src/iconv.c

   1 /* Copyright (C) 2000-2001 Free Software Foundation, Inc.
   2    This file is part of the GNU LIBICONV Library.
   3
   4    The GNU LIBICONV Library is free software; you can redistribute it
   5    and/or modify it under the terms of the GNU Library General Public
   6    License as published by the Free Software Foundation; either version 2
   7    of the License, or (at your option) any later version.
   8
   9    The GNU LIBICONV Library is distributed in the hope that it will be
  10    useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  12    Library General Public License for more details.
  13
  14    You should have received a copy of the GNU Library General Public
  15    License along with the GNU LIBICONV Library; see the file COPYING.LIB.
  16    If not, write to the Free Software Foundation, Inc., 59 Temple Place -
  17    Suite 330, Boston, MA 02111-1307, USA.  */
  18
  19 #include "config.h"
  20 #include <stddef.h>
  21 #include <stdio.h>
  22 #include <stdlib.h>
  23 #include <string.h>
  24 #include <iconv.h>
  25 #include <errno.h>
  26 #if HAVE_LOCALE_H
  27 #include <locale.h>
  28 #endif
  29 #include <fcntl.h>
  30
  31 /* For systems that distinguish between text and binary I/O.
  32    O_BINARY is usually declared in <fcntl.h>. */
  33 #if !defined O_BINARY && defined _O_BINARY
  34   /* For MSC-compatible compilers.  */
  35 # define O_BINARY _O_BINARY
  36 # define O_TEXT _O_TEXT
  37 #endif
  38 #ifdef __BEOS__
  39   /* BeOS 5 has O_BINARY and O_TEXT, but they have no effect.  */
  40 # undef O_BINARY
  41 # undef O_TEXT
  42 #endif
  43 #if O_BINARY
  44 # if !(defined(__EMX__) || defined(__DJGPP__))
  45 #  define setmode _setmode
  46 #  define fileno _fileno
  47 # endif
  48 # ifdef __DJGPP__
  49 #  include <io.h> /* declares setmode() */
  50 #  include <unistd.h> /* declares isatty() */
  51 #  /* Avoid putting stdin/stdout in binary mode if it is connected to the
  52 #     console, because that would make it impossible for the user to
  53 #     interrupt the program through Ctrl-C or Ctrl-Break.  */
  54 #  define SET_BINARY(fd) (!isatty(fd) ? (setmode(fd,O_BINARY), 0) : 0)
  55 # else
  56 #  define SET_BINARY(fd) setmode(fd,O_BINARY)
  57 # endif
  58 #endif
  59
  60 #if O_BINARY
  61   static int force_binary = 0;
  62 #endif
  63
  64 static void usage (int exitcode)
  65 {
  66   const char* helpstring =
  67 #if O_BINARY
  68     "Usage: iconv [--binary] [-f fromcode] [-t tocode] [file ...]\n";
  69 #else
  70     "Usage: iconv [-f fromcode] [-t tocode] [file ...]\n";
  71 #endif
  72   fprintf(exitcode ? stderr : stdout, helpstring);
  73   exit(exitcode);
  74 }
  75
  76 static void print_version (void)
  77 {
  78   printf("iconv (GNU libiconv %d.%d)\n",
  79          _libiconv_version >> 8, _libiconv_version & 0xff);
  80   printf("\
  81 Copyright (C) 2000-2001 Free Software Foundation, Inc.\n\
  82 This is free software; see the source for copying conditions.  There is NO\n\
  83 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\
  84 Written by Bruno Haible.\n");
  85   exit(0);
  86 }
  87
  88 static int convert (iconv_t cd, FILE* infile, const char* infilename)
  89 {
  90   char inbuf[4096+4096];
  91   size_t inbufrest = 0;
  92   char outbuf[4096];
  93
  94 #if O_BINARY
  95   if (force_binary)
  96     SET_BINARY(fileno(infile));
  97 #endif
  98   iconv(cd,NULL,NULL,NULL,NULL);
  99   for (;;) {
 100     size_t inbufsize = fread(inbuf+4096,1,4096,infile);
 101     if (inbufsize == 0) {
 102       if (inbufrest == 0)
 103         break;
 104       else {
 105         fprintf(stderr,"iconv: %s: incomplete character or shift sequence\n",infilename);
 106         return 1;
 107       }
 108     } else {
 109       const char* inptr = inbuf+4096-inbufrest;
 110       size_t insize = inbufrest+inbufsize;
 111       inbufrest = 0;
 112       while (insize > 0) {
 113         char* outptr = outbuf;
 114         size_t outsize = sizeof(outbuf);
 115         size_t res = iconv(cd,(ICONV_CONST char**)&inptr,&insize,&outptr,&outsize);
 116         if (outptr != outbuf) {
 117           int saved_errno = errno;
 118           if (fwrite(outbuf,1,outptr-outbuf,stdout) < outptr-outbuf)
 119             return 1;
 120           errno = saved_errno;
 121         }
 122         if (res == (size_t)(-1)) {
 123           if (errno == EILSEQ) {
 124             fprintf(stderr,"iconv: %s: cannot convert\n",infilename);
 125             return 1;
 126           } else if (errno == EINVAL) {
 127             if (inbufsize == 0 || insize > 4096) {
