1 ///////////////////////////////////////////////////////////////////////////////
3 // Purpose: wxString and wxArrayString classes
4 // Author: Vadim Zeitlin
8 // Copyright: (c) 1998 Vadim Zeitlin <zeitlin@dptmaths.ens-cachan.fr>
9 // Licence: wxWindows license
10 ///////////////////////////////////////////////////////////////////////////////
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.
18 #ifndef _WX_WXSTRINGH__
19 #define _WX_WXSTRINGH__
22 #pragma interface "string.h"
25 // ----------------------------------------------------------------------------
26 // conditinal compilation
27 // ----------------------------------------------------------------------------
29 // compile the std::string compatibility functions if defined
30 #define wxSTD_STRING_COMPATIBILITY
32 // define to derive wxString from wxObject (deprecated!)
33 #ifdef WXSTRING_IS_WXOBJECT
34 #undef WXSTRING_IS_WXOBJECT
37 // ----------------------------------------------------------------------------
39 // ----------------------------------------------------------------------------
41 #if defined(__WXMAC__) || defined(__VISAGECPP__)
49 #if defined(__VISAGECPP__) && __IBMCPP__ >= 400
50 // problem in VACPP V4 with including stdlib.h multiple times
51 // strconv includes it anyway
65 #include <strings.h> // for strcasecmp()
66 #endif // HAVE_STRINGS_H
68 #include "wx/defs.h" // everybody should include this
69 #include "wx/wxchar.h" // for wxChar
70 #include "wx/buffer.h" // for wxCharBuffer
71 #include "wx/strconv.h" // for wxConvertXXX() macros and wxMBConv classes
74 #ifdef WXSTRING_IS_WXOBJECT
75 #include "wx/object.h" // base class
79 // ---------------------------------------------------------------------------
81 // ---------------------------------------------------------------------------
84 #define WXSTRINGCAST (wxChar *)(const wxChar *)
85 #define wxCSTRINGCAST (wxChar *)(const wxChar *)
86 #define wxMBSTRINGCAST (char *)(const char *)
87 #define wxWCSTRINGCAST (wchar_t *)(const wchar_t *)
89 // implementation only
90 #define ASSERT_VALID_INDEX(i) wxASSERT( (unsigned)(i) <= Len() )
92 // ----------------------------------------------------------------------------
94 // ----------------------------------------------------------------------------
96 #if defined(__VISAGECPP__) && __IBMCPP__ >= 400
97 // must define this static for VA or else you get multiply defined symbols everywhere
98 extern const unsigned int wxSTRING_MAXLEN
;
101 // maximum possible length for a string means "take all string" everywhere
102 // (as sizeof(StringData) is unknown here, we substract 100)
103 const unsigned int wxSTRING_MAXLEN
= UINT_MAX
- 100;
107 // ----------------------------------------------------------------------------
109 // ----------------------------------------------------------------------------
111 // global pointer to empty string
112 WXDLLEXPORT_DATA(extern const wxChar
*) wxEmptyString
;
114 // ---------------------------------------------------------------------------
115 // global functions complementing standard C string library replacements for
116 // strlen() and portable strcasecmp()
117 //---------------------------------------------------------------------------
119 // Use wxXXX() functions from wxchar.h instead! These functions are for
120 // backwards compatibility only.
122 // checks whether the passed in pointer is NULL and if the string is empty
123 inline bool IsEmpty(const char *p
) { return (!p
|| !*p
); }
125 // safe version of strlen() (returns 0 if passed NULL pointer)
126 inline size_t Strlen(const char *psz
)
127 { return psz
? strlen(psz
) : 0; }
129 // portable strcasecmp/_stricmp
130 inline int Stricmp(const char *psz1
, const char *psz2
)
132 #if defined(__VISUALC__) || ( defined(__MWERKS__) && defined(__INTEL__) )
133 return _stricmp(psz1
, psz2
);
134 #elif defined(__SC__)
135 return _stricmp(psz1
, psz2
);
136 #elif defined(__SALFORDC__)
137 return stricmp(psz1
, psz2
);
138 #elif defined(__BORLANDC__)
139 return stricmp(psz1
, psz2
);
140 #elif defined(__WATCOMC__)
141 return stricmp(psz1
, psz2
);
142 #elif defined(__DJGPP__)
143 return stricmp(psz1
, psz2
);
144 #elif defined(__EMX__)
145 return stricmp(psz1
, psz2
);
146 #elif defined(__WXPM__)
147 return stricmp(psz1
, psz2
);
148 #elif defined(__UNIX__) || defined(__GNUWIN32__)
149 return strcasecmp(psz1
, psz2
);
150 #elif defined(__MWERKS__) && !defined(__INTEL__)
151 register char c1
, c2
;
153 c1
= tolower(*psz1
++);
154 c2
= tolower(*psz2
++);
155 } while ( c1
&& (c1
== c2
) );
159 // almost all compilers/libraries provide this function (unfortunately under
160 // different names), that's why we don't implement our own which will surely
161 // be more efficient than this code (uncomment to use):
163 register char c1, c2;
165 c1 = tolower(*psz1++);
166 c2 = tolower(*psz2++);
167 } while ( c1 && (c1 == c2) );
172 #error "Please define string case-insensitive compare for your OS/compiler"
173 #endif // OS/compiler
176 // wxSnprintf() is like snprintf() if it's available and sprintf() (always
177 // available, but dangerous!) if not
178 extern int WXDLLEXPORT
wxSnprintf(wxChar
*buf
, size_t len
,
179 const wxChar
*format
, ...);
181 // and wxVsnprintf() is like vsnprintf() or vsprintf()
182 extern int WXDLLEXPORT
wxVsnprintf(wxChar
*buf
, size_t len
,
183 const wxChar
*format
, va_list argptr
);
185 // return an empty wxString
186 class WXDLLEXPORT wxString
; // not yet defined
187 inline const wxString
& wxGetEmptyString() { return *(wxString
*)&wxEmptyString
; }
189 // ---------------------------------------------------------------------------
190 // string data prepended with some housekeeping info (used by wxString class),
191 // is never used directly (but had to be put here to allow inlining)
192 // ---------------------------------------------------------------------------
194 struct WXDLLEXPORT wxStringData
196 int nRefs
; // reference count
197 size_t nDataLength
, // actual string length
198 nAllocLength
; // allocated memory size
200 // mimics declaration 'wxChar data[nAllocLength]'
201 wxChar
* data() const { return (wxChar
*)(this + 1); }
203 // empty string has a special ref count so it's never deleted
204 bool IsEmpty() const { return (nRefs
== -1); }
205 bool IsShared() const { return (nRefs
> 1); }
208 void Lock() { if ( !IsEmpty() ) nRefs
++; }
210 // VC++ will refuse to inline this function but profiling shows that it
212 #if defined(__VISUALC__) && (__VISUALC__ >= 1200)
215 void Unlock() { if ( !IsEmpty() && --nRefs
== 0) free(this); }
217 // if we had taken control over string memory (GetWriteBuf), it's
218 // intentionally put in invalid state
219 void Validate(bool b
) { nRefs
= (b
? 1 : 0); }
220 bool IsValid() const { return (nRefs
!= 0); }
223 // ---------------------------------------------------------------------------
224 // This is (yet another one) String class for C++ programmers. It doesn't use
225 // any of "advanced" C++ features (i.e. templates, exceptions, namespaces...)
