utilities/pspp-convert.c

   1 /* PSPP - a program for statistical analysis.
   2    Copyright (C) 2013, 2014 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation, either version 3 of the License, or
   7    (at your option) any later version.
   8
   9    This program 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
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program.  If not, see <http://www.gnu.org/licenses/>. */
  16
  17 #include <config.h>
  18
  19 #include <errno.h>
  20 #include <getopt.h>
  21 #include <limits.h>
  22 #include <stdlib.h>
  23 #include <unistd.h>
  24
  25 #include "data/any-reader.h"
  26 #include "data/casereader.h"
  27 #include "data/casewriter.h"
  28 #include "data/csv-file-writer.h"
  29 #include "data/file-name.h"
  30 #include "data/por-file-writer.h"
  31 #include "data/settings.h"
  32 #include "data/sys-file-encryption.h"
  33 #include "data/sys-file-writer.h"
  34 #include "data/file-handle-def.h"
  35 #include "libpspp/assertion.h"
  36 #include "libpspp/cast.h"
  37 #include "libpspp/i18n.h"
  38
  39 #include "gl/error.h"
  40 #include "gl/getpass.h"
  41 #include "gl/progname.h"
  42 #include "gl/version-etc.h"
  43
  44 #include "gettext.h"
  45 #define _(msgid) gettext (msgid)
  46
  47 static void usage (void);
  48
  49 static void decrypt_sav_file (struct encrypted_sys_file *enc,
  50                               const char *input_filename,
  51                               const char *output_filename,
  52                               const char *password);
  53
  54 int
  55 main (int argc, char *argv[])
  56 {
  57   const char *input_filename;
  58   const char *output_filename;
  59
  60   long long int max_cases = LLONG_MAX;
  61   struct dictionary *dict;
  62   struct casereader *reader;
  63   struct file_handle *input_fh;
  64   const char *encoding = NULL;
  65   struct encrypted_sys_file *enc;
  66
  67   const char *output_format = NULL;
  68   struct file_handle *output_fh;
  69   struct casewriter *writer;
  70   const char *password = NULL;
  71
  72   long long int i;
  73
  74   set_program_name (argv[0]);
  75   i18n_init ();
  76   fh_init ();
  77   settings_init ();
  78
  79   for (;;)
  80     {
  81       static const struct option long_options[] =
  82         {
  83           { "cases",    required_argument, NULL, 'c' },
  84           { "encoding", required_argument, NULL, 'e' },
  85           { "password", required_argument, NULL, 'p' },
  86
  87           { "output-format", required_argument, NULL, 'O' },
  88
  89           { "help",    no_argument,       NULL, 'h' },
  90           { "version", no_argument,       NULL, 'v' },
  91           { NULL,      0,                 NULL, 0 },
  92         };
  93
  94       int c;
  95
  96       c = getopt_long (argc, argv, "c:e:p:O:hv", long_options, NULL);
  97       if (c == -1)
  98         break;
  99
 100       switch (c)
 101         {
 102         case 'c':
 103           max_cases = strtoull (optarg, NULL, 0);
 104           break;
 105
 106         case 'e':
 107           encoding = optarg;
 108           break;
 109
 110         case 'p':
 111           password = optarg;
 112           break;
 113
 114         case 'O':
 115           output_format = optarg;
 116           break;
 117
 118         case 'v':
 119           version_etc (stdout, "pspp-convert", PACKAGE_NAME, PACKAGE_VERSION,
 120                        "Ben Pfaff", "John Darrington", NULL_SENTINEL);
 121           exit (EXIT_SUCCESS);
 122
 123         case 'h':
 124           usage ();
 125           exit (EXIT_SUCCESS);
 126
 127         default:
 128           exit (EXIT_FAILURE);
 129         }
 130     }
 131
 132   if (optind + 2 != argc)
 133     error (1, 0, _("exactly two non-option arguments are required; "
 134                    "use --help for help"));
 135
 136   input_filename = argv[optind];
 137   output_filename = argv[optind + 1];
 138   if (output_format == NULL)
 139     {
 140       const char *dot = strrchr (output_filename, '.');
 141       if (dot == NULL)
 142         error (1, 0, _("%s: cannot guess output format (use -O option)"),
 143                output_filename);
 144
 145       output_format = dot + 1;
 146     }
 147
 148   if (encrypted_sys_file_open (&enc, input_filename) > 0)
 149     {
 150       if (strcmp (output_format, "sav") && strcmp (output_format, "sys"))