 128               fprintf(stderr,"iconv: %s: incomplete character or shift sequence\n",infilename);
 129               return 1;
 130             } else {
 131               inbufrest = insize;
 132               if (insize > 0) {
 133                 /* Like memcpy(inbuf+4096-insize,inptr,insize), except that
 134                    we cannot use memcpy here, because source and destination
 135                    regions may overlap. */
 136                 char* restptr = inbuf+4096-insize;
 137                 do { *restptr++ = *inptr++; } while (--insize > 0);
 138               }
 139               break;
 140             }
 141           } else if (errno != E2BIG) {
 142             int saved_errno = errno;
 143             fprintf(stderr,"iconv: %s: ",infilename);
 144             errno = saved_errno;
 145             perror("");
 146             return 1;
 147           }
 148         }
 149       }
 150     }
 151   }
 152   {
 153     char* outptr = outbuf;
 154     size_t outsize = sizeof(outbuf);
 155     size_t res = iconv(cd,NULL,NULL,&outptr,&outsize);
 156     if (outptr != outbuf) {
 157       int saved_errno = errno;
 158       if (fwrite(outbuf,1,outptr-outbuf,stdout) < outptr-outbuf)
 159         return 1;
 160       errno = saved_errno;
 161     }
 162     if (res == (size_t)(-1)) {
 163       if (errno == EILSEQ) {
 164         fprintf(stderr,"iconv: %s: cannot convert\n",infilename);
 165         return 1;
 166       } else if (errno == EINVAL) {
 167         fprintf(stderr,"iconv: %s: incomplete character or shift sequence\n",infilename);
 168         return 1;
 169       } else {
 170         int saved_errno = errno;
 171         fprintf(stderr,"iconv: %s: ",infilename);
 172         errno = saved_errno;
 173         perror("");
 174         return 1;
 175       }
 176     }
 177   }
 178   if (ferror(infile)) {
 179     fprintf(stderr,"iconv: %s: I/O error\n",infilename);
 180     return 1;
 181   }
 182   return 0;
 183 }
 184
 185 int main (int argc, char* argv[])
 186 {
 187   const char* fromcode = NULL;
 188   const char* tocode = NULL;
 189   iconv_t cd;
 190   int i;
 191   int status;
 192
 193 #if HAVE_SETLOCALE
 194   /* Needed for the locale dependent encodings, "char" and "wchar_t". */
 195   setlocale(LC_CTYPE,"");
 196 #endif
 197   for (i = 1; i < argc;) {
 198     if (!strcmp(argv[i],"-f")) {
 199       if (i == argc-1) usage(1);
 200       if (fromcode != NULL) usage(1);
 201       fromcode = argv[i+1];
 202       i += 2;
 203       continue;
 204     }
 205     if (!strcmp(argv[i],"-t")) {
 206       if (i == argc-1) usage(1);
 207       if (tocode != NULL) usage(1);
 208       tocode = argv[i+1];
 209       i += 2;
 210       continue;
 211     }
 212     if (!strcmp(argv[i],"--help")) {
 213       usage(0);
 214     }
 215     if (!strcmp(argv[i],"--version")) {
 216       print_version();
 217     }
 218 #if O_BINARY
 219     if (!strcmp(argv[i],"--binary")) {
 220       force_binary = 1;
 221       i++;
 222       continue;
 223     }
 224 #endif
 225     if (argv[i][0] == '-')
 226       usage(1);
 227     break;
 228   }
 229 #if O_BINARY
 230   if (force_binary)
 231     SET_BINARY(fileno(stdout));
 232 #endif
 233   if (fromcode == NULL)
 234     fromcode = "char";
 235   if (tocode == NULL)
 236     tocode = "char";
 237   cd = iconv_open(tocode,fromcode);
 238   if (cd == (iconv_t)(-1)) {
 239     if (iconv_open("UCS-4",fromcode) == (iconv_t)(-1))
 240       fprintf(stderr,"iconv: conversion from %s unsupported\n",fromcode);
 241     else if (iconv_open(tocode,"UCS-4") == (iconv_t)(-1))
 242       fprintf(stderr,"iconv: conversion to %s unsupported\n",tocode);
 243     else
 244       fprintf(stderr,"iconv: conversion from %s to %s unsupported\n",fromcode,tocode);
 245     exit(1);
 246   }
 247   if (i == argc)
 248     status = convert(cd,stdin,"(stdin)");
 249   else {
 250     status = 0;
 251     for (; i < argc; i++) {
 252       const char* infilename = argv[i];
 253       FILE* infile = fopen(infilename,"r");
 254       if (infile == NULL) {
 255         int saved_errno = errno;
 256         fprintf(stderr,"iconv: %s: ",infilename);
 257         errno = saved_errno;
 258         perror("");
 259         status = 1;
 260       } else {
 261         status |= convert(cd,infile,infilename);
 262         fclose(infile);
 263       }
 264     }
 265   }
 266   iconv_close(cd);
 267   if (ferror(stdout)) {
 268     fprintf(stderr,"iconv: I/O error\n");
 269     status = 1;
 270   }
 271   exit(status);
 272 }