226 // thus you should be able to compile it with practicaly any C++ compiler.
227 // This class uses copy-on-write technique, i.e. identical strings share the
228 // same memory as long as neither of them is changed.
230 // This class aims to be as compatible as possible with the new standard
231 // std::string class, but adds some additional functions and should be at
232 // least as efficient than the standard implementation.
234 // Performance note: it's more efficient to write functions which take "const
235 // String&" arguments than "const char *" if you assign the argument to
238 // It was compiled and tested under Win32, Linux (libc 5 & 6), Solaris 5.5.
241 // - ressource support (string tables in ressources)
242 // - more wide character (UNICODE) support
243 // - regular expressions support
244 // ---------------------------------------------------------------------------
246 #ifdef WXSTRING_IS_WXOBJECT
247 class WXDLLEXPORT wxString
: public wxObject
249 DECLARE_DYNAMIC_CLASS(wxString
)
250 #else //WXSTRING_IS_WXOBJECT
251 class WXDLLEXPORT wxString
253 #endif //WXSTRING_IS_WXOBJECT
255 friend class WXDLLEXPORT wxArrayString
;
257 // NB: special care was taken in arranging the member functions in such order
258 // that all inline functions can be effectively inlined, verify that all
259 // performace critical functions are still inlined if you change order!
261 // points to data preceded by wxStringData structure with ref count info
264 // accessor to string data
265 wxStringData
* GetStringData() const { return (wxStringData
*)m_pchData
- 1; }
267 // string (re)initialization functions
268 // initializes the string to the empty value (must be called only from
269 // ctors, use Reinit() otherwise)
270 void Init() { m_pchData
= (wxChar
*)wxEmptyString
; }
271 // initializaes the string with (a part of) C-string
272 void InitWith(const wxChar
*psz
, size_t nPos
= 0, size_t nLen
= wxSTRING_MAXLEN
);
273 // as Init, but also frees old data
274 void Reinit() { GetStringData()->Unlock(); Init(); }
277 // allocates memory for string of lenght nLen
278 void AllocBuffer(size_t nLen
);
279 // copies data to another string
280 void AllocCopy(wxString
&, int, int) const;
281 // effectively copies data to string
282 void AssignCopy(size_t, const wxChar
*);
284 // append a (sub)string
285 void ConcatSelf(int nLen
, const wxChar
*src
);
287 // functions called before writing to the string: they copy it if there
288 // are other references to our data (should be the only owner when writing)
289 void CopyBeforeWrite();
290 void AllocBeforeWrite(size_t);
292 // if we hadn't made these operators private, it would be possible to
293 // compile "wxString s; s = 17;" without any warnings as 17 is implicitly
294 // converted to char in C and we do have operator=(char)
296 // NB: we don't need other versions (short/long and unsigned) as attempt
297 // to assign another numeric type to wxString will now result in
298 // ambiguity between operator=(char) and operator=(int)
299 wxString
& operator=(int);
301 // these methods are not implemented - there is _no_ conversion from int to
302 // string, you're doing something wrong if the compiler wants to call it!
304 // try `s << i' or `s.Printf("%d", i)' instead
308 // constructors and destructor
309 // ctor for an empty string
310 wxString() { Init(); }
312 wxString(const wxString
& stringSrc
)
314 wxASSERT( stringSrc
.GetStringData()->IsValid() );
316 if ( stringSrc
.IsEmpty() ) {
317 // nothing to do for an empty string
321 m_pchData
= stringSrc
.m_pchData
; // share same data
322 GetStringData()->Lock(); // => one more copy
325 // string containing nRepeat copies of ch
326 wxString(wxChar ch
, size_t nRepeat
= 1);
327 // ctor takes first nLength characters from C string
328 // (default value of wxSTRING_MAXLEN means take all the string)
329 wxString(const wxChar
*psz
, size_t nLength
= wxSTRING_MAXLEN
)
330 { InitWith(psz
, 0, nLength
); }
331 wxString(const wxChar
*psz
, wxMBConv
& WXUNUSED(conv
), size_t nLength
= wxSTRING_MAXLEN
)
332 { InitWith(psz
, 0, nLength
); }
335 // from multibyte string
336 // (NB: nLength is right now number of Unicode characters, not
337 // characters in psz! So try not to use it yet!)
338 wxString(const char *psz
, wxMBConv
& conv
= wxConvLibc
, size_t nLength
= wxSTRING_MAXLEN
);
339 // from wxWCharBuffer (i.e. return from wxGetString)
340 wxString(const wxWCharBuffer
& psz
)
341 { InitWith(psz
, 0, wxSTRING_MAXLEN
); }
343 // from C string (for compilers using unsigned char)
344 wxString(const unsigned char* psz
, size_t nLength
= wxSTRING_MAXLEN
)
345 { InitWith((const char*)psz
, 0, nLength
); }
348 // from wide (Unicode) string
349 wxString(const wchar_t *pwz
, wxMBConv
& conv
= wxConvLibc
, size_t nLength
= wxSTRING_MAXLEN
);
350 #endif // !wxUSE_WCHAR_T
353 wxString(const wxCharBuffer
& psz
)
354 { InitWith(psz
, 0, wxSTRING_MAXLEN
); }
355 #endif // Unicode/ANSI
357 // dtor is not virtual, this class must not be inherited from!
