]> git.saurik.com Git - wxWidgets.git/blob - include/wx/string.h
made wxTextInputStream Unicode safe and general cleanup (patch 653775)
[wxWidgets.git] / include / wx / string.h
1 ///////////////////////////////////////////////////////////////////////////////
2 // Name: string.h
3 // Purpose: wxString and wxArrayString classes
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 wxWindows version 1 wxString and std::string and some handy functions
15 missing from string.h.
16 */
17
18 #ifndef _WX_WXSTRINGH__
19 #define _WX_WXSTRINGH__
20
21 #if defined(__GNUG__) && !defined(__APPLE__)
22 #pragma interface "string.h"
23 #endif
24
25 // ----------------------------------------------------------------------------
26 // conditinal compilation
27 // ----------------------------------------------------------------------------
28
29 // compile the std::string compatibility functions if defined
30 #define wxSTD_STRING_COMPATIBILITY
31
32 // ----------------------------------------------------------------------------
33 // headers
34 // ----------------------------------------------------------------------------
35
36 #include "wx/defs.h" // everybody should include this
37
38 #if defined(__WXMAC__) || defined(__VISAGECPP__)
39 #include <ctype.h>
40 #endif
41
42 #ifdef __EMX__
43 #include <std.h>
44 #endif
45
46 #if defined(__VISAGECPP__) && __IBMCPP__ >= 400
47 // problem in VACPP V4 with including stdlib.h multiple times
48 // strconv includes it anyway
49 # include <stdio.h>
50 # include <string.h>
51 # include <stdarg.h>
52 # include <limits.h>
53 #else
54 # include <string.h>
55 # include <stdio.h>
56 # include <stdarg.h>
57 # include <limits.h>
58 # include <stdlib.h>
59 #endif
60
61 #ifdef HAVE_STRINGS_H
62 #include <strings.h> // for strcasecmp()
63 #endif // HAVE_STRINGS_H
64
65 #include "wx/wxchar.h" // for wxChar
66 #include "wx/buffer.h" // for wxCharBuffer
67 #include "wx/strconv.h" // for wxConvertXXX() macros and wxMBConv classes
68
69 // ---------------------------------------------------------------------------
70 // macros
71 // ---------------------------------------------------------------------------
72
73 // casts [unfortunately!] needed to call some broken functions which require
74 // "char *" instead of "const char *"
75 #define WXSTRINGCAST (wxChar *)(const wxChar *)
76 #define wxCSTRINGCAST (wxChar *)(const wxChar *)
77 #define wxMBSTRINGCAST (char *)(const char *)
78 #define wxWCSTRINGCAST (wchar_t *)(const wchar_t *)
79
80 // implementation only
81 #define wxASSERT_VALID_INDEX(i) \
82 wxASSERT_MSG( (size_t)(i) <= Len(), _T("invalid index in wxString") )
83
84 // ----------------------------------------------------------------------------
85 // constants
86 // ----------------------------------------------------------------------------
87
88 #if defined(__VISAGECPP__) && __IBMCPP__ >= 400
89 // must define this static for VA or else you get multiply defined symbols everywhere
90 extern const unsigned int wxSTRING_MAXLEN;
91
92 #else
93 // maximum possible length for a string means "take all string" everywhere
94 // (as sizeof(StringData) is unknown here, we substract 100)
95 const unsigned int wxSTRING_MAXLEN = UINT_MAX - 100;
96
97 #endif
98
99 // ----------------------------------------------------------------------------
100 // global data
101 // ----------------------------------------------------------------------------
102
103 // global pointer to empty string
104 extern WXDLLIMPEXP_DATA_BASE(const wxChar*) wxEmptyString;
105
106 // ---------------------------------------------------------------------------
107 // global functions complementing standard C string library replacements for
108 // strlen() and portable strcasecmp()
109 //---------------------------------------------------------------------------
110
111 // Use wxXXX() functions from wxchar.h instead! These functions are for
112 // backwards compatibility only.
113
114 // checks whether the passed in pointer is NULL and if the string is empty
115 inline bool IsEmpty(const char *p) { return (!p || !*p); }
116
117 // safe version of strlen() (returns 0 if passed NULL pointer)
118 inline size_t Strlen(const char *psz)
119 { return psz ? strlen(psz) : 0; }
120
121 // portable strcasecmp/_stricmp
122 inline int Stricmp(const char *psz1, const char *psz2)
123 {
124 #if defined(__VISUALC__) || ( defined(__MWERKS__) && defined(__INTEL__) )
125 return _stricmp(psz1, psz2);
126 #elif defined(__SC__)
127 return _stricmp(psz1, psz2);
128 #elif defined(__SALFORDC__)
129 return stricmp(psz1, psz2);
130 #elif defined(__BORLANDC__)
131 return stricmp(psz1, psz2);
132 #elif defined(__WATCOMC__)
133 return stricmp(psz1, psz2);
134 #elif defined(__DJGPP__)
135 return stricmp(psz1, psz2);
136 #elif defined(__EMX__)
137 return stricmp(psz1, psz2);
138 #elif defined(__WXPM__)
139 return stricmp(psz1, psz2);
140 #elif defined(__UNIX__) || defined(__GNUWIN32__)
141 return strcasecmp(psz1, psz2);
142 #elif defined(__MWERKS__) && !defined(__INTEL__)
143 register char c1, c2;
144 do {
145 c1 = tolower(*psz1++);
146 c2 = tolower(*psz2++);
147 } while ( c1 && (c1 == c2) );
148
149 return c1 - c2;
150 #else
151 // almost all compilers/libraries provide this function (unfortunately under
152 // different names), that's why we don't implement our own which will surely
153 // be more efficient than this code (uncomment to use):
154 /*
155 register char c1, c2;
156 do {
157 c1 = tolower(*psz1++);
158 c2 = tolower(*psz2++);
159 } while ( c1 && (c1 == c2) );
160
161 return c1 - c2;
162 */
163
164 #error "Please define string case-insensitive compare for your OS/compiler"
165 #endif // OS/compiler
166 }
167
168 // return an empty wxString
169 class WXDLLIMPEXP_BASE wxString; // not yet defined
170 inline const wxString& wxGetEmptyString() { return *(wxString *)&wxEmptyString; }
171
172 // ---------------------------------------------------------------------------
173 // string data prepended with some housekeeping info (used by wxString class),
174 // is never used directly (but had to be put here to allow inlining)
175 // ---------------------------------------------------------------------------
