 m_buf = m_str.DoGetWriteBuf(lenWanted);
2754 wxASSERT(m_buf != NULL);
2755 }
2756
2757 ~wxStringInternalBufferLength()
2758 {
2759 wxASSERT(m_lenSet);
2760 m_str.DoUngetWriteBuf(m_len);
2761 }
2762
2763 operator wxStringCharType*() const { return m_buf; }
2764 void SetLength(size_t length) { m_len = length; m_lenSet = true; }
2765
2766 private:
2767 wxString& m_str;
2768 wxStringCharType *m_buf;
2769 size_t m_len;
2770 bool m_lenSet;
2771
2772 DECLARE_NO_COPY_CLASS(wxStringInternalBufferLength)
2773 };
2774
2775 #endif // !wxUSE_STL_BASED_WXSTRING
2776
2777 template<typename T>
2778 class WXDLLIMPEXP_BASE wxStringTypeBufferBase
2779 {
2780 public:
2781 typedef T CharType;
2782
2783 wxStringTypeBufferBase(wxString& str, size_t lenWanted = 1024)
2784 : m_str(str), m_buf(lenWanted)
2785 { }
2786
2787
2788 operator CharType*() { return m_buf.data(); }
2789
2790 protected:
2791 wxString& m_str;
2792 wxCharTypeBuffer<CharType> m_buf;
2793 };
2794
2795 template<typename T>
2796 class WXDLLIMPEXP_BASE wxStringTypeBufferLengthBase
2797 {
2798 public:
2799 typedef T CharType;
2800
2801 wxStringTypeBufferLengthBase(wxString& str, size_t lenWanted = 1024)
2802 : m_str(str), m_buf(lenWanted), m_len(0), m_lenSet(false)
2803 { }
2804
2805 operator CharType*() { return m_buf.data(); }
2806 void SetLength(size_t length) { m_len = length; m_lenSet = true; }
2807
2808 protected:
2809 wxString& m_str;
2810 wxCharTypeBuffer<CharType> m_buf;
2811 size_t m_len;
2812 bool m_lenSet;
2813 };
2814
2815 template<typename T>
2816 class wxStringTypeBuffer : public wxStringTypeBufferBase<T>
2817 {
2818 public:
2819 wxStringTypeBuffer(wxString& str, size_t lenWanted = 1024)
2820 : wxStringTypeBufferBase<T>(str, lenWanted) {}
2821 ~wxStringTypeBuffer()
2822 {
2823 this->m_str.assign(this->m_buf.data());
2824 }
2825
2826 DECLARE_NO_COPY_CLASS(wxStringTypeBuffer)
2827 };
2828
2829 template<typename T>
2830 class wxStringTypeBufferLength : public wxStringTypeBufferLengthBase<T>
2831 {
2832 public:
2833 wxStringTypeBufferLength(wxString& str, size_t lenWanted = 1024)
2834 : wxStringTypeBufferLengthBase<T>(str, lenWanted) {}
2835
2836 ~wxStringTypeBufferLength()
2837 {
2838 wxASSERT(this->m_lenSet);
2839 this->m_str.assign(this->m_buf.data(), this->m_len);
2840 }
2841
2842 DECLARE_NO_COPY_CLASS(wxStringTypeBufferLength)
2843 };
2844
2845 #if wxUSE_STL_BASED_WXSTRING
2846
2847 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferBase<wxStringCharType> )
2848
2849 class wxStringInternalBuffer : public wxStringTypeBufferBase<wxStringCharType>
2850 {
2851 public:
2852 wxStringInternalBuffer(wxString& str, size_t lenWanted = 1024)
2853 : wxStringTypeBufferBase<wxStringCharType>(str, lenWanted) {}
2854 ~wxStringInternalBuffer()
2855 { m_str.m_impl.assign(m_buf.data()); }
2856
2857 DECLARE_NO_COPY_CLASS(wxStringInternalBuffer)
2858 };
2859
2860 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE(
2861 wxStringTypeBufferLengthBase<wxStringCharType> )
2862
2863 class wxStringInternalBufferLength
2864 : public wxStringTypeBufferLengthBase<wxStringCharType>
2865 {
2866 public:
2867 wxStringInternalBufferLength(wxString& str, size_t lenWanted = 1024)
2868 : wxStringTypeBufferLengthBase<wxStringCharType>(str, lenWanted) {}
2869
2870 ~wxStringInternalBufferLength()
2871 {
2872 wxASSERT(m_lenSet);
2873 m_str.m_impl.assign(m_buf.data(), m_len);
2874 }
2875