 151         error (1, 0, _("can only convert encrypted data file to sav or sys "
 152                        "format"));
 153
 154       decrypt_sav_file (enc, input_filename, output_filename, password);
 155       goto exit;
 156     }
 157
 158   input_fh = fh_create_file (NULL, input_filename, fh_default_properties ());
 159   reader = any_reader_open_and_decode (input_fh, encoding, &dict, NULL);
 160   if (reader == NULL)
 161     exit (1);
 162
 163   output_fh = fh_create_file (NULL, output_filename, fh_default_properties ());
 164   if (!strcmp (output_format, "csv") || !strcmp (output_format, "txt"))
 165     {
 166       struct csv_writer_options options;
 167
 168       csv_writer_options_init (&options);
 169       options.include_var_names = true;
 170       writer = csv_writer_open (output_fh, dict, &options);
 171     }
 172   else if (!strcmp (output_format, "sav") || !strcmp (output_format, "sys"))
 173     {
 174       struct sfm_write_options options;
 175
 176       options = sfm_writer_default_options ();
 177       writer = sfm_open_writer (output_fh, dict, options);
 178     }
 179   else if (!strcmp (output_format, "por"))
 180     {
 181       struct pfm_write_options options;
 182
 183       options = pfm_writer_default_options ();
 184       writer = pfm_open_writer (output_fh, dict, options);
 185     }
 186   else
 187     {
 188       error (1, 0, _("%s: unknown output format (use -O option)"),
 189              output_filename);
 190       NOT_REACHED ();
 191     }
 192
 193   for (i = 0; i < max_cases; i++)
 194     {
 195       struct ccase *c;
 196
 197       c = casereader_read (reader);
 198       if (c == NULL)
 199         break;
 200
 201       casewriter_write (writer, c);
 202     }
 203
 204   if (!casereader_destroy (reader))
 205     error (1, 0, _("%s: error reading input file"), input_filename);
 206   if (!casewriter_destroy (writer))
 207     error (1, 0, _("%s: error writing output file"), output_filename);
 208
 209 exit:
 210   fh_done ();
 211   i18n_done ();
 212
 213   return 0;
 214 }
 215
 216 static void
 217 decrypt_sav_file (struct encrypted_sys_file *enc,
 218                   const char *input_filename,
 219                   const char *output_filename,
 220                   const char *password)
 221 {
 222   FILE *out;
 223   int err;
 224
 225   if (password == NULL)
 226     {
 227       password = getpass ("password: ");
 228       if (password == NULL)
 229         exit (1);
 230     }
 231
 232   if (!encrypted_sys_file_unlock (enc, password))
 233     error (1, 0, _("sorry, wrong password"));
 234
 235   out = fn_open (output_filename, "wb");
 236   if (out == NULL)
 237     error (1, errno, ("%s: error opening output file"), output_filename);
 238
 239   for (;;)
 240     {
 241       uint8_t buffer[1024];
 242       size_t n;
 243
 244       n = encrypted_sys_file_read (enc, buffer, sizeof buffer);
 245       if (n == 0)
 246         break;
 247
 248       if (fwrite (buffer, 1, n, out) != n)
 249         error (1, errno, ("%s: write error"), output_filename);
 250     }
 251
 252   err = encrypted_sys_file_close (enc);
 253   if (err)
 254     error (1, err, ("%s: read error"), input_filename);
 255
 256   if (fflush (out) == EOF)
 257     error (1, errno, ("%s: write error"), output_filename);
 258   fn_close (output_filename, out);
 259 }
 260
 261 static void
 262 usage (void)
 263 {
 264   printf ("\
 265 %s, a utility for converting SPSS data files to other formats.\n\
 266 Usage: %s [OPTION]... INPUT OUTPUT\n\
 267 where INPUT is an SPSS system or portable file\n\
 268   and OUTPUT is the name of the desired output file.\n\
 269 \n\
 270 The desired format of OUTPUT is by default inferred from its extension:\n\
 271   csv txt             comma-separated value\n\
 272   sav sys             SPSS system file\n\
 273   por                 SPSS portable file\n\
 274 \n\
 275 Options:\n\
 276   -O, --output-format=FORMAT  set specific output format, where FORMAT\n\
 277                       is one of the extensions listed above\n\
 278   -e, --encoding=CHARSET  override encoding of input data file\n\
 279   -c MAXCASES         limit number of cases to copy (default is all cases)\n\
 280   -p PASSWORD         password for encrypted .sav files\n\
 281   --help              display this help and exit\n\
 282   --version           output version information and exit\n",
 283           program_name, program_name);
 284 }