176
177 struct WXDLLIMPEXP_BASE wxStringData
178 {
179 int nRefs; // reference count
180 size_t nDataLength, // actual string length
181 nAllocLength; // allocated memory size
182
183 // mimics declaration 'wxChar data[nAllocLength]'
184 wxChar* data() const { return (wxChar*)(this + 1); }
185
186 // empty string has a special ref count so it's never deleted
187 bool IsEmpty() const { return (nRefs == -1); }
188 bool IsShared() const { return (nRefs > 1); }
189
190 // lock/unlock
191 void Lock() { if ( !IsEmpty() ) nRefs++; }
192
193 // VC++ will refuse to inline Unlock but profiling shows that it is wrong
194 #if defined(__VISUALC__) && (__VISUALC__ >= 1200)
195 __forceinline
196 #endif
197 // VC++ free must take place in same DLL as allocation when using non dll
198 // run-time library (e.g. Multithreaded instead of Multithreaded DLL)
199 #if defined(__VISUALC__) && defined(_MT) && !defined(_DLL)
200 void Unlock() { if ( !IsEmpty() && --nRefs == 0) Free(); }
201 // we must not inline deallocation since allocation is not inlined
202 void Free();
203 #else
204 void Unlock() { if ( !IsEmpty() && --nRefs == 0) free(this); }
205 #endif
206
207 // if we had taken control over string memory (GetWriteBuf), it's
208 // intentionally put in invalid state
209 void Validate(bool b) { nRefs = (b ? 1 : 0); }
210 bool IsValid() const { return (nRefs != 0); }
211 };
212
213 // ---------------------------------------------------------------------------
214 // This is (yet another one) String class for C++ programmers. It doesn't use
215 // any of "advanced" C++ features (i.e. templates, exceptions, namespaces...)
216 // thus you should be able to compile it with practicaly any C++ compiler.
217 // This class uses copy-on-write technique, i.e. identical strings share the
218 // same memory as long as neither of them is changed.
219 //
220 // This class aims to be as compatible as possible with the new standard
221 // std::string class, but adds some additional functions and should be at
222 // least as efficient than the standard implementation.
223 //
224 // Performance note: it's more efficient to write functions which take "const
225 // String&" arguments than "const char *" if you assign the argument to
226 // another string.
227 //
228 // It was compiled and tested under Win32, Linux (libc 5 & 6), Solaris 5.5.
229 //
230 // To do:
231 // - ressource support (string tables in ressources)
232 // - more wide character (UNICODE) support
233 // - regular expressions support
234 // ---------------------------------------------------------------------------
235
236 class WXDLLIMPEXP_BASE wxString
237 {
238 #if !wxUSE_STL
239 friend class WXDLLIMPEXP_BASE wxArrayString;
240 #endif
241
242 // NB: special care was taken in arranging the member functions in such order
243 // that all inline functions can be effectively inlined, verify that all
244 // performace critical functions are still inlined if you change order!
245 private:
246 // points to data preceded by wxStringData structure with ref count info
247 wxChar *m_pchData;
248
249 // accessor to string data
250 wxStringData* GetStringData() const { return (wxStringData*)m_pchData - 1; }
251
252 // string (re)initialization functions
253 // initializes the string to the empty value (must be called only from
254 // ctors, use Reinit() otherwise)
255 void Init() { m_pchData = (wxChar *)wxEmptyString; }
256 // initializaes the string with (a part of) C-string
257 void InitWith(const wxChar *psz, size_t nPos = 0, size_t nLen = wxSTRING_MAXLEN);
258 // as Init, but also frees old data
259 void Reinit() { GetStringData()->Unlock(); Init(); }
260
261 // memory allocation
262 // allocates memory for string of length nLen
263 bool AllocBuffer(size_t nLen);
264 // copies data to another string
265 bool AllocCopy(wxString&, int, int) const;
266 // effectively copies data to string
267 bool AssignCopy(size_t, const wxChar *);
268
269 // append a (sub)string
270 bool ConcatSelf(size_t nLen, const wxChar *src);
271
272 // functions called before writing to the string: they copy it if there
273 // are other references to our data (should be the only owner when writing)
274 bool CopyBeforeWrite();
275 bool AllocBeforeWrite(size_t);
276
277 // if we hadn't made these operators private, it would be possible to
278 // compile "wxString s; s = 17;" without any warnings as 17 is implicitly
279 // converted to char in C and we do have operator=(char)
280 //
281 // NB: we don't need other versions (short/long and unsigned) as attempt
282 // to assign another numeric type to wxString will now result in
283 // ambiguity between operator=(char) and operator=(int)
284 wxString& operator=(int);
285
286 // these methods are not implemented - there is _no_ conversion from int to
287 // string, you're doing something wrong if the compiler wants to call it!
288 //
289 // try `s << i' or `s.Printf("%d", i)' instead
290 wxString(int);
291
292 public:
293 // constructors and destructor
294 // ctor for an empty string
295 wxString() : m_pchData(NULL) { Init(); }
296 // copy ctor
297 wxString(const wxString& stringSrc) : m_pchData(NULL)
298 {
299 wxASSERT_MSG( stringSrc.GetStringData()->IsValid(),
300 _T("did you forget to call UngetWriteBuf()?") );
301
302 if ( stringSrc.IsEmpty() ) {
303 // nothing to do for an empty string
304 Init();
305 }
306 else {
307 m_pchData = stringSrc.m_pchData; // share same data
308 GetStringData()->Lock(); // => one more copy
309 }
310 }
311 // string containing nRepeat copies of ch
312 wxString(wxChar ch, size_t nRepeat = 1);
313 // ctor takes first nLength characters from C string
314 // (default value of wxSTRING_MAXLEN means take all the string)
315 wxString(const wxChar *psz, size_t nLength = wxSTRING_MAXLEN)
316 : m_pchData(NULL)
317 { InitWith(psz, 0, nLength); }
318 wxString(const wxChar *psz, wxMBConv& WXUNUSED(conv), size_t nLength = wxSTRING_MAXLEN)
319 : m_pchData(NULL)
320 { InitWith(psz, 0, nLength); }
321
322 #if wxUSE_UNICODE
323 // from multibyte string
324 // (NB: nLength is right now number of Unicode characters, not
325 // characters in psz! So try not to use it yet!)