2876 DECLARE_NO_COPY_CLASS(wxStringInternalBufferLength)
2877 };
2878 #endif // wxUSE_STL_BASED_WXSTRING
2879
2880
2881 #if wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8
2882 typedef wxStringTypeBuffer<wxChar> wxStringBuffer;
2883 typedef wxStringTypeBufferLength<wxChar> wxStringBufferLength;
2884 #else // if !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2885 typedef wxStringInternalBuffer wxStringBuffer;
2886 typedef wxStringInternalBufferLength wxStringBufferLength;
2887 #endif // !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
2888
2889 #if wxUSE_UNICODE_UTF8
2890 typedef wxStringInternalBuffer wxUTF8StringBuffer;
2891 typedef wxStringInternalBufferLength wxUTF8StringBufferLength;
2892 #elif wxUSE_UNICODE_WCHAR
2893
2894 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferBase<char> )
2895
2896 class WXDLLIMPEXP_BASE wxUTF8StringBuffer : public wxStringTypeBufferBase<char>
2897 {
2898 public:
2899 wxUTF8StringBuffer(wxString& str, size_t lenWanted = 1024)
2900 : wxStringTypeBufferBase<char>(str, lenWanted) {}
2901 ~wxUTF8StringBuffer();
2902
2903 DECLARE_NO_COPY_CLASS(wxUTF8StringBuffer)
2904 };
2905
2906 WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferLengthBase<char> )
2907
2908 class WXDLLIMPEXP_BASE wxUTF8StringBufferLength
2909 : public wxStringTypeBufferLengthBase<char>
2910 {
2911 public:
2912 wxUTF8StringBufferLength(wxString& str, size_t lenWanted = 1024)
2913 : wxStringTypeBufferLengthBase<char>(str, lenWanted) {}
2914 ~wxUTF8StringBufferLength();
2915
2916 DECLARE_NO_COPY_CLASS(wxUTF8StringBufferLength)
2917 };
2918 #endif // wxUSE_UNICODE_UTF8/wxUSE_UNICODE_WCHAR
2919
2920
2921 // ---------------------------------------------------------------------------
2922 // wxString comparison functions: operator versions are always case sensitive
2923 // ---------------------------------------------------------------------------
2924
2925 #define wxCMP_WXCHAR_STRING(p, s, op) 0 op s.Cmp(p)
2926
2927 wxDEFINE_ALL_COMPARISONS(const wxChar *, const wxString&, wxCMP_WXCHAR_STRING)
2928
2929 #undef wxCMP_WXCHAR_STRING
2930
2931 inline bool operator==(const wxString& s1, const wxString& s2)
2932 { return s1.IsSameAs(s2); }
2933 inline bool operator!=(const wxString& s1, const wxString& s2)
2934 { return !s1.IsSameAs(s2); }
2935 inline bool operator< (const wxString& s1, const wxString& s2)
2936 { return s1.Cmp(s2) < 0; }
2937 inline bool operator> (const wxString& s1, const wxString& s2)
2938 { return s1.Cmp(s2) > 0; }
2939 inline bool operator<=(const wxString& s1, const wxString& s2)
2940 { return s1.Cmp(s2) <= 0; }
2941 inline bool operator>=(const wxString& s1, const wxString& s2)
2942 { return s1.Cmp(s2) >= 0; }
2943
2944 inline bool operator==(const wxString& s1, const wxCStrData& s2)
2945 { return s1 == s2.AsString(); }
2946 inline bool operator==(const wxCStrData& s1, const wxString& s2)
2947 { return s1.AsString() == s2; }
2948 inline bool operator!=(const wxString& s1, const wxCStrData& s2)
2949 { return s1 != s2.AsString(); }
2950 inline bool operator!=(const wxCStrData& s1, const wxString& s2)
2951 { return s1.AsString() != s2; }
2952
2953 inline bool operator==(const wxString& s1, const wxWCharBuffer& s2)
2954 { return (s1.Cmp((const wchar_t *)s2) == 0); }
2955 inline bool operator==(const wxWCharBuffer& s1, const wxString& s2)
2956 { return (s2.Cmp((const wchar_t *)s1) == 0); }