358 ~wxString() { GetStringData()->Unlock(); }
360 // generic attributes & operations
361 // as standard strlen()
362 size_t Len() const { return GetStringData()->nDataLength
; }
363 // string contains any characters?
364 bool IsEmpty() const { return Len() == 0; }
365 // empty string is "FALSE", so !str will return TRUE
366 bool operator!() const { return IsEmpty(); }
367 // truncate the string to given length
368 wxString
& Truncate(size_t uiLen
);
369 // empty string contents
374 wxASSERT_MSG( IsEmpty(), _T("string not empty after call to Empty()?") );
376 // empty the string and free memory
379 if ( !GetStringData()->IsEmpty() )
382 wxASSERT( GetStringData()->nDataLength
== 0 ); // should be empty
383 wxASSERT( GetStringData()->nAllocLength
== 0 ); // and not own any memory
388 bool IsAscii() const;
390 bool IsNumber() const;
394 // data access (all indexes are 0 based)
396 wxChar
GetChar(size_t n
) const
397 { ASSERT_VALID_INDEX( n
); return m_pchData
[n
]; }
399 wxChar
& GetWritableChar(size_t n
)
400 { ASSERT_VALID_INDEX( n
); CopyBeforeWrite(); return m_pchData
[n
]; }
402 void SetChar(size_t n
, wxChar ch
)
403 { ASSERT_VALID_INDEX( n
); CopyBeforeWrite(); m_pchData
[n
] = ch
; }
405 // get last character
407 { wxASSERT( !IsEmpty() ); return m_pchData
[Len() - 1]; }
408 // get writable last character
410 { wxASSERT( !IsEmpty() ); CopyBeforeWrite(); return m_pchData
[Len()-1]; }
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).
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
430 The only real fix is, of course, to remove all versions but the one
434 // operator version of GetChar
435 wxChar
operator[](size_t n
) const
436 { ASSERT_VALID_INDEX( n
); return m_pchData
[n
]; }
438 // operator version of GetChar
439 wxChar
operator[](int n
) const
440 { ASSERT_VALID_INDEX( n
); return m_pchData
[n
]; }
442 // operator version of GetWriteableChar
443 wxChar
& operator[](size_t n
)
444 { ASSERT_VALID_INDEX( n
); CopyBeforeWrite(); return m_pchData
[n
]; }
446 #ifndef wxSIZE_T_IS_UINT
447 // operator version of GetChar
448 wxChar
operator[](unsigned int n
) const
449 { ASSERT_VALID_INDEX( n
); return m_pchData
[n
]; }
451 // operator version of GetWriteableChar
452 wxChar
& operator[](unsigned int n
)
453 { ASSERT_VALID_INDEX( n
); CopyBeforeWrite(); return m_pchData
[n
]; }
454 #endif // size_t != unsigned int
456 // implicit conversion to C string
457 operator const wxChar
*() const { return m_pchData
; }
458 // explicit conversion to C string (use this with printf()!)
459 const wxChar
* c_str() const { return m_pchData
; }
460 // identical to c_str()
461 const wxChar
* wx_str() const { return m_pchData
; }
462 // identical to c_str()
463 const wxChar
* GetData() const { return m_pchData
; }
465 // conversions with (possible) format convertions: have to return a
466 // buffer with temporary data
468 // the functions defined (in either Unicode or ANSI) mode are mb_str() to
469 // return an ANSI (multibyte) string, wc_str() to return a wide string and
470 // fn_str() to return a string which should be used with the OS APIs
471 // accepting the file names. The return value is always the same, but the
472 // type differs because a function may either return pointer to the buffer
473 // directly or have to use intermediate buffer for translation.
475 const wxCharBuffer
mb_str(wxMBConv
& conv
= wxConvLibc
) const
476 { return conv
.cWC2MB(m_pchData
); }
478 const wxWX2MBbuf
mbc_str() const { return mb_str(*wxConvCurrent
); }
480 const wxChar
* wc_str() const { return m_pchData
; }
482 // for compatibility with !wxUSE_UNICODE version
483 const wxChar
* wc_str(wxMBConv
& WXUNUSED(conv
)) const { return m_pchData
; }
486 const wxCharBuffer
fn_str() const { return mb_str(wxConvFile
); }
488 const wxChar
* fn_str() const { return m_pchData
; }
489 #endif // wxMBFILES/!wxMBFILES
491 const wxChar
* mb_str() const { return m_pchData
; }
493 // for compatibility with wxUSE_UNICODE version
494 const wxChar
* mb_str(wxMBConv
& WXUNUSED(conv
)) const { return m_pchData
; }
496 const wxWX2MBbuf
mbc_str() const { return mb_str(); }
499 const wxWCharBuffer
wc_str(wxMBConv
& conv
) const
500 { return conv
.cMB2WC(m_pchData
); }
501 #endif // wxUSE_WCHAR_T
503 const wxChar
* fn_str() const { return m_pchData
; }
504 #endif // Unicode/ANSI
506 // overloaded assignment
507 // from another wxString
508 wxString
& operator=(const wxString
& stringSrc
);
510 wxString
& operator=(wxChar ch
);
512 wxString
& operator=(const wxChar
*psz
);
514 // from wxWCharBuffer
515 wxString
& operator=(const wxWCharBuffer
& psz
) { return operator=((const wchar_t *)psz
); }
517 // from another kind of C string
518 wxString
& operator=(const unsigned char* psz
);
520 // from a wide string
521 wxString
& operator=(const wchar_t *pwz
);
524 wxString
& operator=(const wxCharBuffer
& psz
) { return operator=((const char *)psz
); }
525 #endif // Unicode/ANSI
527 // string concatenation
528 // in place concatenation
530 Concatenate and return the result. Note that the left to right
531 associativity of << allows to write things like "str << str1 << str2
532 << ..." (unlike with +=)
535 wxString
& operator<<(const wxString
& s
)
537 wxASSERT( s
.GetStringData()->IsValid() );
539 ConcatSelf(s
.Len(), s
);
542 // string += C string
543 wxString
& operator<<(const wxChar
*psz
)
544 { ConcatSelf(wxStrlen(psz
), psz
); return *this; }
546 wxString
& operator<<(wxChar ch
) { ConcatSelf(1, &ch
); return *this; }
549 void operator+=(const wxString
& s
) { (void)operator<<(s
); }
550 // string += C string
551 void operator+=(const wxChar
*psz
) { (void)operator<<(psz
); }
553 void operator+=(wxChar ch
) { (void)operator<<(ch
); }
555 // string += buffer (i.e. from wxGetString)
557 wxString
