codecvt.h

Go to the documentation of this file.
00001 // Locale support (codecvt) -*- C++ -*-
00002 
00003 // Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008,
00004 // 2009  Free Software Foundation, Inc.
00005 //
00006 // This file is part of the GNU ISO C++ Library.  This library is free
00007 // software; you can redistribute it and/or modify it under the
00008 // terms of the GNU General Public License as published by the
00009 // Free Software Foundation; either version 3, or (at your option)
00010 // any later version.
00011 
00012 // This library is distributed in the hope that it will be useful,
00013 // but WITHOUT ANY WARRANTY; without even the implied warranty of
00014 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
00015 // GNU General Public License for more details.
00016 
00017 // Under Section 7 of GPL version 3, you are granted additional
00018 // permissions described in the GCC Runtime Library Exception, version
00019 // 3.1, as published by the Free Software Foundation.
00020 
00021 // You should have received a copy of the GNU General Public License and
00022 // a copy of the GCC Runtime Library Exception along with this program;
00023 // see the files COPYING3 and COPYING.RUNTIME respectively.  If not, see
00024 // <http://www.gnu.org/licenses/>.
00025 
00026 /** @file bits/codecvt.h
00027  *  This is an internal header file, included by other library headers.
00028  *  You should not attempt to use it directly.
00029  */
00030 
00031 //
00032 // ISO C++ 14882: 22.2.1.5 Template class codecvt
00033 //
00034 
00035 // Written by Benjamin Kosnik <bkoz@redhat.com>
00036 
00037 #ifndef _CODECVT_H
00038 #define _CODECVT_H 1
00039 
00040 #pragma GCC system_header
00041 
00042 _GLIBCXX_BEGIN_NAMESPACE(std)
00043 
00044   /// Empty base class for codecvt facet [22.2.1.5].
00045   class codecvt_base
00046   {
00047   public:
00048     enum result
00049     {
00050       ok,
00051       partial,
00052       error,
00053       noconv
00054     };
00055   };
00056 
00057   /**
00058    *  @brief  Common base for codecvt functions.
00059    *
00060    *  This template class provides implementations of the public functions
00061    *  that forward to the protected virtual functions.
00062    *
00063    *  This template also provides abstract stubs for the protected virtual
00064    *  functions.
00065   */
00066   template<typename _InternT, typename _ExternT, typename _StateT>
00067     class __codecvt_abstract_base
00068     : public locale::facet, public codecvt_base
00069     {
00070     public:
00071       // Types:
00072       typedef codecvt_base::result  result;
00073       typedef _InternT          intern_type;
00074       typedef _ExternT          extern_type;
00075       typedef _StateT           state_type;
00076 
00077       // 22.2.1.5.1 codecvt members
00078       /**
00079        *  @brief  Convert from internal to external character set.
00080        *
00081        *  Converts input string of intern_type to output string of
00082        *  extern_type.  This is analogous to wcsrtombs.  It does this by
00083        *  calling codecvt::do_out.
00084        *
00085        *  The source and destination character sets are determined by the
00086        *  facet's locale, internal and external types.
00087        *
00088        *  The characters in [from,from_end) are converted and written to
00089        *  [to,to_end).  from_next and to_next are set to point to the
00090        *  character following the last successfully converted character,
00091        *  respectively.  If the result needed no conversion, from_next and
00092        *  to_next are not affected.
00093        *
00094        *  The @a state argument should be initialized if the input is at the
00095        *  beginning and carried from a previous call if continuing
00096        *  conversion.  There are no guarantees about how @a state is used.
00097        *
00098        *  The result returned is a member of codecvt_base::result.  If
00099        *  all the input is converted, returns codecvt_base::ok.  If no
00100        *  conversion is necessary, returns codecvt_base::noconv.  If
00101        *  the input ends early or there is insufficient space in the
00102        *  output, returns codecvt_base::partial.  Otherwise the
00103        *  conversion failed and codecvt_base::error is returned.
00104        *
00105        *  @param  state  Persistent conversion state data.
00106        *  @param  from  Start of input.
00107        *  @param  from_end  End of input.
00108        *  @param  from_next  Returns start of unconverted data.
00109        *  @param  to  Start of output buffer.
00110        *  @param  to_end  End of output buffer.
00111        *  @param  to_next  Returns start of unused output area.
00112        *  @return  codecvt_base::result.
00113       */
00114       result
00115       out(state_type& __state, const intern_type* __from,
00116       const intern_type* __from_end, const intern_type*& __from_next,
00117       extern_type* __to, extern_type* __to_end,
00118       extern_type*& __to_next) const
00119       {
00120     return this->do_out(__state, __from, __from_end, __from_next,
00121                 __to, __to_end, __to_next);
00122       }
00123 
00124       /**
00125        *  @brief  Reset conversion state.