2957 inline bool operator!=(const wxString& s1, const wxWCharBuffer& s2)
2958 { return (s1.Cmp((const wchar_t *)s2) != 0); }
2959 inline bool operator!=(const wxWCharBuffer& s1, const wxString& s2)
2960 { return (s2.Cmp((const wchar_t *)s1) != 0); }
2961
2962 inline bool operator==(const wxString& s1, const wxCharBuffer& s2)
2963 { return (s1.Cmp((const char *)s2) == 0); }
2964 inline bool operator==(const wxCharBuffer& s1, const wxString& s2)
2965 { return (s2.Cmp((const char *)s1) == 0); }
2966 inline bool operator!=(const wxString& s1, const wxCharBuffer& s2)
2967 { return (s1.Cmp((const char *)s2) != 0); }
2968 inline bool operator!=(const wxCharBuffer& s1, const wxString& s2)
2969 { return (s2.Cmp((const char *)s1) != 0); }
2970
2971 inline wxString operator+(const wxString& string, const wxWCharBuffer& buf)
2972 { return string + (const wchar_t *)buf; }
2973 inline wxString operator+(const wxWCharBuffer& buf, const wxString& string)
2974 { return (const wchar_t *)buf + string; }
2975
2976 inline wxString operator+(const wxString& string, const wxCharBuffer& buf)
2977 { return string + (const char *)buf; }
2978 inline wxString operator+(const wxCharBuffer& buf, const wxString& string)
2979 { return (const char *)buf + string; }
2980
2981 // comparison with char
2982 inline bool operator==(const wxUniChar& c, const wxString& s) { return s.IsSameAs(c); }
2983 inline bool operator==(const wxUniCharRef& c, const wxString& s) { return s.IsSameAs(c); }
2984 inline bool operator==(char c, const wxString& s) { return s.IsSameAs(c); }
2985 inline bool operator==(wchar_t c, const wxString& s) { return s.IsSameAs(c); }
2986 inline bool operator==(int c, const wxString& s) { return s.IsSameAs(c); }
2987 inline bool operator==(const wxString& s, const wxUniChar& c) { return s.IsSameAs(c); }
2988 inline bool operator==(const wxString& s, const wxUniCharRef& c) { return s.IsSameAs(c); }
2989 inline bool operator==(const wxString& s, char c) { return s.IsSameAs(c); }
2990 inline bool operator==(const wxString& s, wchar_t c) { return s.IsSameAs(c); }
2991 inline bool operator!=(const wxUniChar& c, const wxString& s) { return !s.IsSameAs(c); }
2992 inline bool operator!=(const wxUniCharRef& c, const wxString& s) { return !s.IsSameAs(c); }
2993 inline bool operator!=(char c, const wxString& s) { return !s.IsSameAs(c); }
2994 inline bool operator!=(wchar_t c, const wxString& s) { return !s.IsSameAs(c); }
2995 inline bool operator!=(int c, const wxString& s) { return !s.IsSameAs(c); }
2996 inline bool operator!=(const wxString& s, const wxUniChar& c) { return !s.IsSameAs(c); }
2997 inline bool operator!=(const wxString& s, const wxUniCharRef& c) { return !s.IsSameAs(c); }
2998 inline bool operator!=(const wxString& s, char c) { return !s.IsSameAs(c); }
2999 inline bool operator!=(const wxString& s, wchar_t c) { return !s.IsSameAs(c); }
3000
3001 // comparison with C string in Unicode build
3002 #if wxUSE_UNICODE
3003
3004 #define wxCMP_CHAR_STRING(p, s, op) wxString(p) op s
3005
3006 wxDEFINE_ALL_COMPARISONS(const char *, const wxString&, wxCMP_CHAR_STRING)
3007
3008 #undef wxCMP_CHAR_STRING
3009
3010 #endif // wxUSE_UNICODE
3011
3012 // we also need to provide the operators for comparison with wxCStrData to
3013 // resolve ambiguity between operator(const wxChar *,const wxString &) and
3014 // operator(const wxChar *, const wxChar *) for "p == s.c_str()"