& operator<<(const wxWCharBuffer
& s
) { (void)operator<<((const wchar_t *)s
); return *this; }
558 void operator+=(const wxWCharBuffer
& s
) { (void)operator<<((const wchar_t *)s
); }
560 wxString
& operator<<(const wxCharBuffer
& s
) { (void)operator<<((const char *)s
); return *this; }
561 void operator+=(const wxCharBuffer
& s
) { (void)operator<<((const char *)s
); }
564 // string += C string
565 wxString
& Append(const wxChar
* psz
)
566 { ConcatSelf(wxStrlen(psz
), psz
); return *this; }
567 // append count copies of given character
568 wxString
& Append(wxChar ch
, size_t count
= 1u)
569 { wxString
str(ch
, count
); return *this << str
; }
570 wxString
& Append(const wxChar
* psz
, size_t nLen
)
571 { ConcatSelf(nLen
, psz
); return *this; }
573 // prepend a string, return the string itself
574 wxString
& Prepend(const wxString
& str
)
575 { *this = str
+ *this; return *this; }
577 // non-destructive concatenation
579 friend wxString WXDLLEXPORT
operator+(const wxString
& string1
, const wxString
& string2
);
581 friend wxString WXDLLEXPORT
operator+(const wxString
& string
, wxChar ch
);
583 friend wxString WXDLLEXPORT
operator+(wxChar ch
, const wxString
& string
);
585 friend wxString WXDLLEXPORT
operator+(const wxString
& string
, const wxChar
*psz
);
587 friend wxString WXDLLEXPORT
operator+(const wxChar
*psz
, const wxString
& string
);
589 // stream-like functions
590 // insert an int into string
591 wxString
& operator<<(int i
)
592 { return (*this) << Format(_T("%d"), i
); }
593 // insert an unsigned int into string
594 wxString
& operator<<(unsigned int ui
)
595 { return (*this) << Format(_T("%u"), ui
); }
596 // insert a long into string
597 wxString
& operator<<(long l
)
598 { return (*this) << Format(_T("%ld"), l
); }
599 // insert an unsigned long into string
600 wxString
& operator<<(unsigned long ul
)
601 { return (*this) << Format(_T("%lu"), ul
); }
602 // insert a float into string
603 wxString
& operator<<(float f
)
604 { return (*this) << Format(_T("%f"), f
); }
605 // insert a double into string
606 wxString
& operator<<(double d
)
607 { return (*this) << Format(_T("%g"), d
); }
610 // case-sensitive comparison (returns a value < 0, = 0 or > 0)
611 int Cmp(const wxChar
*psz
) const { return wxStrcmp(c_str(), psz
); }
612 // same as Cmp() but not case-sensitive
613 int CmpNoCase(const wxChar
*psz
) const { return wxStricmp(c_str(), psz
); }
614 // test for the string equality, either considering case or not
615 // (if compareWithCase then the case matters)
616 bool IsSameAs(const wxChar
*psz
, bool compareWithCase
= TRUE
) const
617 { return (compareWithCase
? Cmp(psz
) : CmpNoCase(psz
)) == 0; }
618 // comparison with a signle character: returns TRUE if equal
619 bool IsSameAs(wxChar c
, bool compareWithCase
= TRUE
) const
621 return (Len() == 1) && (compareWithCase
? GetChar(0u) == c
622 : wxToupper(GetChar(0u)) == wxToupper(c
));
625 // simple sub-string extraction
626 // return substring starting at nFirst of length nCount (or till the end
627 // if nCount = default value)
628 wxString
Mid(size_t nFirst
, size_t nCount
= wxSTRING_MAXLEN
) const;
630 // operator version of Mid()
631 wxString
operator()(size_t start
, size_t len
) const
632 { return Mid(start
, len
); }
634 // check that the string starts with prefix and return the rest of the
635 // string in the provided pointer if it is not NULL, otherwise return
637 bool StartsWith(const wxChar
*prefix
, wxString
*rest
= NULL
) const;
639 // get first nCount characters
640 wxString
Left(size_t nCount
) const;
641 // get last nCount characters
642 wxString
Right(size_t nCount
) const;
643 // get all characters before the first occurance of ch
644 // (returns the whole string if ch not found)
645 wxString
BeforeFirst(wxChar ch
) const;
646 // get all characters before the last occurence of ch
647 // (returns empty string if ch not found)
648 wxString
BeforeLast(wxChar ch
) const;
649 // get all characters after the first occurence of ch
650 // (returns empty string if ch not found)
651 wxString
AfterFirst(wxChar ch
) const;
652 // get all characters after the last occurence of ch
653 // (returns the whole string if ch not found)
654 wxString
AfterLast(wxChar ch
) const;
656 // for compatibility only, use more explicitly named functions above
657 wxString
Before(wxChar ch
) const { return BeforeLast(ch
); }
658 wxString
After(wxChar ch
) const { return AfterFirst(ch
); }
661 // convert to upper case in place, return the string itself
662 wxString
& MakeUpper();
663 // convert to upper case, return the copy of the string
664 // Here's something to remember: BC++ doesn't like returns in inlines.
665 wxString
Upper() const ;
666 // convert to lower case in place, return the string itself
667 wxString
& MakeLower();
668 // convert to lower case, return the copy of the string
669 wxString
Lower() const ;
671 // trimming/padding whitespace (either side) and truncating
672 // remove spaces from left or from right (default) side
673 wxString
& Trim(bool bFromRight
= TRUE
);
674 // add nCount copies chPad in the beginning or at the end (default)
675 wxString
& Pad(size_t nCount
, wxChar chPad
= wxT(' '), bool bFromRight
= TRUE
);
677 // searching and replacing
678 // searching (return starting index, or -1 if not found)
679 int Find(wxChar ch
, bool bFromEnd
= FALSE
) const; // like strchr/strrchr
680 // searching (return starting index, or -1 if not found)
681 int Find(const wxChar
*pszSub
) const; // like strstr
682 // replace first (or all of bReplaceAll) occurences of substring with
683 // another string, returns the number of replacements made
684 size_t Replace(const wxChar
*szOld
,
686 bool bReplaceAll
= TRUE
);
688 // check if the string contents matches a mask containing '*' and '?'