00126        *
00127        *  Writes characters to output that would restore @a state to initial
00128        *  conditions.  The idea is that if a partial conversion occurs, then
00129        *  the converting the characters written by this function would leave
00130        *  the state in initial conditions, rather than partial conversion
00131        *  state.  It does this by calling codecvt::do_unshift().
00132        *
00133        *  For example, if 4 external characters always converted to 1 internal
00134        *  character, and input to in() had 6 external characters with state
00135        *  saved, this function would write two characters to the output and
00136        *  set the state to initialized conditions.
00137        *
00138        *  The source and destination character sets are determined by the
00139        *  facet's locale, internal and external types.
00140        *
00141        *  The result returned is a member of codecvt_base::result.  If the
00142        *  state could be reset and data written, returns codecvt_base::ok.  If
00143        *  no conversion is necessary, returns codecvt_base::noconv.  If the
00144        *  output has insufficient space, returns codecvt_base::partial.
00145        *  Otherwise the reset failed and codecvt_base::error is returned.
00146        *
00147        *  @param  state  Persistent conversion state data.
00148        *  @param  to  Start of output buffer.
00149        *  @param  to_end  End of output buffer.
00150        *  @param  to_next  Returns start of unused output area.
00151        *  @return  codecvt_base::result.
00152       */
00153       result
00154       unshift(state_type& __state, extern_type* __to, extern_type* __to_end,
00155           extern_type*& __to_next) const
00156       { return this->do_unshift(__state, __to,__to_end,__to_next); }
00157 
00158       /**
00159        *  @brief  Convert from external to internal character set.
00160        *
00161        *  Converts input string of extern_type to output string of
00162        *  intern_type.  This is analogous to mbsrtowcs.  It does this by
00163        *  calling codecvt::do_in.
00164        *
00165        *  The source and destination character sets are determined by the
00166        *  facet's locale, internal and external types.
00167        *
00168        *  The characters in [from,from_end) are converted and written to
00169        *  [to,to_end).  from_next and to_next are set to point to the
00170        *  character following the last successfully converted character,
00171        *  respectively.  If the result needed no conversion, from_next and
00172        *  to_next are not affected.
00173        *
00174        *  The @a state argument should be initialized if the input is at the
00175        *  beginning and carried from a previous call if continuing
00176        *  conversion.  There are no guarantees about how @a state is used.
00177        *
00178        *  The result returned is a member of codecvt_base::result.  If
00179        *  all the input is converted, returns codecvt_base::ok.  If no
00180        *  conversion is necessary, returns codecvt_base::noconv.  If
00181        *  the input ends early or there is insufficient space in the
00182        *  output, returns codecvt_base::partial.  Otherwise the
00183        *  conversion failed and codecvt_base::error is returned.
00184        *
00185        *  @param  state  Persistent conversion state data.
00186        *  @param  from  Start of input.
00187        *  @param  from_end  End of input.
00188        *  @param  from_next  Returns start of unconverted data.
00189        *  @param  to  Start of output buffer.
00190        *  @param  to_end  End of output buffer.
00191        *  @param  to_next  Returns start of unused output area.
00192        *  @return  codecvt_base::result.
00193       */
00194       result
00195       in(state_type& __state, const extern_type* __from,
00196      const extern_type* __from_end, const extern_type*& __from_next,
00197      intern_type* __to, intern_type* __to_end,
00198      intern_type*& __to_next) const
00199       {
00200     return this->do_in(__state, __from, __from_end, __from_next,
00201                __to, __to_end, __to_next);
00202       }
00203 
00204       int
00205       encoding() const throw()
00206       { return this->do_encoding(); }
00207 
00208       bool
00209       always_noconv() const throw()
00210       { return this->do_always_noconv(); }
00211 
00212       int
00213       length(state_type& __state, const extern_type* __from,
00214          const extern_type* __end, size_t __max) const
00215       { return this->do_length(__state, __from, __end, __max); }
00216 
00217       int
00218       max_length() const throw()
00219       { return this->do_max_length(); }
00220 
00221     protected:
00222       explicit
00223       __codecvt_abstract_base(size_t __refs = 0) : locale::facet(__refs) { }
00224 
00225       virtual
00226       ~__codecvt_abstract_base() { }
00227 
00228       /**
00229        *  @brief  Convert from internal to external character set.
00230        *
00231        *  Converts input string of intern_type to output string of
00232        *  extern_type.  This function is a hook for derived classes to change
00233        *  the value returned.  @see out for more information.