326 wxString(const char *psz, wxMBConv& conv, size_t nLength = wxSTRING_MAXLEN);
327 // from wxWCharBuffer (i.e. return from wxGetString)
328 wxString(const wxWCharBuffer& psz)
329 { InitWith(psz, 0, wxSTRING_MAXLEN); }
330 #else // ANSI
331 // from C string (for compilers using unsigned char)
332 wxString(const unsigned char* psz, size_t nLength = wxSTRING_MAXLEN)
333 : m_pchData(NULL)
334 { InitWith((const char*)psz, 0, nLength); }
335
336 #if wxUSE_WCHAR_T
337 // from wide (Unicode) string
338 wxString(const wchar_t *pwz, wxMBConv& conv = wxConvLibc, size_t nLength = wxSTRING_MAXLEN);
339 #endif // !wxUSE_WCHAR_T
340
341 // from wxCharBuffer
342 wxString(const wxCharBuffer& psz)
343 : m_pchData(NULL)
344 { InitWith(psz, 0, wxSTRING_MAXLEN); }
345 #endif // Unicode/ANSI
346
347 // dtor is not virtual, this class must not be inherited from!
348 ~wxString() { GetStringData()->Unlock(); }
349
350 // generic attributes & operations
351 // as standard strlen()
352 size_t Len() const { return GetStringData()->nDataLength; }
353 // string contains any characters?
354 bool IsEmpty() const { return Len() == 0; }
355 // empty string is "FALSE", so !str will return TRUE
356 bool operator!() const { return IsEmpty(); }
357 // truncate the string to given length
358 wxString& Truncate(size_t uiLen);
359 // empty string contents
360 void Empty()
361 {
362 Truncate(0);
363
364 wxASSERT_MSG( IsEmpty(), _T("string not empty after call to Empty()?") );
365 }
366 // empty the string and free memory
367 void Clear()
368 {
369 if ( !GetStringData()->IsEmpty() )
370 Reinit();
371
372 wxASSERT_MSG( !GetStringData()->nDataLength &&
373 !GetStringData()->nAllocLength,
374 _T("string should be empty after Clear()") );
375 }
376
377 // contents test
378 // Is an ascii value
379 bool IsAscii() const;
380 // Is a number
381 bool IsNumber() const;
382 // Is a word
383 bool IsWord() const;
384
385 // data access (all indexes are 0 based)
386 // read access
387 wxChar GetChar(size_t n) const
388 { wxASSERT_VALID_INDEX( n ); return m_pchData[n]; }
389 // read/write access
390 wxChar& GetWritableChar(size_t n)
391 { wxASSERT_VALID_INDEX( n ); CopyBeforeWrite(); return m_pchData[n]; }
392 // write access
393 void SetChar(size_t n, wxChar ch)
394 { wxASSERT_VALID_INDEX( n ); CopyBeforeWrite(); m_pchData[n] = ch; }
395
396 // get last character
397 wxChar Last() const
398 {
399 wxASSERT_MSG( !IsEmpty(), _T("wxString: index out of bounds") );
400
401 return m_pchData[Len() - 1];
402 }
403
404 // get writable last character
405 wxChar& Last()
406 {
407 wxASSERT_MSG( !IsEmpty(), _T("wxString: index out of bounds") );
408 CopyBeforeWrite();
409 return m_pchData[Len()-1];
410 }
411
412 /*
413 So why do we have all these overloaded operator[]s? A bit of history:
414 initially there was only one of them, taking size_t. Then people
415 started complaining because they wanted to use ints as indices (I
416 wonder why) and compilers were giving warnings about it, so we had to
417 add the operator[](int). Then it became apparent that you couldn't
418 write str[0] any longer because there was ambiguity between two
419 overloads and so you now had to write str[0u] (or, of course, use the
420 explicit casts to either int or size_t but nobody did this).
421
422 Finally, someone decided to compile wxWin on an Alpha machine and got
423 a surprize: str[0u] didn't compile there because it is of type
424 unsigned int and size_t is unsigned _long_ on Alpha and so there was
425 ambiguity between converting uint to int or ulong. To fix this one we
426 now add operator[](uint) for the machines where size_t is not already
427 the same as unsigned int - hopefully this fixes the problem (for some
428 time)
429
430 The only real fix is, of course, to remove all versions but the one
431 taking size_t...
432 */
433
434 // operator version of GetChar
435 wxChar operator[](size_t n) const
436 { wxASSERT_VALID_INDEX( n ); return m_pchData[n]; }
437
438 // operator version of GetChar
439 wxChar operator[](int n) const
440 { wxASSERT_VALID_INDEX( n ); return m_pchData[n]; }
441
442 // operator version of GetWriteableChar
443 wxChar& operator[](size_t n)
444 { wxASSERT_VALID_INDEX( n ); CopyBeforeWrite(); return m_pchData[n]; }
445
446 #ifndef wxSIZE_T_IS_UINT
447 // operator version of GetChar
448 wxChar operator[](unsigned int n) const
449 { wxASSERT_VALID_INDEX( n ); return m_pchData[n]; }
450
451 // operator version of GetWriteableChar
452 wxChar& operator[](unsigned int n)
453 { wxASSERT_VALID_INDEX( n ); CopyBeforeWrite(); return m_pchData[n]; }
454 #endif // size_t != unsigned int
455
456 // implicit conversion to C string
457 operator const wxChar*() const { return m_pchData; }
458
459 // explicit conversion to C string (use this with printf()!)
460 const wxChar* c_str() const { return m_pchData; }
461 // identical to c_str(), for wxWin 1.6x compatibility
462 const wxChar* wx_str() const { return m_pchData; }
463 // identical to c_str(), for MFC compatibility
464 const wxChar* GetData() const { return m_pchData; }
465
466 // conversion to/from plain (i.e. 7 bit) ASCII: this is useful for
467 // converting numbers or strings which are certain not to contain special
468 // chars (typically system functions, X atoms, environment variables etc.)