689 bool Matches(const wxChar
*szMask
) const;
691 // conversion to numbers: all functions return TRUE only if the whole string
692 // is a number and put the value of this number into the pointer provided
693 // convert to a signed integer
694 bool ToLong(long *val
, int base
= 0) const;
695 // convert to an unsigned integer
696 bool ToULong(unsigned long *val
, int base
= 0) const;
697 // convert to a double
698 bool ToDouble(double *val
) const;
700 // formated input/output
701 // as sprintf(), returns the number of characters written or < 0 on error
702 int Printf(const wxChar
*pszFormat
, ...);
703 // as vprintf(), returns the number of characters written or < 0 on error
704 int PrintfV(const wxChar
* pszFormat
, va_list argptr
);
706 // returns the string containing the result of Printf() to it
707 static wxString
Format(const wxChar
*pszFormat
, ...);
708 // the same as above, but takes a va_list
709 static wxString
FormatV(const wxChar
*pszFormat
, va_list argptr
);
711 // raw access to string memory
712 // ensure that string has space for at least nLen characters
713 // only works if the data of this string is not shared
714 void Alloc(size_t nLen
);
715 // minimize the string's memory
716 // only works if the data of this string is not shared
718 // get writable buffer of at least nLen bytes. Unget() *must* be called
719 // a.s.a.p. to put string back in a reasonable state!
720 wxChar
*GetWriteBuf(size_t nLen
);
721 // call this immediately after GetWriteBuf() has been used
722 void UngetWriteBuf();
723 void UngetWriteBuf(size_t nLen
);
725 // wxWindows version 1 compatibility functions
728 wxString
SubString(size_t from
, size_t to
) const
729 { return Mid(from
, (to
- from
+ 1)); }
730 // values for second parameter of CompareTo function
731 enum caseCompare
{exact
, ignoreCase
};
732 // values for first parameter of Strip function
733 enum stripType
{leading
= 0x1, trailing
= 0x2, both
= 0x3};
736 int sprintf(const wxChar
*pszFormat
, ...);
739 inline int CompareTo(const wxChar
* psz
, caseCompare cmp
= exact
) const
740 { return cmp
== exact
? Cmp(psz
) : CmpNoCase(psz
); }
743 size_t Length() const { return Len(); }
744 // Count the number of characters
745 int Freq(wxChar ch
) const;
747 void LowerCase() { MakeLower(); }
749 void UpperCase() { MakeUpper(); }
750 // use Trim except that it doesn't change this string
751 wxString
Strip(stripType w
= trailing
) const;
753 // use Find (more general variants not yet supported)
754 size_t Index(const wxChar
* psz
) const { return Find(psz
); }
755 size_t Index(wxChar ch
) const { return Find(ch
); }
757 wxString
& Remove(size_t pos
) { return Truncate(pos
); }
758 wxString
& RemoveLast(size_t n
= 1) { return Truncate(Len() - n
); }
760 wxString
& Remove(size_t nStart
, size_t nLen
) { return erase( nStart
, nLen
); }
763 int First( const wxChar ch
) const { return Find(ch
); }
764 int First( const wxChar
* psz
) const { return Find(psz
); }
765 int First( const wxString
&str
) const { return Find(str
); }
766 int Last( const wxChar ch
) const { return Find(ch
, TRUE
); }
767 bool Contains(const wxString
& str
) const { return Find(str
) != -1; }
770 bool IsNull() const { return IsEmpty(); }
772 #ifdef wxSTD_STRING_COMPATIBILITY
773 // std::string compatibility functions
776 typedef wxChar value_type
;
777 typedef const value_type
*const_iterator
;
779 // an 'invalid' value for string index
780 static const size_t npos
;
783 // take nLen chars starting at nPos
784 wxString(const wxString
& str
, size_t nPos
, size_t nLen
)
786 wxASSERT( str
.GetStringData()->IsValid() );
787 InitWith(str
.c_str(), nPos
, nLen
== npos
? 0 : nLen
);
789 // take all characters from pStart to pEnd
790 wxString(const void *pStart
, const void *pEnd
);
792 // lib.string.capacity
793 // return the length of the string
794 size_t size() const { return Len(); }
795 // return the length of the string
796 size_t length() const { return Len(); }
797 // return the maximum size of the string
798 size_t max_size() const { return wxSTRING_MAXLEN
; }
799 // resize the string, filling the space with c if c != 0
800 void resize(size_t nSize
, wxChar ch
= wxT('\0'));
801 // delete the contents of the string
802 void clear() { Empty(); }
803 // returns true if the string is empty
804 bool empty() const { return IsEmpty(); }
805 // inform string about planned change in size
806 void reserve(size_t size
) { Alloc(size
); }
809 // return the character at position n
810 wxChar
at(size_t n
) const { return GetChar(n
); }
811 // returns the writable character at position n
812 wxChar
& at(size_t n
) { return GetWritableChar(n
); }
814 // first valid index position
815 const_iterator
begin() const { return wx_str(); }
816 // position one after the last valid one
817 const_iterator
end() const { return wx_str() + length(); }
819 // lib.string.modifiers
821 wxString
& append(const wxString
& str
)
822 { *this += str
; return *this; }
823 // append elements str[pos], ..., str[pos+n]
824 wxString
& append(const wxString
& str
, size_t pos
, size_t n
)
825 { ConcatSelf(n
, str
.c_str() + pos
); return *this; }
826 // append first n (or all if n == npos) characters of sz
827 wxString
& append(const wxChar
*sz
, size_t n
= npos
)
828 { ConcatSelf(n
== npos
? wxStrlen(sz
) : n
, sz
); return *this; }
830 // append n copies of ch
831 wxString
& append(size_t n
, wxChar ch
) { return Pad(n
, ch
); }
833 // same as `this_string = str'
834 wxString
& assign(const wxString
& str
)
835 { return *this = str
; }
836 // same as ` = str[pos..pos + n]
837 wxString
& assign(const wxString
& str
, size_t pos
, size_t n
)
838 { Empty(); return Append(str
.c_str() + pos
, n
); }
839 // same as `= first n (or all if n == npos) characters of sz'
840 wxString
& assign(const wxChar
*sz
, size_t n
= npos
)