00234       */
00235       virtual result
00236       do_out(state_type& __state, const intern_type* __from,
00237          const intern_type* __from_end, const intern_type*& __from_next,
00238          extern_type* __to, extern_type* __to_end,
00239          extern_type*& __to_next) const = 0;
00240 
00241       virtual result
00242       do_unshift(state_type& __state, extern_type* __to,
00243          extern_type* __to_end, extern_type*& __to_next) const = 0;
00244 
00245       virtual result
00246       do_in(state_type& __state, const extern_type* __from,
00247         const extern_type* __from_end, const extern_type*& __from_next,
00248         intern_type* __to, intern_type* __to_end,
00249         intern_type*& __to_next) const = 0;
00250 
00251       virtual int
00252       do_encoding() const throw() = 0;
00253 
00254       virtual bool
00255       do_always_noconv() const throw() = 0;
00256 
00257       virtual int
00258       do_length(state_type&, const extern_type* __from,
00259         const extern_type* __end, size_t __max) const = 0;
00260 
00261       virtual int
00262       do_max_length() const throw() = 0;
00263     };
00264 
00265   /// @brief class codecvt [22.2.1.5].
00266   /// NB: Generic, mostly useless implementation.
00267   template<typename _InternT, typename _ExternT, typename _StateT>
00268     class codecvt
00269     : public __codecvt_abstract_base<_InternT, _ExternT, _StateT>
00270     {
00271     public:
00272       // Types:
00273       typedef codecvt_base::result  result;
00274       typedef _InternT          intern_type;
00275       typedef _ExternT          extern_type;
00276       typedef _StateT           state_type;
00277 
00278     protected:
00279       __c_locale            _M_c_locale_codecvt;
00280 
00281     public:
00282       static locale::id         id;
00283 
00284       explicit
00285       codecvt(size_t __refs = 0)
00286       : __codecvt_abstract_base<_InternT, _ExternT, _StateT> (__refs) { }
00287 
00288       explicit
00289       codecvt(__c_locale __cloc, size_t __refs = 0);
00290 
00291     protected:
00292       virtual
00293       ~codecvt() { }
00294 
00295       virtual result
00296       do_out(state_type& __state, const intern_type* __from,
00297          const intern_type* __from_end, const intern_type*& __from_next,
00298          extern_type* __to, extern_type* __to_end,
00299          extern_type*& __to_next) const;
00300 
00301       virtual result
00302       do_unshift(state_type& __state, extern_type* __to,
00303          extern_type* __to_end, extern_type*& __to_next) const;
00304 
00305       virtual result
00306       do_in(state_type& __state, const extern_type* __from,
00307         const extern_type* __from_end, const extern_type*& __from_next,
00308         intern_type* __to, intern_type* __to_end,
00309         intern_type*& __to_next) const;
00310 
00311       virtual int
00312       do_encoding() const throw();
00313 
00314       virtual bool
00315       do_always_noconv() const throw();
00316 
00317       virtual int
00318       do_length(state_type&, const extern_type* __from,
00319         const extern_type* __end, size_t __max) const;
00320 
00321       virtual int
00322       do_max_length() const throw();
00323     };
00324 
00325   template<typename _InternT, typename _ExternT, typename _StateT>
00326     locale::id codecvt<_InternT, _ExternT, _StateT>::id;
00327 
00328   /// class codecvt<char, char, mbstate_t> specialization.
00329   template<>
00330     class codecvt<char, char, mbstate_t>
00331     : public __codecvt_abstract_base<char, char, mbstate_t>
00332     {
00333     public:
00334       // Types:
00335       typedef char          intern_type;
00336       typedef char          extern_type;
00337       typedef mbstate_t         state_type;
00338 
00339     protected:
00340       __c_locale            _M_c_locale_codecvt;
00341 
00342     public:
00343       static locale::id id;
00344 
00345       explicit
00346       codecvt(size_t __refs = 0);
00347 
00348       explicit
00349       codecvt(__c_locale __cloc, size_t __refs = 0);
00350 
00351     protected:
00352       virtual
00353       ~codecvt();
00354 
00355       virtual result
00356       do_out(state_type& __state, const intern_type* __from,
00357          const intern_type* __from_end, const intern_type*& __from_next,
00358          extern_type* __to, extern_type* __to_end,
00359          extern_type*& __to_next) const;
00360 
00361       virtual result
00362       do_unshift(state_type& __state, extern_type* __to,
00363          extern_type* __to_end, extern_type*& __to_next) const;
00364 
00365       virtual result
00366       do_in(state_type& __state, const extern_type* __from,
00367         const extern_type* __from_end, const extern_type*& __from_next,
00368         intern_type* __to, intern_type* __to_end,
00369         intern_type*& __to_next) const;
00370 
00371       virtual int
00372       do_encoding() const throw();
00373 
00374       virtual bool
00375       do_always_noconv() const throw();
00376 
00377       virtual int
00378       do_length(state_type&, const extern_type* __from,
00379         const extern_type* __end, size_t __max) const;
00380 
00381       virtual int
00382       do_max_length() const throw();
00383   };
00384 
00385 #ifdef _GLIBCXX_USE_WCHAR_T
00386   /// class codecvt<wchar_t, char, mbstate_t> specialization.