3015 //
3016 // notice that these are (shallow) pointer comparisons, not (deep) string ones
3017 #define wxCMP_CHAR_CSTRDATA(p, s, op) p op s.AsChar()
3018 #define wxCMP_WCHAR_CSTRDATA(p, s, op) p op s.AsWChar()
3019
3020 wxDEFINE_ALL_COMPARISONS(const wchar_t *, const wxCStrData&, wxCMP_WCHAR_CSTRDATA)
3021 wxDEFINE_ALL_COMPARISONS(const char *, const wxCStrData&, wxCMP_CHAR_CSTRDATA)
3022
3023 #undef wxCMP_CHAR_CSTRDATA
3024 #undef wxCMP_WCHAR_CSTRDATA
3025
3026 // ---------------------------------------------------------------------------
3027 // Implementation only from here until the end of file
3028 // ---------------------------------------------------------------------------
3029
3030 #if wxUSE_STD_IOSTREAM
3031
3032 #include "wx/iosfwrap.h"
3033
3034 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxString&);
3035 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCStrData&);
3036 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCharBuffer&);
3037 #ifndef __BORLANDC__
3038 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxWCharBuffer&);
3039 #endif
3040
3041 #if wxUSE_UNICODE && defined(HAVE_WOSTREAM)
3042
3043 WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxString&);
3044 WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxCStrData&);
3045 WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxWCharBuffer&);
3046
3047 #endif // wxUSE_UNICODE && defined(HAVE_WOSTREAM)
3048
3049 #endif // wxUSE_STD_IOSTREAM
3050
3051 // ---------------------------------------------------------------------------
3052 // wxCStrData implementation
3053 // ---------------------------------------------------------------------------
3054
3055 inline wxCStrData::wxCStrData(char *buf)
3056 : m_str(new wxString(buf)), m_offset(0), m_owned(true) {}
3057 inline wxCStrData::wxCStrData(wchar_t *buf)
3058 : m_str(new wxString(buf)), m_offset(0), m_owned(true) {}
3059
3060 inline wxCStrData::wxCStrData(const wxCStrData& data)
3061 : m_str(data.m_owned ? new wxString(*data.m_str) : data.m_str),
3062 m_offset(data.m_offset),
3063 m_owned(data.m_owned)
3064 {
3065 }
3066
3067 inline wxCStrData::~wxCStrData()
3068 {
3069 if ( m_owned )
3070 delete wx_const_cast(wxString*, m_str); // cast to silence warnings
3071 }
3072
3073 // simple cases for AsChar() and AsWChar(), the complicated ones are
3074 // in string.cpp
3075 #if wxUSE_UNICODE_WCHAR
3076 inline const wchar_t* wxCStrData::AsWChar() const
3077 {
3078 return m_str->wx_str() + m_offset;
3079 }
3080 #endif // wxUSE_UNICODE_WCHAR
3081
3082 #if !wxUSE_UNICODE
3083 inline const char* wxCStrData::AsChar() const
3084 {
3085 return m_str->wx_str() + m_offset;
3086 }
3087 #endif // !wxUSE_UNICODE
3088
3089 #if wxUSE_UTF8_LOCALE_ONLY
3090 inline const char* wxCStrData::AsChar() const
3091 {
3092 return wxStringOperations::AddToIter(m_str->wx_str(), m_offset);
3093 }
3094 #endif // wxUSE_UTF8_LOCALE_ONLY
3095
3096 inline const wxCharBuffer wxCStrData::AsCharBuf() const
3097 {
3098 #if !wxUSE_UNICODE
3099 return wxCharBuffer::CreateNonOwned(AsChar());
3100 #else
3101 return AsString().mb_str();
3102 #endif
3103 }
3104
3105 inline const wxWCharBuffer wxCStrData::AsWCharBuf() const
3106 {
3107 #if wxUSE_UNICODE_WCHAR
3108 return wxWCharBuffer::CreateNonOwned(AsWChar());
3109 #else
3110 return AsString().wc_str();
3111 #endif
3112 }
3113
3114 inline wxString wxCStrData::AsString() const