469 //
470 // the behaviour of these functions with the strings containing anything
471 // else than 7 bit ASCII characters is undefined, use at your own risk.
472 #if wxUSE_UNICODE
473 static wxString FromAscii(const char *ascii); // string
474 static wxString FromAscii(const char ascii); // char
475 const wxCharBuffer ToAscii() const;
476 #else // ANSI
477 static wxString FromAscii(const char *ascii) { return wxString( ascii ); }
478 static wxString FromAscii(const char ascii) { return wxString( ascii ); }
479 const char *ToAscii() const { return c_str(); }
480 #endif // Unicode/!Unicode
481
482 // conversions with (possible) format conversions: have to return a
483 // buffer with temporary data
484 //
485 // the functions defined (in either Unicode or ANSI) mode are mb_str() to
486 // return an ANSI (multibyte) string, wc_str() to return a wide string and
487 // fn_str() to return a string which should be used with the OS APIs
488 // accepting the file names. The return value is always the same, but the
489 // type differs because a function may either return pointer to the buffer
490 // directly or have to use intermediate buffer for translation.
491 #if wxUSE_UNICODE
492 const wxCharBuffer mb_str(wxMBConv& conv = wxConvLibc) const
493 { return conv.cWC2MB(m_pchData); }
494
495 const wxWX2MBbuf mbc_str() const { return mb_str(*wxConvCurrent); }
496
497 const wxChar* wc_str() const { return m_pchData; }
498
499 // for compatibility with !wxUSE_UNICODE version
500 const wxChar* wc_str(wxMBConv& WXUNUSED(conv)) const { return m_pchData; }
501
502 #if wxMBFILES
503 const wxCharBuffer fn_str() const { return mb_str(wxConvFile); }
504 #else // !wxMBFILES
505 const wxChar* fn_str() const { return m_pchData; }
506 #endif // wxMBFILES/!wxMBFILES
507 #else // ANSI
508 const wxChar* mb_str() const { return m_pchData; }
509
510 // for compatibility with wxUSE_UNICODE version
511 const wxChar* mb_str(wxMBConv& WXUNUSED(conv)) const { return m_pchData; }
512
513 const wxWX2MBbuf mbc_str() const { return mb_str(); }
514
515 #if wxUSE_WCHAR_T
516 const wxWCharBuffer wc_str(wxMBConv& conv) const
517 { return conv.cMB2WC(m_pchData); }
518 #endif // wxUSE_WCHAR_T
519
520 const wxChar* fn_str() const { return m_pchData; }
521 #endif // Unicode/ANSI
522
523 // overloaded assignment
524 // from another wxString
525 wxString& operator=(const wxString& stringSrc);
526 // from a character
527 wxString& operator=(wxChar ch);
528 // from a C string
529 wxString& operator=(const wxChar *psz);
530 #if wxUSE_UNICODE
531 // from wxWCharBuffer
532 wxString& operator=(const wxWCharBuffer& psz)
533 { (void) operator=((const wchar_t *)psz); return *this; }
534 #else // ANSI
535 // from another kind of C string
536 wxString& operator=(const unsigned char* psz);
537 #if wxUSE_WCHAR_T
538 // from a wide string
539 wxString& operator=(const wchar_t *pwz);
540 #endif
541 // from wxCharBuffer
542 wxString& operator=(const wxCharBuffer& psz)
543 { (void) operator=((const char *)psz); return *this; }
544 #endif // Unicode/ANSI
545
546 // string concatenation
547 // in place concatenation
548 /*
549 Concatenate and return the result. Note that the left to right
550 associativity of << allows to write things like "str << str1 << str2
551 << ..." (unlike with +=)
552 */
553 // string += string
554 wxString& operator<<(const wxString& s)
555 {
556 wxASSERT_MSG( s.GetStringData()->IsValid(),
557 _T("did you forget to call UngetWriteBuf()?") );
558
559 ConcatSelf(s.Len(), s);
560 return *this;
561 }
562 // string += C string
563 wxString& operator<<(const wxChar *psz)
564 { ConcatSelf(wxStrlen(psz), psz); return *this; }
565 // string += char
566 wxString& operator<<(wxChar ch) { ConcatSelf(1, &ch); return *this; }
567
568 // string += string
569 void operator+=(const wxString& s) { (void)operator<<(s); }
570 // string += C string
571 void operator+=(const wxChar *psz) { (void)operator<<(psz); }
572 // string += char
573 void operator+=(wxChar ch) { (void)operator<<(ch); }
574
575 // string += buffer (i.e. from wxGetString)
576 #if wxUSE_UNICODE
577 wxString& operator<<(const wxWCharBuffer& s)
578 { (void)operator<<((const wchar_t *)s); return *this; }
579 void operator+=(const wxWCharBuffer& s)
580 { (void)operator<<((const wchar_t *)s); }
581 #else // !wxUSE_UNICODE
582 wxString& operator<<(const wxCharBuffer& s)
583 { (void)operator<<((const char *)s); return *this; }
584 void operator+=(const wxCharBuffer& s)
585 { (void)operator<<((const char *)s); }
586 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
587
588 // string += C string
589 wxString& Append(const wxString& s)
590 {
591 // test for IsEmpty() to share the string if possible
592 if ( IsEmpty() )
593 *this = s;
594 else
595 ConcatSelf(s.Length(), s.c_str());
596 return *this;
597 }
598 wxString& Append(const wxChar* psz)
599 { ConcatSelf(wxStrlen(psz), psz); return *this; }
600 // append count copies of given character
601 wxString& Append(wxChar ch, size_t count = 1u)
602 { wxString str(ch, count); return *this << str; }
603 wxString& Append(const wxChar* psz, size_t nLen)
604 { ConcatSelf(nLen, psz); return *this; }
605
606 // prepend a string, return the string itself
607 wxString& Prepend(const wxString& str)
608 { *this = str + *this; return *this; }
609
610 // non-destructive concatenation
611 //
612 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string1, const wxString& string2);
613 //
614 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxChar ch);
615 //
616 friend wxString WXDLLIMPEXP_BASE operator+(wxChar ch, const wxString& string);
617 //
618 friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const wxChar *psz);
619 //
620 friend wxString WXDLLIMPEXP_BASE operator+(const wxChar *psz, const wxString& string);
621
622 // stream-like functions
623 // insert an int into string
624 wxString& operator<<(int i)
625 { return (*this) << Format(_T("%d"), i); }
626 // insert an unsigned int into string
627 wxString& operator<<(unsigned int ui)
