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