3115 {
3116 if ( m_offset == 0 )
3117 return *m_str;
3118 else
3119 return m_str->Mid(m_offset);
3120 }
3121
3122 inline const wxStringCharType *wxCStrData::AsInternal() const
3123 {
3124 #if wxUSE_UNICODE_UTF8
3125 return wxStringOperations::AddToIter(m_str->wx_str(), m_offset);
3126 #else
3127 return m_str->wx_str() + m_offset;
3128 #endif
3129 }
3130
3131 inline wxUniChar wxCStrData::operator*() const
3132 {
3133 if ( m_str->empty() )
3134 return wxUniChar(_T('\0'));
3135 else
3136 return (*m_str)[m_offset];
3137 }
3138
3139 inline wxUniChar wxCStrData::operator[](size_t n) const
3140 {
3141 // NB: we intentionally use operator[] and not at() here because the former
3142 // works for the terminating NUL while the latter does not
3143 return (*m_str)[m_offset + n];
3144 }
3145
3146 // ----------------------------------------------------------------------------
3147 // more wxCStrData operators
3148 // ----------------------------------------------------------------------------
3149
3150 // we need to define those to allow "size_t pos = p - s.c_str()" where p is
3151 // some pointer into the string
3152 inline size_t operator-(const char *p, const wxCStrData& cs)
3153 {
3154 return p - cs.AsChar();
3155 }
3156
3157 inline size_t operator-(const wchar_t *p, const wxCStrData& cs)
3158 {
3159 return p - cs.AsWChar();
3160 }
3161
3162 // ----------------------------------------------------------------------------
3163 // implementation of wx[W]CharBuffer inline methods using wxCStrData
3164 // ----------------------------------------------------------------------------
3165
3166 // FIXME-UTF8: move this to buffer.h
3167 inline wxCharBuffer::wxCharBuffer(const wxCStrData& cstr)
3168 : wxCharTypeBufferBase(cstr.AsCharBuf())
3169 {
3170 }
3171
3172 inline wxWCharBuffer::wxWCharBuffer(const wxCStrData& cstr)
3173 : wxCharTypeBufferBase(cstr.AsWCharBuf())
3174 {
3175 }
3176
3177 #if wxUSE_UNICODE_UTF8
3178 // ----------------------------------------------------------------------------
3179 // implementation of wxStringIteratorNode inline methods
3180 // ----------------------------------------------------------------------------
3181
3182 void wxStringIteratorNode::DoSet(const wxString *str,
3183 wxStringImpl::const_iterator *citer,
3184 wxStringImpl::iterator *iter)
3185 {
3186 m_prev = NULL;
3187 m_iter = iter;
3188 m_citer = citer;
3189 m_str = str;
3190 if ( str )
3191 {
3192 m_next = str->m_iterators.ptr;
3193 wx_const_cast(wxString*, m_str)->m_iterators.ptr = this;
3194 if ( m_next )
3195 m_next->m_prev = this;
3196 }
3197 else
3198 {
3199 m_next = NULL;
3200 }
3201 }
3202
3203 void wxStringIteratorNode::clear()
3204 {
3205 if ( m_next )
3206 m_next->m_prev = m_prev;
3207 if ( m_prev )
3208 m_prev->m_next = m_next;
3209 else if ( m_str ) // first in the list
3210 wx_const_cast(wxString*, m_str)->m_iterators.ptr = m_next;
3211
3212 m_next = m_prev = NULL;
3213 m_citer = NULL;
3214 m_iter = NULL;
3215 m_str = NULL;
3216 }
3217 #endif // wxUSE_UNICODE_UTF8
3218
3219 #if WXWIN_COMPATIBILITY_2_8
3220 // lot of code out there doesn't explicitly include wx/crt.h, but uses
3221 // CRT wrappers that are now declared in wx/wxcrt.h and wx/wxcrtvararg.h,
3222 // so let's include this header now that wxString is defined and it's safe
3223 // to do it:
3224 #include "wx/crt.h"
3225 #endif
3226
3227 #endif // _WX_WXSTRING_H_