libstdc++-v3/include/bits/codecvt.h

   1 // Locale support (codecvt) -*- C++ -*-
   2
   3 // Copyright (C) 2000, 2001, 2002, 2003, 2004 Free Software Foundation, Inc.
   4 //
   5 // This file is part of the GNU ISO C++ Library.  This library is free
   6 // software; you can redistribute it and/or modify it under the
   7 // terms of the GNU General Public License as published by the
   8 // Free Software Foundation; either version 2, or (at your option)
   9 // any later version.
  10
  11 // This library is distributed in the hope that it will be useful,
  12 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14 // GNU General Public License for more details.
  15
  16 // You should have received a copy of the GNU General Public License along
  17 // with this library; see the file COPYING.  If not, write to the Free
  18 // Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307,
  19 // USA.
  20
  21 // As a special exception, you may use this file as part of a free software
  22 // library without restriction.  Specifically, if other files instantiate
  23 // templates or use macros or inline functions from this file, or you compile
  24 // this file and link it with other files to produce an executable, this
  25 // file does not by itself cause the resulting executable to be covered by
  26 // the GNU General Public License.  This exception does not however
  27 // invalidate any other reasons why the executable file might be covered by
  28 // the GNU General Public License.
  29
  30 //
  31 // ISO C++ 14882: 22.2.1.5 Template class codecvt
  32 //
  33
  34 // Written by Benjamin Kosnik <bkoz@cygnus.com>
  35
  36 /** @file bits/codecvt.h
  37  *  This is an internal header file, included by other library headers.
  38  *  You should not attempt to use it directly.
  39  */
  40
  41 #ifndef _CODECVT_H
  42 #define _CODECVT_H 1
  43
  44 #pragma GCC system_header
  45
  46   /// @brief  Empty base class for codecvt facet [22.2.1.5].
  47   class codecvt_base
  48   {
  49   public:
  50     enum result
  51     {
  52       ok,
  53       partial,
  54       error,
  55       noconv
  56     };
  57   };
  58
  59   /**
  60    *  @brief  Common base for codecvt functions.
  61    *
  62    *  This template class provides implementations of the public functions
  63    *  that forward to the protected virtual functions.
  64    *
  65    *  This template also provides abstract stubs for the protected virtual
  66    *  functions.
  67   */
  68   template<typename _InternT, typename _ExternT, typename _StateT>
  69     class __codecvt_abstract_base
  70     : public locale::facet, public codecvt_base
  71     {
  72     public:
  73       // Types:
  74       typedef codecvt_base::result      result;
  75       typedef _InternT                  intern_type;
  76       typedef _ExternT                  extern_type;
  77       typedef _StateT                   state_type;
  78
  79       // 22.2.1.5.1 codecvt members
  80       /**
  81        *  @brief  Convert from internal to external character set.
  82        *
  83        *  Converts input string of intern_type to output string of
  84        *  extern_type.  This is analogous to wcsrtombs.  It does this by
  85        *  calling codecvt::do_out.
  86        *
  87        *  The source and destination character sets are determined by the
  88        *  facet's locale, internal and external types.
  89        *
  90        *  The characters in [from,from_end) are converted and written to
  91        *  [to,to_end).  from_next and to_next are set to point to the
  92        *  character following the last successfully converted character,
  93        *  respectively.  If the result needed no conversion, from_next and
  94        *  to_next are not affected.
  95        *
  96        *  The @a state argument should be intialized if the input is at the
  97        *  beginning and carried from a previous call if continuing
  98        *  conversion.  There are no guarantees about how @a state is used.
  99        *
 100        *  The result returned is a member of codecvt_base::result.  If
 101        *  all the input is converted, returns codecvt_base::ok.  If no
 102        *  conversion is necessary, returns codecvt_base::noconv.  If
 103        *  the input ends early or there is insufficient space in the
 104        *  output, returns codecvt_base::partial.  Otherwise the
 105        *  conversion failed and codecvt_base::error is returned.
 106        *
 107        *  @param  state  Persistent conversion state data.
 108        *  @param  from  Start of input.
 109        *  @param  from_end  End of input.
 110        *  @param  from_next  Returns start of unconverted data.
 111        *  @param  to  Start of output buffer.
 112        *  @param  to_end  End of output buffer.
 113        *  @param  to_next  Returns start of unused output area.
 114        *  @return  codecvt_base::result.