628 { return (*this) << Format(_T("%u"), ui); }
629 // insert a long into string
630 wxString& operator<<(long l)
631 { return (*this) << Format(_T("%ld"), l); }
632 // insert an unsigned long into string
633 wxString& operator<<(unsigned long ul)
634 { return (*this) << Format(_T("%lu"), ul); }
635 // insert a float into string
636 wxString& operator<<(float f)
637 { return (*this) << Format(_T("%f"), f); }
638 // insert a double into string
639 wxString& operator<<(double d)
640 { return (*this) << Format(_T("%g"), d); }
641
642 // string comparison
643 // case-sensitive comparison (returns a value < 0, = 0 or > 0)
644 int Cmp(const wxChar *psz) const { return wxStrcmp(c_str(), psz); }
645 // same as Cmp() but not case-sensitive
646 int CmpNoCase(const wxChar *psz) const { return wxStricmp(c_str(), psz); }
647 // test for the string equality, either considering case or not
648 // (if compareWithCase then the case matters)
649 bool IsSameAs(const wxChar *psz, bool compareWithCase = TRUE) const
650 { return (compareWithCase ? Cmp(psz) : CmpNoCase(psz)) == 0; }
651 // comparison with a signle character: returns TRUE if equal
652 bool IsSameAs(wxChar c, bool compareWithCase = TRUE) const
653 {
654 return (Len() == 1) && (compareWithCase ? GetChar(0u) == c
655 : wxToupper(GetChar(0u)) == wxToupper(c));
656 }
657
658 // simple sub-string extraction
659 // return substring starting at nFirst of length nCount (or till the end
660 // if nCount = default value)
661 wxString Mid(size_t nFirst, size_t nCount = wxSTRING_MAXLEN) const;
662
663 // operator version of Mid()
664 wxString operator()(size_t start, size_t len) const
665 { return Mid(start, len); }
666
667 // check that the string starts with prefix and return the rest of the
668 // string in the provided pointer if it is not NULL, otherwise return
669 // FALSE
670 bool StartsWith(const wxChar *prefix, wxString *rest = NULL) const;
671
672 // get first nCount characters
673 wxString Left(size_t nCount) const;
674 // get last nCount characters
675 wxString Right(size_t nCount) const;
676 // get all characters before the first occurance of ch
677 // (returns the whole string if ch not found)
678 wxString BeforeFirst(wxChar ch) const;
679 // get all characters before the last occurence of ch
680 // (returns empty string if ch not found)
681 wxString BeforeLast(wxChar ch) const;
682 // get all characters after the first occurence of ch
683 // (returns empty string if ch not found)
684 wxString AfterFirst(wxChar ch) const;
685 // get all characters after the last occurence of ch
686 // (returns the whole string if ch not found)
687 wxString AfterLast(wxChar ch) const;
688
689 // for compatibility only, use more explicitly named functions above
690 wxString Before(wxChar ch) const { return BeforeLast(ch); }
691 wxString After(wxChar ch) const { return AfterFirst(ch); }
692
693 // case conversion
694 // convert to upper case in place, return the string itself
695 wxString& MakeUpper();
696 // convert to upper case, return the copy of the string
697 // Here's something to remember: BC++ doesn't like returns in inlines.
698 wxString Upper() const ;
699 // convert to lower case in place, return the string itself
700 wxString& MakeLower();
701 // convert to lower case, return the copy of the string
702 wxString Lower() const ;
703
704 // trimming/padding whitespace (either side) and truncating
705 // remove spaces from left or from right (default) side
706 wxString& Trim(bool bFromRight = TRUE);
707 // add nCount copies chPad in the beginning or at the end (default)
708 wxString& Pad(size_t nCount, wxChar chPad = wxT(' '), bool bFromRight = TRUE);
709
710 // searching and replacing
711 // searching (return starting index, or -1 if not found)
712 int Find(wxChar ch, bool bFromEnd = FALSE) const; // like strchr/strrchr
713 // searching (return starting index, or -1 if not found)
714 int Find(const wxChar *pszSub) const; // like strstr
715 // replace first (or all of bReplaceAll) occurences of substring with
716 // another string, returns the number of replacements made
717 size_t Replace(const wxChar *szOld,
718 const wxChar *szNew,
719 bool bReplaceAll = TRUE);
720
721 // check if the string contents matches a mask containing '*' and '?'
722 bool Matches(const wxChar *szMask) const;
723
724 // conversion to numbers: all functions return TRUE only if the whole
725 // string is a number and put the value of this number into the pointer
726 // provided, the base is the numeric base in which the conversion should be
727 // done and must be comprised between 2 and 36 or be 0 in which case the
728 // standard C rules apply (leading '0' => octal, "0x" => hex)
729 // convert to a signed integer
730 bool ToLong(long *val, int base = 10) const;
731 // convert to an unsigned integer
732 bool ToULong(unsigned long *val, int base = 10) const;
733 // convert to a double
734 bool ToDouble(double *val) const;
735
736 // formated input/output
737 // as sprintf(), returns the number of characters written or < 0 on error
738 // (take 'this' into account in attribute parameter count)
739 int Printf(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_2;
740 // as vprintf(), returns the number of characters written or < 0 on error
741 int PrintfV(const wxChar* pszFormat, va_list argptr);
742
743 // returns the string containing the result of Printf() to it
744 static wxString Format(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_1;
745 // the same as above, but takes a va_list
746 static wxString FormatV(const wxChar *pszFormat, va_list argptr);
747
748 // raw access to string memory
749 // ensure that string has space for at least nLen characters
750 // only works if the data of this string is not shared
751 bool Alloc(size_t nLen);
752 // minimize the string's memory
753 // only works if the data of this string is not shared
754 bool Shrink();
755 // get writable buffer of at least nLen bytes. Unget() *must* be called
756 // a.s.a.p. to put string back in a reasonable state!