00387   template<>
00388     class codecvt<wchar_t, char, mbstate_t>
00389     : public __codecvt_abstract_base<wchar_t, char, mbstate_t>
00390     {
00391     public:
00392       // Types:
00393       typedef wchar_t           intern_type;
00394       typedef char          extern_type;
00395       typedef mbstate_t         state_type;
00396 
00397     protected:
00398       __c_locale            _M_c_locale_codecvt;
00399 
00400     public:
00401       static locale::id         id;
00402 
00403       explicit
00404       codecvt(size_t __refs = 0);
00405 
00406       explicit
00407       codecvt(__c_locale __cloc, size_t __refs = 0);
00408 
00409     protected:
00410       virtual
00411       ~codecvt();
00412 
00413       virtual result
00414       do_out(state_type& __state, const intern_type* __from,
00415          const intern_type* __from_end, const intern_type*& __from_next,
00416          extern_type* __to, extern_type* __to_end,
00417          extern_type*& __to_next) const;
00418 
00419       virtual result
00420       do_unshift(state_type& __state,
00421          extern_type* __to, extern_type* __to_end,
00422          extern_type*& __to_next) const;
00423 
00424       virtual result
00425       do_in(state_type& __state,
00426          const extern_type* __from, const extern_type* __from_end,
00427          const extern_type*& __from_next,
00428          intern_type* __to, intern_type* __to_end,
00429          intern_type*& __to_next) const;
00430 
00431       virtual
00432       int do_encoding() const throw();
00433 
00434       virtual
00435       bool do_always_noconv() const throw();
00436 
00437       virtual
00438       int do_length(state_type&, const extern_type* __from,
00439             const extern_type* __end, size_t __max) const;
00440 
00441       virtual int
00442       do_max_length() const throw();
00443     };
00444 #endif //_GLIBCXX_USE_WCHAR_T
00445 
00446   /// class codecvt_byname [22.2.1.6].
00447   template<typename _InternT, typename _ExternT, typename _StateT>
00448     class codecvt_byname : public codecvt<_InternT, _ExternT, _StateT>
00449     {
00450     public:
00451       explicit
00452       codecvt_byname(const char* __s, size_t __refs = 0)
00453       : codecvt<_InternT, _ExternT, _StateT>(__refs)
00454       {
00455     if (__builtin_strcmp(__s, "C") != 0
00456         && __builtin_strcmp(__s, "POSIX") != 0)
00457       {
00458         this->_S_destroy_c_locale(this->_M_c_locale_codecvt);
00459         this->_S_create_c_locale(this->_M_c_locale_codecvt, __s);
00460       }
00461       }
00462 
00463     protected:
00464       virtual
00465       ~codecvt_byname() { }
00466     };
00467 
00468   // Inhibit implicit instantiations for required instantiations,
00469   // which are defined via explicit instantiations elsewhere.
00470   // NB: This syntax is a GNU extension.
00471 #if _GLIBCXX_EXTERN_TEMPLATE
00472   extern template class codecvt_byname<char, char, mbstate_t>;
00473 
00474   extern template
00475     const codecvt<char, char, mbstate_t>&
00476     use_facet<codecvt<char, char, mbstate_t> >(const locale&);
00477 
00478   extern template
00479     bool
00480     has_facet<codecvt<char, char, mbstate_t> >(const locale&);
00481 
00482 #ifdef _GLIBCXX_USE_WCHAR_T
00483   extern template class codecvt_byname<wchar_t, char, mbstate_t>;
00484 
00485   extern template
00486     const codecvt<wchar_t, char, mbstate_t>&
00487     use_facet<codecvt<wchar_t, char, mbstate_t> >(const locale&);
00488 
00489   extern template
00490     bool
00491     has_facet<codecvt<wchar_t, char, mbstate_t> >(const locale&);
00492 #endif
00493 #endif
00494 
00495 _GLIBCXX_END_NAMESPACE
00496 
00497 #endif // _CODECVT_H

Generated on Thu Jul 23 21:16:02 2009 for libstdc++ by  doxygen 1.5.8