libio/iofwide.c

   1 /* Copyright (C) 1999-2017 Free Software Foundation, Inc.
   2    This file is part of the GNU C Library.
   3
   4    The GNU C Library is free software; you can redistribute it and/or
   5    modify it under the terms of the GNU Lesser General Public
   6    License as published by the Free Software Foundation; either
   7    version 2.1 of the License, or (at your option) any later version.
   8
   9    The GNU C Library is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  12    Lesser General Public License for more details.
  13
  14    You should have received a copy of the GNU Lesser General Public
  15    License along with the GNU C Library; if not, see
  16    <http://www.gnu.org/licenses/>.
  17
  18    As a special exception, if you link the code in this file with
  19    files compiled with a GNU compiler to produce an executable,
  20    that does not cause the resulting executable to be covered by
  21    the GNU Lesser General Public License.  This exception does not
  22    however invalidate any other reasons why the executable file
  23    might be covered by the GNU Lesser General Public License.
  24    This exception applies to code released by its copyright holders
  25    in files containing the exception.  */
  26
  27 #include <libioP.h>
  28 #include <dlfcn.h>
  29 #include <wchar.h>
  30 #include <assert.h>
  31 #include <stdlib.h>
  32 #include <string.h>
  33
  34 #include <langinfo.h>
  35 #include <locale/localeinfo.h>
  36 #include <wcsmbs/wcsmbsload.h>
  37 #include <iconv/gconv_int.h>
  38 #include <shlib-compat.h>
  39 #include <sysdep.h>
  40
  41
  42 /* Prototypes of libio's codecvt functions.  */
  43 static enum __codecvt_result do_out (struct _IO_codecvt *codecvt,
  44                                      __mbstate_t *statep,
  45                                      const wchar_t *from_start,
  46                                      const wchar_t *from_end,
  47                                      const wchar_t **from_stop, char *to_start,
  48                                      char *to_end, char **to_stop);
  49 static enum __codecvt_result do_unshift (struct _IO_codecvt *codecvt,
  50                                          __mbstate_t *statep, char *to_start,
  51                                          char *to_end, char **to_stop);
  52 static enum __codecvt_result do_in (struct _IO_codecvt *codecvt,
  53                                     __mbstate_t *statep,
  54                                     const char *from_start,
  55                                     const char *from_end,
  56                                     const char **from_stop, wchar_t *to_start,
  57                                     wchar_t *to_end, wchar_t **to_stop);
  58 static int do_encoding (struct _IO_codecvt *codecvt);
  59 static int do_length (struct _IO_codecvt *codecvt, __mbstate_t *statep,
  60                       const char *from_start,
  61                       const char *from_end, _IO_size_t max);
  62 static int do_max_length (struct _IO_codecvt *codecvt);
  63 static int do_always_noconv (struct _IO_codecvt *codecvt);
  64
  65
  66 /* The functions used in `codecvt' for libio are always the same.  */
  67 const struct _IO_codecvt __libio_codecvt =
  68 {
  69   .__codecvt_destr = NULL,              /* Destructor, never used.  */
  70   .__codecvt_do_out = do_out,
  71   .__codecvt_do_unshift = do_unshift,
  72   .__codecvt_do_in = do_in,
  73   .__codecvt_do_encoding = do_encoding,
  74   .__codecvt_do_always_noconv = do_always_noconv,
  75   .__codecvt_do_length = do_length,
  76   .__codecvt_do_max_length = do_max_length
  77 };
  78
  79
  80 /* Return orientation of stream.  If mode is nonzero try to change
  81    the orientation first.  */
  82 #undef _IO_fwide
  83 int
  84 _IO_fwide (_IO_FILE *fp, int mode)
  85 {
  86   /* Normalize the value.  */
  87   mode = mode < 0 ? -1 : (mode == 0 ? 0 : 1);
  88
  89 #if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_1)
  90   if (__builtin_expect (&_IO_stdin_used == NULL, 0)
  91       && (fp == _IO_stdin || fp == _IO_stdout || fp == _IO_stderr))
  92     /* This is for a stream in the glibc 2.0 format.  */
  93     return -1;
  94 #endif
  95
  96   /* The orientation already has been determined.  */
  97   if (fp->_mode != 0