757 wxChar *GetWriteBuf(size_t nLen);
758 // call this immediately after GetWriteBuf() has been used
759 void UngetWriteBuf();
760 void UngetWriteBuf(size_t nLen);
761
762 // wxWindows version 1 compatibility functions
763
764 // use Mid()
765 wxString SubString(size_t from, size_t to) const
766 { return Mid(from, (to - from + 1)); }
767 // values for second parameter of CompareTo function
768 enum caseCompare {exact, ignoreCase};
769 // values for first parameter of Strip function
770 enum stripType {leading = 0x1, trailing = 0x2, both = 0x3};
771
772 // use Printf()
773 // (take 'this' into account in attribute parameter count)
774 int sprintf(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_2;
775
776 // use Cmp()
777 inline int CompareTo(const wxChar* psz, caseCompare cmp = exact) const
778 { return cmp == exact ? Cmp(psz) : CmpNoCase(psz); }
779
780 // use Len
781 size_t Length() const { return Len(); }
782 // Count the number of characters
783 int Freq(wxChar ch) const;
784 // use MakeLower
785 void LowerCase() { MakeLower(); }
786 // use MakeUpper
787 void UpperCase() { MakeUpper(); }
788 // use Trim except that it doesn't change this string
789 wxString Strip(stripType w = trailing) const;
790
791 // use Find (more general variants not yet supported)
792 size_t Index(const wxChar* psz) const { return Find(psz); }
793 size_t Index(wxChar ch) const { return Find(ch); }
794 // use Truncate
795 wxString& Remove(size_t pos) { return Truncate(pos); }
796 wxString& RemoveLast(size_t n = 1) { return Truncate(Len() - n); }
797
798 wxString& Remove(size_t nStart, size_t nLen) { return erase( nStart, nLen ); }
799
800 // use Find()
801 int First( const wxChar ch ) const { return Find(ch); }
802 int First( const wxChar* psz ) const { return Find(psz); }
803 int First( const wxString &str ) const { return Find(str); }
804 int Last( const wxChar ch ) const { return Find(ch, TRUE); }
805 bool Contains(const wxString& str) const { return Find(str) != -1; }
806
807 // use IsEmpty()
808 bool IsNull() const { return IsEmpty(); }
809
810 #ifdef wxSTD_STRING_COMPATIBILITY
811 // std::string compatibility functions
812
813 // standard types
814 typedef wxChar value_type;
815 typedef size_t size_type;
816 typedef value_type *iterator;
817 typedef const value_type *const_iterator;
818
819 // an 'invalid' value for string index
820 static const size_t npos;
821
822 // constructors
823 // take nLen chars starting at nPos
824 wxString(const wxString& str, size_t nPos, size_t nLen)
825 : m_pchData(NULL)
826 {
827 wxASSERT_MSG( str.GetStringData()->IsValid(),
828 _T("did you forget to call UngetWriteBuf()?") );
829
830 InitWith(str.c_str(), nPos, nLen == npos ? 0 : nLen);
831 }
832 // take all characters from pStart to pEnd
833 wxString(const void *pStart, const void *pEnd);
834
835 // lib.string.capacity
836 // return the length of the string
837 size_t size() const { return Len(); }
838 // return the length of the string
839 size_t length() const { return Len(); }
840 // return the maximum size of the string
841 size_t max_size() const { return wxSTRING_MAXLEN; }
842 // resize the string, filling the space with c if c != 0
843 void resize(size_t nSize, wxChar ch = wxT('\0'));
844 // delete the contents of the string
845 void clear() { Empty(); }
846 // returns true if the string is empty
847 bool empty() const { return IsEmpty(); }
848 // inform string about planned change in size
849 void reserve(size_t sz) { Alloc(sz); }
850
851 // lib.string.access
852 // return the character at position n
853 wxChar at(size_t n) const { return GetChar(n); }
854 // returns the writable character at position n
855 wxChar& at(size_t n) { return GetWritableChar(n); }
856
857 // first valid index position
858 const_iterator begin() const { return wx_str(); }
859 // position one after the last valid one
860 const_iterator end() const { return wx_str() + length(); }
861
862 // first valid index position
863 iterator begin() { CopyBeforeWrite(); return m_pchData; }
864 // position one after the last valid one
865 iterator end() { CopyBeforeWrite(); return m_pchData + length(); }
866
867 // lib.string.modifiers
868 // append a string
869 wxString& append(const wxString& str)
870 { *this += str; return *this; }
871 // append elements str[pos], ..., str[pos+n]
872 wxString& append(const wxString& str, size_t pos, size_t n)
873 { ConcatSelf(n, str.c_str() + pos); return *this; }
874 // append first n (or all if n == npos) characters of sz
875 wxString& append(const wxChar *sz, size_t n = npos)
876 { ConcatSelf(n == npos ? wxStrlen(sz) : n, sz); return *this; }
877
878 // append n copies of ch
879 wxString& append(size_t n, wxChar ch) { return Pad(n, ch); }
880
881 // same as `this_string = str'
882 wxString& assign(const wxString& str)
883 { return *this = str; }
884 // same as ` = str[pos..pos + n]
885 wxString& assign(const wxString& str, size_t pos, size_t n)
886 { Empty(); return Append(str.c_str() + pos, n); }
887 // same as `= first n (or all if n == npos) characters of sz'
888 wxString& assign(const wxChar *sz, size_t n = npos)
889 { Empty(); return Append(sz, n == npos ? wxStrlen(sz) : n); }
890 // same as `= n copies of ch'
891 wxString& assign(size_t n, wxChar ch)
892 { Empty(); return Append(ch, n); }
893
894 // insert another string
895 wxString& insert(size_t nPos, const wxString& str);
896 // insert n chars of str starting at nStart (in str)
897 wxString& insert(size_t nPos, const wxString& str, size_t nStart, size_t n)
898 { return insert(nPos, wxString((const wxChar *)str + nStart, n)); }
899
900 // insert first n (or all if n == npos) characters of sz
901 wxString& insert(size_t nPos, const wxChar *sz, size_t n = npos)
902 { return insert(nPos, wxString(sz, n)); }
903 // insert n copies of ch
904 wxString& insert(size_t nPos, size_t n, wxChar ch)
905 { return insert(nPos, wxString(ch, n)); }
906
907 // delete characters from nStart to nStart + nLen
908 wxString& erase(size_t nStart = 0, size_t nLen = npos);
909
910 // replaces the substring of length nLen starting at nStart
911 wxString& replace(size_t nStart, size_t nLen, const wxChar* sz);
912 // replaces the substring with nCount copies of ch
913 wxString& replace(size_t nStart, size_t nLen, size_t nCount, wxChar ch);
914 // replaces a substring with another substring
915 wxString& replace(size_t nStart, size_t nLen,
916 const wxString& str, size_t nStart2, size_t nLen2);
917 // replaces the substring with first nCount chars of sz
918 wxString& replace(size_t nStart, size_t nLen,
919 const wxChar* sz, size_t nCount);
920
921 // swap two strings
922 void swap(wxString& str);
923
924 // All find() functions take the nStart argument which specifies the
925 // position to start the search on, the default value is 0. All functions
926 // return npos if there were no match.