 115       */
 116       result
 117       out(state_type& __state, const intern_type* __from,
 118           const intern_type* __from_end, const intern_type*& __from_next,
 119           extern_type* __to, extern_type* __to_end,
 120           extern_type*& __to_next) const
 121       {
 122         return this->do_out(__state, __from, __from_end, __from_next,
 123                             __to, __to_end, __to_next);
 124       }
 125
 126       /**
 127        *  @brief  Reset conversion state.
 128        *
 129        *  Writes characters to output that would restore @a state to initial
 130        *  conditions.  The idea is that if a partial conversion occurs, then
 131        *  the converting the characters written by this function would leave
 132        *  the state in initial conditions, rather than partial conversion
 133        *  state.  It does this by calling codecvt::do_unshift().
 134        *
 135        *  For example, if 4 external characters always converted to 1 internal
 136        *  character, and input to in() had 6 external characters with state
 137        *  saved, this function would write two characters to the output and
 138        *  set the state to initialized conditions.
 139        *
 140        *  The source and destination character sets are determined by the
 141        *  facet's locale, internal and external types.
 142        *
 143        *  The result returned is a member of codecvt_base::result.  If the
 144        *  state could be reset and data written, returns codecvt_base::ok.  If
 145        *  no conversion is necessary, returns codecvt_base::noconv.  If the
 146        *  output has insufficient space, returns codecvt_base::partial.
 147        *  Otherwise the reset failed and codecvt_base::error is returned.
 148        *
 149        *  @param  state  Persistent conversion state data.
 150        *  @param  to  Start of output buffer.
 151        *  @param  to_end  End of output buffer.
 152        *  @param  to_next  Returns start of unused output area.
 153        *  @return  codecvt_base::result.
 154       */
 155       result
 156       unshift(state_type& __state, extern_type* __to, extern_type* __to_end,
 157               extern_type*& __to_next) const
 158       { return this->do_unshift(__state, __to,__to_end,__to_next); }
 159
 160       /**
 161        *  @brief  Convert from external to internal character set.
 162        *
 163        *  Converts input string of extern_type to output string of
 164        *  intern_type.  This is analogous to mbsrtowcs.  It does this by
 165        *  calling codecvt::do_in.
 166        *
 167        *  The source and destination character sets are determined by the
 168        *  facet's locale, internal and external types.
 169        *
 170        *  The characters in [from,from_end) are converted and written to
 171        *  [to,to_end).  from_next and to_next are set to point to the
 172        *  character following the last successfully converted character,
 173        *  respectively.  If the result needed no conversion, from_next and
 174        *  to_next are not affected.
 175        *
 176        *  The @a state argument should be intialized if the input is at the
 177        *  beginning and carried from a previous call if continuing
 178        *  conversion.  There are no guarantees about how @a state is used.
 179        *
 180        *  The result returned is a member of codecvt_base::result.  If
 181        *  all the input is converted, returns codecvt_base::ok.  If no
 182        *  conversion is necessary, returns codecvt_base::noconv.  If
 183        *  the input ends early or there is insufficient space in the
 184        *  output, returns codecvt_base::partial.  Otherwise the
 185        *  conversion failed and codecvt_base::error is returned.
 186        *
 187        *  @param  state  Persistent conversion state data.
 188        *  @param  from  Start of input.
 189        *  @param  from_end  End of input.
 190        *  @param  from_next  Returns start of unconverted data.
 191        *  @param  to  Start of output buffer.
 192        *  @param  to_end  End of output buffer.
 193        *  @param  to_next  Returns start of unused output area.
 194        *  @return  codecvt_base::result.
 195       */
 196       result
 197       in(state_type& __state, const extern_type* __from,
 198          const extern_type* __from_end, const extern_type*& __from_next,
 199          intern_type* __to, intern_type* __to_end,
 200          intern_type*& __to_next) const
 201       {
 202         return this->do_in(__state, __from, __from_end, __from_next,
 203                            __to, __to_end, __to_next);
 204       }
 205
 206       int
 207       encoding() const throw()
 208       { return this->do_encoding(); }
 209
 210       bool
 211       always_noconv() const throw()
 212       { return this->do_always_noconv(); }
 213
 214       int
 215       length(state_type& __state, const extern_type* __from,
 216              const extern_type* __end, size_t __max) const
 217       { return this->do_length(__state, __from, __end, __max); }
 218
 219       int
 220       max_length() const throw()
 221       { return this->do_max_length(); }
 222
 223     protected:
 224       explicit
 225       __codecvt_abstract_base(size_t __refs = 0) : locale::facet(__refs) { }
 226
 227       virtual
 228       ~__codecvt_abstract_base() { }
 229
 230       /**
 231        *  @brief  Convert from internal to external character set.