841 { Empty(); return Append(sz
, n
== npos
? wxStrlen(sz
) : n
); }
842 // same as `= n copies of ch'
843 wxString
& assign(size_t n
, wxChar ch
)
844 { Empty(); return Append(ch
, n
); }
846 // insert another string
847 wxString
& insert(size_t nPos
, const wxString
& str
);
848 // insert n chars of str starting at nStart (in str)
849 wxString
& insert(size_t nPos
, const wxString
& str
, size_t nStart
, size_t n
)
850 { return insert(nPos
, wxString((const wxChar
*)str
+ nStart
, n
)); }
852 // insert first n (or all if n == npos) characters of sz
853 wxString
& insert(size_t nPos
, const wxChar
*sz
, size_t n
= npos
)
854 { return insert(nPos
, wxString(sz
, n
)); }
855 // insert n copies of ch
856 wxString
& insert(size_t nPos
, size_t n
, wxChar ch
)
857 { return insert(nPos
, wxString(ch
, n
)); }
859 // delete characters from nStart to nStart + nLen
860 wxString
& erase(size_t nStart
= 0, size_t nLen
= npos
);
862 // replaces the substring of length nLen starting at nStart
863 wxString
& replace(size_t nStart
, size_t nLen
, const wxChar
* sz
);
864 // replaces the substring with nCount copies of ch
865 wxString
& replace(size_t nStart
, size_t nLen
, size_t nCount
, wxChar ch
);
866 // replaces a substring with another substring
867 wxString
& replace(size_t nStart
, size_t nLen
,
868 const wxString
& str
, size_t nStart2
, size_t nLen2
);
869 // replaces the substring with first nCount chars of sz
870 wxString
& replace(size_t nStart
, size_t nLen
,
871 const wxChar
* sz
, size_t nCount
);
874 void swap(wxString
& str
);
876 // All find() functions take the nStart argument which specifies the
877 // position to start the search on, the default value is 0. All functions
878 // return npos if there were no match.
881 size_t find(const wxString
& str
, size_t nStart
= 0) const;
883 // VC++ 1.5 can't cope with this syntax.
884 #if !defined(__VISUALC__) || defined(__WIN32__)
885 // find first n characters of sz
886 size_t find(const wxChar
* sz
, size_t nStart
= 0, size_t n
= npos
) const;
889 // Gives a duplicate symbol (presumably a case-insensitivity problem)
890 #if !defined(__BORLANDC__)
891 // find the first occurence of character ch after nStart
892 size_t find(wxChar ch
, size_t nStart
= 0) const;
894 // rfind() family is exactly like find() but works right to left
896 // as find, but from the end
897 size_t rfind(const wxString
& str
, size_t nStart
= npos
) const;
899 // VC++ 1.5 can't cope with this syntax.
900 #if !defined(__VISUALC__) || defined(__WIN32__)
901 // as find, but from the end
902 size_t rfind(const wxChar
* sz
, size_t nStart
= npos
,
903 size_t n
= npos
) const;
904 // as find, but from the end
905 size_t rfind(wxChar ch
, size_t nStart
= npos
) const;
908 // find first/last occurence of any character in the set
910 // as strpbrk() but starts at nStart, returns npos if not found
911 size_t find_first_of(const wxString
& str
, size_t nStart
= 0) const
912 { return find_first_of(str
.c_str(), nStart
); }
914 size_t find_first_of(const wxChar
* sz
, size_t nStart
= 0) const;
915 // same as find(char, size_t)
916 size_t find_first_of(wxChar c
, size_t nStart
= 0) const
917 { return find(c
, nStart
); }
918 // find the last (starting from nStart) char from str in this string
919 size_t find_last_of (const wxString
& str
, size_t nStart
= npos
) const
920 { return find_last_of(str
.c_str(), nStart
); }
922 size_t find_last_of (const wxChar
* sz
, size_t nStart
= npos
) const;
924 size_t find_last_of(wxChar c
, size_t nStart
= npos
) const
925 { return rfind(c
, nStart
); }
927 // find first/last occurence of any character not in the set
929 // as strspn() (starting from nStart), returns npos on failure
930 size_t find_first_not_of(const wxString
& str
, size_t nStart
= 0) const
931 { return find_first_not_of(str
.c_str(), nStart
); }
933 size_t find_first_not_of(const wxChar
* sz
, size_t nStart
= 0) const;
935 size_t find_first_not_of(wxChar ch
, size_t nStart
= 0) const;
937 size_t find_last_not_of(const wxString
& str
, size_t nStart
= npos
) const
938 { return find_first_not_of(str
.c_str(), nStart
); }
940 size_t find_last_not_of(const wxChar
* sz
, size_t nStart
= npos
) const;
942 size_t find_last_not_of(wxChar ch
, size_t nStart
= npos
) const;
944 // All compare functions return -1, 0 or 1 if the [sub]string is less,
945 // equal or greater than the compare() argument.
947 // just like strcmp()
948 int compare(const wxString
& str
) const { return Cmp(str
); }
949 // comparison with a substring
950 int compare(size_t nStart
, size_t nLen
, const wxString
& str
) const
951 { return Mid(nStart
, nLen
).Cmp(str
); }
952 // comparison of 2 substrings
953 int compare(size_t nStart
, size_t nLen
,
954 const wxString
& str
, size_t nStart2
, size_t nLen2
) const
955 { return Mid(nStart
, nLen
).Cmp(str
.Mid(nStart2
, nLen2
)); }
956 // just like strcmp()
957 int compare(const wxChar
* sz
) const { return Cmp(sz
); }
958 // substring comparison with first nCount characters of sz
959 int compare(size_t nStart
, size_t nLen
,
960 const wxChar
* sz
, size_t nCount
= npos
) const
961 { return Mid(nStart
, nLen
).Cmp(wxString(sz
, nCount
)); }
963 // substring extraction
964 wxString
substr(size_t nStart
= 0, size_t nLen
= npos
) const
965 { return Mid(nStart
, nLen
); }
966 #endif // wxSTD_STRING_COMPATIBILITY
969 // ----------------------------------------------------------------------------
970 // The string array uses it's knowledge of internal structure of the wxString
971 // class to optimize string storage. Normally, we would store pointers to
972 // string, but as wxString is, in fact, itself a pointer (sizeof(wxString) is
973 // sizeof(char *)) we store these pointers instead. The cast to "wxString *" is
974 // really all we need to turn such pointer into a string!