927
928 // find a substring
929 size_t find(const wxString& str, size_t nStart = 0) const;
930
931 // VC++ 1.5 can't cope with this syntax.
932 #if !defined(__VISUALC__) || defined(__WIN32__)
933 // find first n characters of sz
934 size_t find(const wxChar* sz, size_t nStart = 0, size_t n = npos) const;
935 #endif // VC++ 1.5
936
937 // Gives a duplicate symbol (presumably a case-insensitivity problem)
938 #if !defined(__BORLANDC__)
939 // find the first occurence of character ch after nStart
940 size_t find(wxChar ch, size_t nStart = 0) const;
941 #endif
942 // rfind() family is exactly like find() but works right to left
943
944 // as find, but from the end
945 size_t rfind(const wxString& str, size_t nStart = npos) const;
946
947 // VC++ 1.5 can't cope with this syntax.
948 #if !defined(__VISUALC__) || defined(__WIN32__)
949 // as find, but from the end
950 size_t rfind(const wxChar* sz, size_t nStart = npos,
951 size_t n = npos) const;
952 // as find, but from the end
953 size_t rfind(wxChar ch, size_t nStart = npos) const;
954 #endif // VC++ 1.5
955
956 // find first/last occurence of any character in the set
957
958 // as strpbrk() but starts at nStart, returns npos if not found
959 size_t find_first_of(const wxString& str, size_t nStart = 0) const
960 { return find_first_of(str.c_str(), nStart); }
961 // same as above
962 size_t find_first_of(const wxChar* sz, size_t nStart = 0) const;
963 // same as find(char, size_t)
964 size_t find_first_of(wxChar c, size_t nStart = 0) const
965 { return find(c, nStart); }
966 // find the last (starting from nStart) char from str in this string
967 size_t find_last_of (const wxString& str, size_t nStart = npos) const
968 { return find_last_of(str.c_str(), nStart); }
969 // same as above
970 size_t find_last_of (const wxChar* sz, size_t nStart = npos) const;
971 // same as above
972 size_t find_last_of(wxChar c, size_t nStart = npos) const
973 { return rfind(c, nStart); }
974
975 // find first/last occurence of any character not in the set
976
977 // as strspn() (starting from nStart), returns npos on failure
978 size_t find_first_not_of(const wxString& str, size_t nStart = 0) const
979 { return find_first_not_of(str.c_str(), nStart); }
980 // same as above
981 size_t find_first_not_of(const wxChar* sz, size_t nStart = 0) const;
982 // same as above
983 size_t find_first_not_of(wxChar ch, size_t nStart = 0) const;
984 // as strcspn()
985 size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
986 { return find_first_not_of(str.c_str(), nStart); }
987 // same as above
988 size_t find_last_not_of(const wxChar* sz, size_t nStart = npos) const;
989 // same as above
990 size_t find_last_not_of(wxChar ch, size_t nStart = npos) const;
991
992 // All compare functions return -1, 0 or 1 if the [sub]string is less,
993 // equal or greater than the compare() argument.
994
995 // just like strcmp()
996 int compare(const wxString& str) const { return Cmp(str); }
997 // comparison with a substring
998 int compare(size_t nStart, size_t nLen, const wxString& str) const
999 { return Mid(nStart, nLen).Cmp(str); }
1000 // comparison of 2 substrings
1001 int compare(size_t nStart, size_t nLen,
1002 const wxString& str, size_t nStart2, size_t nLen2) const
1003 { return Mid(nStart, nLen).Cmp(str.Mid(nStart2, nLen2)); }
1004 // just like strcmp()
1005 int compare(const wxChar* sz) const { return Cmp(sz); }
1006 // substring comparison with first nCount characters of sz
1007 int compare(size_t nStart, size_t nLen,
1008 const wxChar* sz, size_t nCount = npos) const
1009 { return Mid(nStart, nLen).Cmp(wxString(sz, nCount)); }
1010
1011 // substring extraction
1012 wxString substr(size_t nStart = 0, size_t nLen = npos) const
1013 { return Mid(nStart, nLen); }
1014 #endif // wxSTD_STRING_COMPATIBILITY
1015 };
1016
1017 // define wxArrayString, for compatibility
1018 #if WXWIN_COMPATIBILITY_2_4 && !wxUSE_STL
1019 #include "wx/arrstr.h"
1020 #endif
1021
1022 // ----------------------------------------------------------------------------
1023 // wxStringBuffer: a tiny class allowing to get a writable pointer into string
1024 // ----------------------------------------------------------------------------
1025
1026 class WXDLLIMPEXP_BASE wxStringBuffer
1027 {
1028 public:
1029 wxStringBuffer(wxString& str, size_t lenWanted = 1024)
1030 : m_str(str), m_buf(NULL)
1031 { m_buf = m_str.GetWriteBuf(lenWanted); }
1032
1033 ~wxStringBuffer() { m_str.UngetWriteBuf(); }
1034
1035 operator wxChar*() const { return m_buf; }
1036
1037 private:
1038 wxString& m_str;
1039 wxChar *m_buf;
1040
1041 DECLARE_NO_COPY_CLASS(wxStringBuffer)
1042 };
1043
1044 // ---------------------------------------------------------------------------
1045 // wxString comparison functions: operator versions are always case sensitive
1046 // ---------------------------------------------------------------------------
1047
1048 inline bool operator==(const wxString& s1, const wxString& s2)
