1 // std::codecvt implementation details, GNU version -*- C++ -*-
3 // Copyright (C) 2002, 2003, 2005, 2006 Free Software Foundation, Inc.
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)
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.
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, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301,
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.
31 // ISO C++ 14882: 22.2.1.5 - Template class codecvt
34 // Written by Benjamin Kosnik <bkoz@redhat.com>
37 #include <bits/c++locale_internal.h>
39 _GLIBCXX_BEGIN_NAMESPACE(std)
42 #ifdef _GLIBCXX_USE_WCHAR_T
44 codecvt<wchar_t, char, mbstate_t>::
45 do_out(state_type& __state, const intern_type* __from,
46 const intern_type* __from_end, const intern_type*& __from_next,
47 extern_type* __to, extern_type* __to_end,
48 extern_type*& __to_next) const
51 state_type __tmp_state(__state);
53 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
54 __c_locale __old = __uselocale(_M_c_locale_codecvt);
57 // wcsnrtombs is *very* fast but stops if encounters NUL characters:
58 // in case we fall back to wcrtomb and then continue, in a loop.
59 // NB: wcsnrtombs is a GNU extension
60 for (__from_next = __from, __to_next = __to;
61 __from_next < __from_end && __to_next < __to_end
64 const intern_type* __from_chunk_end = wmemchr(__from_next, L'\0',
65 __from_end - __from_next);
66 if (!__from_chunk_end)
67 __from_chunk_end = __from_end;
70 const size_t __conv = wcsnrtombs(__to_next, &__from_next,
71 __from_chunk_end - __from_next,
72 __to_end - __to_next, &__state);
73 if (__conv == static_cast<size_t>(-1))
75 // In case of error, in order to stop at the exact place we
76 // have to start again from the beginning with a series of
78 for (; __from < __from_next; ++__from)
79 __to_next += wcrtomb(__to_next, *__from, &__tmp_state);
80 __state = __tmp_state;
83 else if (__from_next && __from_next < __from_chunk_end)
90 __from_next = __from_chunk_end;
94 if (__from_next < __from_end && __ret == ok)
96 extern_type __buf[MB_LEN_MAX];
97 __tmp_state = __state;
98 const size_t __conv2 = wcrtomb(__buf, *__from_next, &__tmp_state);
99 if (__conv2 > static_cast<size_t>(__to_end - __to_next))
103 memcpy(__to_next, __buf, __conv2);
104 __state = __tmp_state;
105 __to_next += __conv2;
111 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
119 codecvt<wchar_t, char, mbstate_t>::
120 do_in(state_type& __state, const extern_type* __from,
121 const extern_type* __from_end, const extern_type*& __from_next,
122 intern_type* __to, intern_type* __to_end,
123 intern_type*& __to_next) const
126 state_type __tmp_state(__state);
128 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
129 __c_locale __old = __uselocale(_M_c_locale_codecvt);
132 // mbsnrtowcs is *very* fast but stops if encounters NUL characters:
133 // in case we store a L'\0' and then continue, in a loop.
134 // NB: mbsnrtowcs is a GNU extension
135 for (__from_next = __from, __to_next = __to;
136 __from_next < __from_end && __to_next < __to_end
139 const extern_type* __from_chunk_end;
140 __from_chunk_end = static_cast<const extern_type*>(memchr(__from_next, '\0',
143 if (!__from_chunk_end)
144 __from_chunk_end = __from_end;
146 __from = __from_next;
147 size_t __conv = mbsnrtowcs(__to_next, &__from_next,
148 __from_chunk_end - __from_next,
149 __to_end - __to_next, &__state);
150 if (__conv == static_cast<size_t>(-1))
152 // In case of error, in order to stop at the exact place we
153 // have to start again from the beginning with a series of
155 for (;; ++__to_next, __from += __conv)
157 __conv = mbrtowc(__to_next, __from, __from_end - __from,
159 if (__conv == static_cast<size_t>(-1)
160 || __conv == static_cast<size_t>(-2))
163 __from_next = __from;
164 __state = __tmp_state;
167 else if (__from_next && __from_next < __from_chunk_end)
169 // It is unclear what to return in this case (see DR 382).
175 __from_next = __from_chunk_end;
179 if (__from_next < __from_end && __ret == ok)
181 if (__to_next < __to_end)
183 // XXX Probably wrong for stateful encodings
184 __tmp_state = __state;
186 *__to_next++ = L'\0';
193 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
201 codecvt<wchar_t, char, mbstate_t>::
202 do_encoding() const throw()
204 // XXX This implementation assumes that the encoding is
205 // stateless and is either single-byte or variable-width.
207 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
208 __c_locale __old = __uselocale(_M_c_locale_codecvt);
212 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
219 codecvt<wchar_t, char, mbstate_t>::
220 do_max_length() const throw()
222 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
223 __c_locale __old = __uselocale(_M_c_locale_codecvt);
225 // XXX Probably wrong for stateful encodings.
226 int __ret = MB_CUR_MAX;
227 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
234 codecvt<wchar_t, char, mbstate_t>::
235 do_length(state_type& __state, const extern_type* __from,
236 const extern_type* __end, size_t __max) const
239 state_type __tmp_state(__state);
241 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
242 __c_locale __old = __uselocale(_M_c_locale_codecvt);
245 // mbsnrtowcs is *very* fast but stops if encounters NUL characters:
246 // in case we advance past it and then continue, in a loop.
247 // NB: mbsnrtowcs is a GNU extension
249 // A dummy internal buffer is needed in order for mbsnrtocws to consider
250 // its fourth parameter (it wouldn't with NULL as first parameter).
251 wchar_t* __to = static_cast<wchar_t*>(__builtin_alloca(sizeof(wchar_t)
253 while (__from < __end && __max)
255 const extern_type* __from_chunk_end;
256 __from_chunk_end = static_cast<const extern_type*>(memchr(__from, '\0',
259 if (!__from_chunk_end)
260 __from_chunk_end = __end;
262 const extern_type* __tmp_from = __from;
263 size_t __conv = mbsnrtowcs(__to, &__from,
264 __from_chunk_end - __from,
266 if (__conv == static_cast<size_t>(-1))
268 // In case of error, in order to stop at the exact place we
269 // have to start again from the beginning with a series of
271 for (__from = __tmp_from;; __from += __conv)
273 __conv = mbrtowc(NULL, __from, __end - __from,
275 if (__conv == static_cast<size_t>(-1)
276 || __conv == static_cast<size_t>(-2))
279 __state = __tmp_state;
280 __ret += __from - __tmp_from;
284 __from = __from_chunk_end;
286 __ret += __from - __tmp_from;
289 if (__from < __end && __max)
291 // XXX Probably wrong for stateful encodings
292 __tmp_state = __state;
299 #if __GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ > 2)
307 _GLIBCXX_END_NAMESPACE