976 // Of course, it can be called a dirty hack, but we use twice less memory and
977 // this approach is also more speed efficient, so it's probably worth it.
979 // Usage notes: when a string is added/inserted, a new copy of it is created,
980 // so the original string may be safely deleted. When a string is retrieved
981 // from the array (operator[] or Item() method), a reference is returned.
982 // ----------------------------------------------------------------------------
984 class WXDLLEXPORT wxArrayString
987 // type of function used by wxArrayString::Sort()
988 typedef int (*CompareFunction
)(const wxString
& first
,
989 const wxString
& second
);
991 // constructors and destructor
993 wxArrayString() { Init(FALSE
); }
994 // if autoSort is TRUE, the array is always sorted (in alphabetical order)
996 // NB: the reason for using int and not bool is that like this we can avoid
997 // using this ctor for implicit conversions from "const char *" (which
998 // we'd like to be implicitly converted to wxString instead!)
1000 // of course, using explicit would be even better - if all compilers
1002 wxArrayString(int autoSort
) { Init(autoSort
!= 0); }
1004 wxArrayString(const wxArrayString
& array
);
1005 // assignment operator
1006 wxArrayString
& operator=(const wxArrayString
& src
);
1007 // not virtual, this class should not be derived from
1010 // memory management
1011 // empties the list, but doesn't release memory
1013 // empties the list and releases memory
1015 // preallocates memory for given number of items
1016 void Alloc(size_t nCount
);
1017 // minimzes the memory usage (by freeing all extra memory)
1021 // number of elements in the array
1022 size_t GetCount() const { return m_nCount
; }
1024 bool IsEmpty() const { return m_nCount
== 0; }
1025 // number of elements in the array (GetCount is preferred API)
1026 size_t Count() const { return m_nCount
; }
1028 // items access (range checking is done in debug version)
1029 // get item at position uiIndex
1030 wxString
& Item(size_t nIndex
) const
1031 { wxASSERT( nIndex
< m_nCount
); return *(wxString
*)&(m_pItems
[nIndex
]); }
1033 wxString
& operator[](size_t nIndex
) const { return Item(nIndex
); }
1035 wxString
& Last() const { wxASSERT( !IsEmpty() ); return Item(Count() - 1); }
1037 // return a wxString[], useful for the controls which
1038 // take one in their ctor. You must delete[] it yourself
1039 // once you are done with it. Will return NULL if the
1040 // ArrayString was empty.
1041 wxString
* GetStringArray() const;
1044 // Search the element in the array, starting from the beginning if
1045 // bFromEnd is FALSE or from end otherwise. If bCase, comparison is case
1046 // sensitive (default). Returns index of the first item matched or
1048 int Index (const wxChar
*sz
, bool bCase
= TRUE
, bool bFromEnd
= FALSE
) const;
1049 // add new element at the end (if the array is not sorted), return its
1051 size_t Add(const wxString
& str
);
1052 // add new element at given position
1053 void Insert(const wxString
& str
, size_t uiIndex
);
1054 // expand the array to have count elements
1055 void SetCount(size_t count
);
1056 // remove first item matching this value
1057 void Remove(const wxChar
*sz
);
1058 // remove item by index
1059 void Remove(size_t nIndex
);
1060 void RemoveAt(size_t nIndex
) { Remove(nIndex
); }
1063 // sort array elements in alphabetical order (or reversed alphabetical
1064 // order if reverseOrder parameter is TRUE)
1065 void Sort(bool reverseOrder
= FALSE
);
1066 // sort array elements using specified comparaison function
1067 void Sort(CompareFunction compareFunction
);
1070 // compare two arrays case sensitively
1071 bool operator==(const wxArrayString
& a
) const;
1072 // compare two arrays case sensitively
1073 bool operator!=(const wxArrayString
& a
) const { return !(*this == a
); }
1076 void Init(bool autoSort
); // common part of all ctors
1077 void Copy(const wxArrayString
& src
); // copies the contents of another array
1080 void Grow(); // makes array bigger if needed
1081 void Free(); // free all the strings stored
1083 void DoSort(); // common part of all Sort() variants
1085 size_t m_nSize
, // current size of the array
1086 m_nCount
; // current number of elements
1088 wxChar
**m_pItems
; // pointer to data
1090 bool m_autoSort
; // if TRUE, keep the array always sorted
1093 class WXDLLEXPORT wxSortedArrayString
: public wxArrayString
1096 wxSortedArrayString() : wxArrayString(TRUE
)
1098 wxSortedArrayString(const wxArrayString
& array
) : wxArrayString(TRUE
)
1102 // ----------------------------------------------------------------------------
1103 // wxStringBuffer: a tiny class allowing to get a writable pointer into string
1104 // ----------------------------------------------------------------------------
1106 class WXDLLEXPORT wxStringBuffer
1109 wxStringBuffer(wxString
& str
, size_t lenWanted
= 1024)
1110 : m_str(str
) { m_buf
= m_str
.GetWriteBuf(lenWanted
); }
1112 ~wxStringBuffer() { m_str
.UngetWriteBuf(); }
1114 operator wxChar
*() const { return m_buf
; }
1121 // ---------------------------------------------------------------------------
1122 // wxString comparison functions: operator versions are always case sensitive