1049 { return (s1.Len() == s2.Len()) && (s1.Cmp(s2) == 0); }
1050 inline bool operator==(const wxString& s1, const wxChar * s2)
1051 { return s1.Cmp(s2) == 0; }
1052 inline bool operator==(const wxChar * s1, const wxString& s2)
1053 { return s2.Cmp(s1) == 0; }
1054 inline bool operator!=(const wxString& s1, const wxString& s2)
1055 { return (s1.Len() != s2.Len()) || (s1.Cmp(s2) != 0); }
1056 inline bool operator!=(const wxString& s1, const wxChar * s2)
1057 { return s1.Cmp(s2) != 0; }
1058 inline bool operator!=(const wxChar * s1, const wxString& s2)
1059 { return s2.Cmp(s1) != 0; }
1060 inline bool operator< (const wxString& s1, const wxString& s2)
1061 { return s1.Cmp(s2) < 0; }
1062 inline bool operator< (const wxString& s1, const wxChar * s2)
1063 { return s1.Cmp(s2) < 0; }
1064 inline bool operator< (const wxChar * s1, const wxString& s2)
1065 { return s2.Cmp(s1) > 0; }
1066 inline bool operator> (const wxString& s1, const wxString& s2)
1067 { return s1.Cmp(s2) > 0; }
1068 inline bool operator> (const wxString& s1, const wxChar * s2)
1069 { return s1.Cmp(s2) > 0; }
1070 inline bool operator> (const wxChar * s1, const wxString& s2)
1071 { return s2.Cmp(s1) < 0; }
1072 inline bool operator<=(const wxString& s1, const wxString& s2)
1073 { return s1.Cmp(s2) <= 0; }
1074 inline bool operator<=(const wxString& s1, const wxChar * s2)
1075 { return s1.Cmp(s2) <= 0; }
1076 inline bool operator<=(const wxChar * s1, const wxString& s2)
1077 { return s2.Cmp(s1) >= 0; }
1078 inline bool operator>=(const wxString& s1, const wxString& s2)
1079 { return s1.Cmp(s2) >= 0; }
1080 inline bool operator>=(const wxString& s1, const wxChar * s2)
1081 { return s1.Cmp(s2) >= 0; }
1082 inline bool operator>=(const wxChar * s1, const wxString& s2)
1083 { return s2.Cmp(s1) <= 0; }
1084
1085 // comparison with char
1086 inline bool operator==(wxChar c, const wxString& s) { return s.IsSameAs(c); }
1087 inline bool operator==(const wxString& s, wxChar c) { return s.IsSameAs(c); }
1088 inline bool operator!=(wxChar c, const wxString& s) { return !s.IsSameAs(c); }
1089 inline bool operator!=(const wxString& s, wxChar c) { return !s.IsSameAs(c); }
1090
1091 #if wxUSE_UNICODE
1092 inline bool operator==(const wxString& s1, const wxWCharBuffer& s2)
1093 { return (s1.Cmp((const wchar_t *)s2) == 0); }
1094 inline bool operator==(const wxWCharBuffer& s1, const wxString& s2)
1095 { return (s2.Cmp((const wchar_t *)s1) == 0); }
1096 inline bool operator!=(const wxString& s1, const wxWCharBuffer& s2)
1097 { return (s1.Cmp((const wchar_t *)s2) != 0); }
1098 inline bool operator!=(const wxWCharBuffer& s1, const wxString& s2)
1099 { return (s2.Cmp((const wchar_t *)s1) != 0); }
1100 #else // !wxUSE_UNICODE
1101 inline bool operator==(const wxString& s1, const wxCharBuffer& s2)
1102 { return (s1.Cmp((const char *)s2) == 0); }
1103 inline bool operator==(const wxCharBuffer& s1, const wxString& s2)
1104 { return (s2.Cmp((const char *)s1) == 0); }
1105 inline bool operator!=(const wxString& s1, const wxCharBuffer& s2)
1106 { return (s1.Cmp((const char *)s2) != 0); }
1107 inline bool operator!=(const wxCharBuffer& s1, const wxString& s2)
1108 { return (s2.Cmp((const char *)s1) != 0); }
1109 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
1110
1111 wxString WXDLLIMPEXP_BASE operator+(const wxString& string1, const wxString& string2);
1112 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxChar ch);
1113 wxString WXDLLIMPEXP_BASE operator+(wxChar ch, const wxString& string);
1114 wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const wxChar *psz);
1115 wxString WXDLLIMPEXP_BASE operator+(const wxChar *psz, const wxString& string);
1116 #if wxUSE_UNICODE
1117 inline wxString operator+(const wxString& string, const wxWCharBuffer& buf)
1118 { return string + (const wchar_t *)buf; }
1119 inline wxString operator+(const wxWCharBuffer& buf, const wxString& string)
1120 { return (const wchar_t *)buf + string; }
1121 #else // !wxUSE_UNICODE
1122 inline wxString operator+(const wxString& string, const wxCharBuffer& buf)
1123 { return string + (const char *)buf; }
1124 inline wxString operator+(const wxCharBuffer& buf, const wxString& string)
1125 { return (const char *)buf + string; }
1126 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
1127
1128 // ---------------------------------------------------------------------------
1129 // Implementation only from here until the end of file
1130 // ---------------------------------------------------------------------------
1131
1132 // don't pollute the library user's name space
1133 #undef wxASSERT_VALID_INDEX
1134
1135 #if defined(wxSTD_STRING_COMPATIBILITY) && wxUSE_STD_IOSTREAM
1136
1137 #include "wx/iosfwrap.h"
1138
1139 WXDLLIMPEXP_BASE wxSTD istream& operator>>(wxSTD istream&, wxString&);
1140 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxString&);
1141
1142 #endif // wxSTD_STRING_COMPATIBILITY
1143
1144 #endif // _WX_WXSTRINGH__