  98       /* Or the caller simply wants to know about the current orientation.  */
  99       || mode == 0)
 100     return fp->_mode;
 101
 102   /* Set the orientation appropriately.  */
 103   if (mode > 0)
 104     {
 105       struct _IO_codecvt *cc = fp->_codecvt = &fp->_wide_data->_codecvt;
 106
 107       fp->_wide_data->_IO_read_ptr = fp->_wide_data->_IO_read_end;
 108       fp->_wide_data->_IO_write_ptr = fp->_wide_data->_IO_write_base;
 109
 110       /* Get the character conversion functions based on the currently
 111          selected locale for LC_CTYPE.  */
 112       {
 113         /* Clear the state.  We start all over again.  */
 114         memset (&fp->_wide_data->_IO_state, '\0', sizeof (__mbstate_t));
 115         memset (&fp->_wide_data->_IO_last_state, '\0', sizeof (__mbstate_t));
 116
 117         struct gconv_fcts fcts;
 118         __wcsmbs_clone_conv (&fcts);
 119         assert (fcts.towc_nsteps == 1);
 120         assert (fcts.tomb_nsteps == 1);
 121
 122         /* The functions are always the same.  */
 123         *cc = __libio_codecvt;
 124
 125         cc->__cd_in.__cd.__nsteps = fcts.towc_nsteps;
 126         cc->__cd_in.__cd.__steps = fcts.towc;
 127
 128         cc->__cd_in.__cd.__data[0].__invocation_counter = 0;
 129         cc->__cd_in.__cd.__data[0].__internal_use = 1;
 130         cc->__cd_in.__cd.__data[0].__flags = __GCONV_IS_LAST;
 131         cc->__cd_in.__cd.__data[0].__statep = &fp->_wide_data->_IO_state;
 132
 133         cc->__cd_out.__cd.__nsteps = fcts.tomb_nsteps;
 134         cc->__cd_out.__cd.__steps = fcts.tomb;
 135
 136         cc->__cd_out.__cd.__data[0].__invocation_counter = 0;
 137         cc->__cd_out.__cd.__data[0].__internal_use = 1;
 138         cc->__cd_out.__cd.__data[0].__flags
 139           = __GCONV_IS_LAST | __GCONV_TRANSLIT;
 140         cc->__cd_out.__cd.__data[0].__statep = &fp->_wide_data->_IO_state;
 141       }
 142
 143       /* From now on use the wide character callback functions.  */
 144       _IO_JUMPS_FILE_plus (fp) = fp->_wide_data->_wide_vtable;
 145     }
 146
 147   /* Set the mode now.  */
 148   fp->_mode = mode;
 149
 150   return mode;
 151 }
 152
 153
 154 static enum __codecvt_result
 155 do_out (struct _IO_codecvt *codecvt, __mbstate_t *statep,
 156         const wchar_t *from_start, const wchar_t *from_end,
 157         const wchar_t **from_stop, char *to_start, char *to_end,
 158         char **to_stop)
 159 {
 160   enum __codecvt_result result;
 161
 162   struct __gconv_step *gs = codecvt->__cd_out.__cd.__steps;
 163   int status;
 164   size_t dummy;
 165   const unsigned char *from_start_copy = (unsigned char *) from_start;
 166
 167   codecvt->__cd_out.__cd.__data[0].__outbuf = (unsigned char *) to_start;
 168   codecvt->__cd_out.__cd.__data[0].__outbufend = (unsigned char *) to_end;
 169   codecvt->__cd_out.__cd.__data[0].__statep = statep;
 170
 171   __gconv_fct fct = gs->__fct;
 172 #ifdef PTR_DEMANGLE
 173   if (gs->__shlib_handle != NULL)
 174     PTR_DEMANGLE (fct);
 175 #endif
 176
 177   status = DL_CALL_FCT (fct,
 178                         (gs, codecvt->__cd_out.__cd.__data, &from_start_copy,
 179                          (const unsigned char *) from_end, NULL,
 180                          &dummy, 0, 0));
 181
 182   *from_stop = (wchar_t *) from_start_copy;
 183   *to_stop = (char *) codecvt->__cd_out.__cd.__data[0].__outbuf;
 184
 185   switch (status)
 186     {
 187     case __GCONV_OK:
 188     case __GCONV_EMPTY_INPUT:
 189       result = __codecvt_ok;
 190       break;
 191
 192     case __GCONV_FULL_OUTPUT:
 193     case __GCONV_INCOMPLETE_INPUT:
 194       result = __codecvt_partial;
 195       break;
 196
 197     default:
 198       result = __codecvt_error;
 199       break;
 200     }
 201
 202   return result;
 203 }
 204
 205
 206 static enum __codecvt_result
 207 do_unshift (struct _IO_codecvt *codecvt, __mbstate_t *statep,
 208             char *to_start, char *to_end, char **to_stop)
 209 {
 210   enum __codecvt_result result;
 211
 212   struct __gconv_step *gs = codecvt->__cd_out.__cd.__steps;
 213   int status;
 214   size_t dummy;
 215
 216   codecvt->__cd_out.__cd.__data[0].__outbuf = (unsigned char *) to_start;
 217   codecvt->__cd_out.__cd.__data[0].__outbufend = (unsigned char *) to_end;
 218   codecvt->__cd_out.__cd.__data[0].__statep = statep;