 232        *
 233        *  Converts input string of intern_type to output string of
 234        *  extern_type.  This function is a hook for derived classes to change
 235        *  the value returned.  @see out for more information.
 236       */
 237       virtual result
 238       do_out(state_type& __state, const intern_type* __from,
 239              const intern_type* __from_end, const intern_type*& __from_next,
 240              extern_type* __to, extern_type* __to_end,
 241              extern_type*& __to_next) const = 0;
 242
 243       virtual result
 244       do_unshift(state_type& __state, extern_type* __to,
 245                  extern_type* __to_end, extern_type*& __to_next) const = 0;
 246
 247       virtual result
 248       do_in(state_type& __state, const extern_type* __from,
 249             const extern_type* __from_end, const extern_type*& __from_next,
 250             intern_type* __to, intern_type* __to_end,
 251             intern_type*& __to_next) const = 0;
 252
 253       virtual int
 254       do_encoding() const throw() = 0;
 255
 256       virtual bool
 257       do_always_noconv() const throw() = 0;
 258
 259       virtual int
 260       do_length(state_type&, const extern_type* __from,
 261                 const extern_type* __end, size_t __max) const = 0;
 262
 263       virtual int
 264       do_max_length() const throw() = 0;
 265     };
 266
 267   /// @brief class codecvt [22.2.1.5].
 268   /// NB: Generic, mostly useless implementation.
 269   template<typename _InternT, typename _ExternT, typename _StateT>
 270     class codecvt
 271     : public __codecvt_abstract_base<_InternT, _ExternT, _StateT>
 272     {
 273     public:
 274       // Types:
 275       typedef codecvt_base::result      result;
 276       typedef _InternT                  intern_type;
 277       typedef _ExternT                  extern_type;
 278       typedef _StateT                   state_type;
 279
 280     protected:
 281       __c_locale                        _M_c_locale_codecvt;
 282
 283     public:
 284       static locale::id                 id;
 285
 286       explicit
 287       codecvt(size_t __refs = 0)
 288       : __codecvt_abstract_base<_InternT, _ExternT, _StateT> (__refs) { }
 289
 290       explicit
 291       codecvt(__c_locale __cloc, size_t __refs = 0);
 292
 293     protected:
 294       virtual
 295       ~codecvt() { }
 296
 297       virtual result
 298       do_out(state_type& __state, const intern_type* __from,
 299              const intern_type* __from_end, const intern_type*& __from_next,
 300              extern_type* __to, extern_type* __to_end,
 301              extern_type*& __to_next) const;
 302
 303       virtual result
 304       do_unshift(state_type& __state, extern_type* __to,
 305                  extern_type* __to_end, extern_type*& __to_next) const;
 306
 307       virtual result
 308       do_in(state_type& __state, const extern_type* __from,
 309             const extern_type* __from_end, const extern_type*& __from_next,
 310             intern_type* __to, intern_type* __to_end,
 311             intern_type*& __to_next) const;
 312
 313       virtual int
 314       do_encoding() const throw();
 315
 316       virtual bool
 317       do_always_noconv() const throw();
 318
 319       virtual int
 320       do_length(state_type&, const extern_type* __from,
 321                 const extern_type* __end, size_t __max) const;
 322
 323       virtual int
 324       do_max_length() const throw();
 325     };
 326
 327   template<typename _InternT, typename _ExternT, typename _StateT>
 328     locale::id codecvt<_InternT, _ExternT, _StateT>::id;
 329
 330   /// @brief class codecvt<char, char, mbstate_t> specialization.