1123 // ---------------------------------------------------------------------------
1125 inline bool operator==(const wxString
& s1
, const wxString
& s2
)
1126 { return (s1
.Len() == s2
.Len()) && (s1
.Cmp(s2
) == 0); }
1127 inline bool operator==(const wxString
& s1
, const wxChar
* s2
)
1128 { return s1
.Cmp(s2
) == 0; }
1129 inline bool operator==(const wxChar
* s1
, const wxString
& s2
)
1130 { return s2
.Cmp(s1
) == 0; }
1131 inline bool operator!=(const wxString
& s1
, const wxString
& s2
)
1132 { return (s1
.Len() != s2
.Len()) || (s1
.Cmp(s2
) != 0); }
1133 inline bool operator!=(const wxString
& s1
, const wxChar
* s2
)
1134 { return s1
.Cmp(s2
) != 0; }
1135 inline bool operator!=(const wxChar
* s1
, const wxString
& s2
)
1136 { return s2
.Cmp(s1
) != 0; }
1137 inline bool operator< (const wxString
& s1
, const wxString
& s2
)
1138 { return s1
.Cmp(s2
) < 0; }
1139 inline bool operator< (const wxString
& s1
, const wxChar
* s2
)
1140 { return s1
.Cmp(s2
) < 0; }
1141 inline bool operator< (const wxChar
* s1
, const wxString
& s2
)
1142 { return s2
.Cmp(s1
) > 0; }
1143 inline bool operator> (const wxString
& s1
, const wxString
& s2
)
1144 { return s1
.Cmp(s2
) > 0; }
1145 inline bool operator> (const wxString
& s1
, const wxChar
* s2
)
1146 { return s1
.Cmp(s2
) > 0; }
1147 inline bool operator> (const wxChar
* s1
, const wxString
& s2
)
1148 { return s2
.Cmp(s1
) < 0; }
1149 inline bool operator<=(const wxString
& s1
, const wxString
& s2
)
1150 { return s1
.Cmp(s2
) <= 0; }
1151 inline bool operator<=(const wxString
& s1
, const wxChar
* s2
)
1152 { return s1
.Cmp(s2
) <= 0; }
1153 inline bool operator<=(const wxChar
* s1
, const wxString
& s2
)
1154 { return s2
.Cmp(s1
) >= 0; }
1155 inline bool operator>=(const wxString
& s1
, const wxString
& s2
)
1156 { return s1
.Cmp(s2
) >= 0; }
1157 inline bool operator>=(const wxString
& s1
, const wxChar
* s2
)
1158 { return s1
.Cmp(s2
) >= 0; }
1159 inline bool operator>=(const wxChar
* s1
, const wxString
& s2
)
1160 { return s2
.Cmp(s1
) <= 0; }
1162 // comparison with char
1163 inline bool operator==(wxChar c
, const wxString
& s
) { return s
.IsSameAs(c
); }
1164 inline bool operator==(const wxString
& s
, wxChar c
) { return s
.IsSameAs(c
); }
1165 inline bool operator!=(wxChar c
, const wxString
& s
) { return !s
.IsSameAs(c
); }
1166 inline bool operator!=(const wxString
& s
, wxChar c
) { return !s
.IsSameAs(c
); }
1169 inline bool operator==(const wxString
& s1
, const wxWCharBuffer
& s2
)
1170 { return (s1
.Cmp((const wchar_t *)s2
) == 0); }
1171 inline bool operator==(const wxWCharBuffer
& s1
, const wxString
& s2
)
1172 { return (s2
.Cmp((const wchar_t *)s1
) == 0); }
1173 inline bool operator!=(const wxString
& s1
, const wxWCharBuffer
& s2
)
1174 { return (s1
.Cmp((const wchar_t *)s2
) != 0); }
1175 inline bool operator!=(const wxWCharBuffer
& s1
, const wxString
& s2
)
1176 { return (s2
.Cmp((const wchar_t *)s1
) != 0); }
1177 #else // !wxUSE_UNICODE
1178 inline bool operator==(const wxString
& s1
, const wxCharBuffer
& s2
)
1179 { return (s1
.Cmp((const char *)s2
) == 0); }
1180 inline bool operator==(const wxCharBuffer
& s1
, const wxString
& s2
)
1181 { return (s2
.Cmp((const char *)s1
) == 0); }
1182 inline bool operator!=(const wxString
& s1
, const wxCharBuffer
& s2
)
1183 { return (s1
.Cmp((const char *)s2
) != 0); }
1184 inline bool operator!=(const wxCharBuffer
& s1
, const wxString
& s2
)
1185 { return (s2
.Cmp((const char *)s1
) != 0); }
1186 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
1188 wxString WXDLLEXPORT
operator+(const wxString
& string1
, const wxString
& string2
);
1189 wxString WXDLLEXPORT
operator+(const wxString
& string
, wxChar ch
);
1190 wxString WXDLLEXPORT
operator+(wxChar ch
, const wxString
& string
);
1191 wxString WXDLLEXPORT
operator+(const wxString
& string
, const wxChar
*psz
);
1192 wxString WXDLLEXPORT
operator+(const wxChar
*psz
, const wxString
& string
);
1194 inline wxString
operator+(const wxString
& string
, const wxWCharBuffer
& buf
)
1195 { return string
+ (const wchar_t *)buf
; }
1196 inline wxString
operator+(const wxWCharBuffer
& buf
, const wxString
& string
)
1197 { return (const wchar_t *)buf
+ string
; }
1198 #else // !wxUSE_UNICODE
1199 inline wxString
operator+(const wxString
& string
, const wxCharBuffer
& buf
)
1200 { return string
+ (const char *)buf
; }
1201 inline wxString
operator+(const wxCharBuffer
& buf
, const wxString
& string
)
1202 { return (const char *)buf
+ string
; }
1203 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
1205 // ---------------------------------------------------------------------------
1206 // Implementation only from here until the end of file
1207 // ---------------------------------------------------------------------------
1209 // don't pollute the library user's name space
1210 #undef ASSERT_VALID_INDEX
1212 #if defined(wxSTD_STRING_COMPATIBILITY) && wxUSE_STD_IOSTREAM
1214 #include "wx/ioswrap.h"
1216 WXDLLEXPORT wxSTD istream
& operator>>(wxSTD istream
&, wxString
&);
1217 WXDLLEXPORT wxSTD ostream
& operator<<(wxSTD ostream
&, const wxString
&);
1219 #endif // wxSTD_STRING_COMPATIBILITY
1221 #endif // _WX_WXSTRINGH__