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