 331   template<>
 332     class codecvt<char, char, mbstate_t>
 333     : public __codecvt_abstract_base<char, char, mbstate_t>
 334     {
 335     public:
 336       // Types:
 337       typedef char                      intern_type;
 338       typedef char                      extern_type;
 339       typedef mbstate_t                 state_type;
 340
 341     protected:
 342       __c_locale                        _M_c_locale_codecvt;
 343
 344     public:
 345       static locale::id id;
 346
 347       explicit
 348       codecvt(size_t __refs = 0);
 349
 350       explicit
 351       codecvt(__c_locale __cloc, size_t __refs = 0);
 352
 353     protected:
 354       virtual
 355       ~codecvt();
 356
 357       virtual result
 358       do_out(state_type& __state, const intern_type* __from,
 359              const intern_type* __from_end, const intern_type*& __from_next,
 360              extern_type* __to, extern_type* __to_end,
 361              extern_type*& __to_next) const;
 362
 363       virtual result
 364       do_unshift(state_type& __state, extern_type* __to,
 365                  extern_type* __to_end, extern_type*& __to_next) const;
 366
 367       virtual result
 368       do_in(state_type& __state, const extern_type* __from,
 369             const extern_type* __from_end, const extern_type*& __from_next,
 370             intern_type* __to, intern_type* __to_end,
 371             intern_type*& __to_next) const;
 372
 373       virtual int
 374       do_encoding() const throw();
 375
 376       virtual bool
 377       do_always_noconv() const throw();
 378
 379       virtual int
 380       do_length(state_type&, const extern_type* __from,
 381                 const extern_type* __end, size_t __max) const;
 382
 383       virtual int
 384       do_max_length() const throw();
 385   };
 386
 387 #ifdef _GLIBCXX_USE_WCHAR_T
 388   /// @brief  class codecvt<wchar_t, char, mbstate_t> specialization.
 389   template<>
 390     class codecvt<wchar_t, char, mbstate_t>
 391     : public __codecvt_abstract_base<wchar_t, char, mbstate_t>
 392     {
 393     public:
 394       // Types:
 395       typedef wchar_t                   intern_type;
 396       typedef char                      extern_type;
 397       typedef mbstate_t                 state_type;
 398
 399     protected:
 400       __c_locale                        _M_c_locale_codecvt;
 401
 402     public:
 403       static locale::id                 id;
 404
 405       explicit
 406       codecvt(size_t __refs = 0);
 407
 408       explicit
 409       codecvt(__c_locale __cloc, size_t __refs = 0);
 410
 411     protected:
 412       virtual
 413       ~codecvt();
 414
 415       virtual result
 416       do_out(state_type& __state, const intern_type* __from,
 417              const intern_type* __from_end, const intern_type*& __from_next,
 418              extern_type* __to, extern_type* __to_end,
 419              extern_type*& __to_next) const;
 420
 421       virtual result
 422       do_unshift(state_type& __state,
 423                  extern_type* __to, extern_type* __to_end,
 424                  extern_type*& __to_next) const;
 425
 426       virtual result
 427       do_in(state_type& __state,
 428              const extern_type* __from, const extern_type* __from_end,
 429              const extern_type*& __from_next,
 430              intern_type* __to, intern_type* __to_end,
 431              intern_type*& __to_next) const;
 432
 433       virtual
 434       int do_encoding() const throw();
 435
 436       virtual
 437       bool do_always_noconv() const throw();
 438
 439       virtual
 440       int do_length(state_type&, const extern_type* __from,
 441                     const extern_type* __end, size_t __max) const;
 442
 443       virtual int
 444       do_max_length() const throw();
 445     };
 446 #endif //_GLIBCXX_USE_WCHAR_T
 447
 448   /// @brief class codecvt_byname [22.2.1.6].
 449   template<typename _InternT, typename _ExternT, typename _StateT>
 450     class codecvt_byname : public codecvt<_InternT, _ExternT, _StateT>
 451     {
 452     public:
 453       explicit
 454       codecvt_byname(const char* __s, size_t __refs = 0)
 455       : codecvt<_InternT, _ExternT, _StateT>(__refs)
 456       {
 457         if (std::strcmp(__s, "C") != 0 && std::strcmp(__s, "POSIX") != 0)
 458           {
 459             this->_S_destroy_c_locale(this->_M_c_locale_codecvt);
 460             this->_S_create_c_locale(this->_M_c_locale_codecvt, __s);
 461           }
 462       }
 463
 464     protected:
 465       virtual
 466       ~codecvt_byname() { }
 467     };
 468
 469   // Include host and configuration specific partial specializations
 470   // with additional functionality, if possible.
 471 #ifdef _GLIBCXX_USE_WCHAR_T
 472   #include <bits/codecvt_specializations.h>
 473 #endif
 474
 475 #endif // _CODECVT_H