 219
 220   __gconv_fct fct = gs->__fct;
 221 #ifdef PTR_DEMANGLE
 222   if (gs->__shlib_handle != NULL)
 223     PTR_DEMANGLE (fct);
 224 #endif
 225
 226   status = DL_CALL_FCT (fct,
 227                         (gs, codecvt->__cd_out.__cd.__data, NULL, NULL,
 228                          NULL, &dummy, 1, 0));
 229
 230   *to_stop = (char *) codecvt->__cd_out.__cd.__data[0].__outbuf;
 231
 232   switch (status)
 233     {
 234     case __GCONV_OK:
 235     case __GCONV_EMPTY_INPUT:
 236       result = __codecvt_ok;
 237       break;
 238
 239     case __GCONV_FULL_OUTPUT:
 240     case __GCONV_INCOMPLETE_INPUT:
 241       result = __codecvt_partial;
 242       break;
 243
 244     default:
 245       result = __codecvt_error;
 246       break;
 247     }
 248
 249   return result;
 250 }
 251
 252
 253 static enum __codecvt_result
 254 do_in (struct _IO_codecvt *codecvt, __mbstate_t *statep,
 255        const char *from_start, const char *from_end, const char **from_stop,
 256        wchar_t *to_start, wchar_t *to_end, wchar_t **to_stop)
 257 {
 258   enum __codecvt_result result;
 259
 260   struct __gconv_step *gs = codecvt->__cd_in.__cd.__steps;
 261   int status;
 262   size_t dummy;
 263   const unsigned char *from_start_copy = (unsigned char *) from_start;
 264
 265   codecvt->__cd_in.__cd.__data[0].__outbuf = (unsigned char *) to_start;
 266   codecvt->__cd_in.__cd.__data[0].__outbufend = (unsigned char *) to_end;
 267   codecvt->__cd_in.__cd.__data[0].__statep = statep;
 268
 269   __gconv_fct fct = gs->__fct;
 270 #ifdef PTR_DEMANGLE
 271   if (gs->__shlib_handle != NULL)
 272     PTR_DEMANGLE (fct);
 273 #endif
 274
 275   status = DL_CALL_FCT (fct,
 276                         (gs, codecvt->__cd_in.__cd.__data, &from_start_copy,
 277                          (const unsigned char *) from_end, NULL,
 278                          &dummy, 0, 0));
 279
 280   *from_stop = (const char *) from_start_copy;
 281   *to_stop = (wchar_t *) codecvt->__cd_in.__cd.__data[0].__outbuf;
 282
 283   switch (status)
 284     {
 285     case __GCONV_OK:
 286     case __GCONV_EMPTY_INPUT:
 287       result = __codecvt_ok;
 288       break;
 289
 290     case __GCONV_FULL_OUTPUT:
 291     case __GCONV_INCOMPLETE_INPUT:
 292       result = __codecvt_partial;
 293       break;
 294
 295     default:
 296       result = __codecvt_error;
 297       break;
 298     }
 299
 300   return result;
 301 }
 302
 303
 304 static int
 305 do_encoding (struct _IO_codecvt *codecvt)
 306 {
 307   /* See whether the encoding is stateful.  */
 308   if (codecvt->__cd_in.__cd.__steps[0].__stateful)
 309     return -1;
 310   /* Fortunately not.  Now determine the input bytes for the conversion
 311      necessary for each wide character.  */
 312   if (codecvt->__cd_in.__cd.__steps[0].__min_needed_from
 313       != codecvt->__cd_in.__cd.__steps[0].__max_needed_from)
 314     /* Not a constant value.  */
 315     return 0;
 316
 317   return codecvt->__cd_in.__cd.__steps[0].__min_needed_from;
 318 }
 319
 320
 321 static int
 322 do_always_noconv (struct _IO_codecvt *codecvt)
 323 {
 324   return 0;
 325 }
 326
 327
 328 static int
 329 do_length (struct _IO_codecvt *codecvt, __mbstate_t *statep,
 330            const char *from_start, const char *from_end, _IO_size_t max)
 331 {
 332   int result;
 333   const unsigned char *cp = (const unsigned char *) from_start;
 334   wchar_t to_buf[max];
 335   struct __gconv_step *gs = codecvt->__cd_in.__cd.__steps;
 336   size_t dummy;
 337
 338   codecvt->__cd_in.__cd.__data[0].__outbuf = (unsigned char *) to_buf;
 339   codecvt->__cd_in.__cd.__data[0].__outbufend = (unsigned char *) &to_buf[max];
 340   codecvt->__cd_in.__cd.__data[0].__statep = statep;
 341
 342   __gconv_fct fct = gs->__fct;
 343 #ifdef PTR_DEMANGLE
 344   if (gs->__shlib_handle != NULL)
 345     PTR_DEMANGLE (fct);
 346 #endif
 347
 348   DL_CALL_FCT (fct,
 349                (gs, codecvt->__cd_in.__cd.__data, &cp,
 350                 (const unsigned char *) from_end, NULL,
 351                 &dummy, 0, 0));
 352
 353   result = cp - (const unsigned char *) from_start;
 354
 355   return result;
 356 }
 357
 358
 359 static int
 360 do_max_length (struct _IO_codecvt *codecvt)
 361 {
 362   return codecvt->__cd_in.__cd.__steps[0].__max_needed_from;
 363 }