src/language/data-io/save-translate.c

   1 /* PSPP - a program for statistical analysis.
   2    Copyright (C) 2010, 2011, 2013, 2016 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 <stdlib.h>
  20
  21 #include "data/case-map.h"
  22 #include "data/casereader.h"
  23 #include "data/casewriter.h"
  24 #include "data/csv-file-writer.h"
  25 #include "data/dataset.h"
  26 #include "data/dictionary.h"
  27 #include "data/file-name.h"
  28 #include "data/format.h"
  29 #include "data/settings.h"
  30 #include "language/command.h"
  31 #include "language/data-io/file-handle.h"
  32 #include "language/data-io/trim.h"
  33 #include "language/lexer/lexer.h"
  34 #include "libpspp/message.h"
  35
  36 #include "xalloc.h"
  37
  38 #include "gettext.h"
  39 #define _(msgid) gettext (msgid)
  40 #define N_(msgid) (msgid)
  41
  42 int
  43 cmd_save_translate (struct lexer *lexer, struct dataset *ds)
  44 {
  45   enum { CSV_FILE = 1, TAB_FILE } type;
  46
  47   struct dictionary *dict;
  48   struct case_map_stage *stage;
  49   struct case_map *map;
  50   struct casewriter *writer;
  51   struct file_handle *handle;
  52
  53   struct csv_writer_options csv_opts;
  54
  55   bool replace;
  56
  57   bool retain_unselected;
  58   bool recode_user_missing;
  59   bool include_var_names;
  60   bool use_value_labels;
  61   bool use_print_formats;
  62   char decimal;
  63   char delimiter;
  64   char qualifier;
  65
  66   bool ok;
  67
  68   type = 0;
  69
  70   dict = dict_clone (dataset_dict (ds));
  71   stage = NULL;
  72   map = NULL;
  73
  74   handle = NULL;
  75   replace = false;
  76
  77   retain_unselected = true;
  78   recode_user_missing = false;
  79   include_var_names = false;
  80   use_value_labels = false;
  81   use_print_formats = false;
  82   decimal = settings_get_decimal_char (FMT_F);
  83   delimiter = 0;
  84   qualifier = '"';
  85
  86   stage = case_map_stage_create (dict);
  87   dict_delete_scratch_vars (dict);
  88
  89   while (lex_token (lexer) != T_ENDCMD)
  90     {
  91       if (!lex_force_match (lexer, T_SLASH))
  92         goto error;
  93
  94       if (lex_match_id (lexer, "OUTFILE"))
  95         {
  96           if (handle != NULL)
  97             {
  98               lex_sbc_only_once ("OUTFILE");
  99               goto error;
 100             }
 101
 102           lex_match (lexer, T_EQUALS);
 103
 104           handle = fh_parse (lexer, FH_REF_FILE, NULL);
 105           if (handle == NULL)
 106             goto error;
 107         }
 108       else if (lex_match_id (lexer, "TYPE"))
 109         {
 110           if (type != 0)
 111             {
 112               lex_sbc_only_once ("TYPE");
 113               goto error;
 114             }
 115
 116           lex_match (lexer, T_EQUALS);
 117           if (lex_match_id (lexer, "CSV"))
 118             type = CSV_FILE;
 119           else if (lex_match_id (lexer, "TAB"))
 120             type = TAB_FILE;
 121           else
 122             {
 123               lex_error_expecting (lexer, "CSV", "TAB", NULL_SENTINEL);
 124               goto error;
 125             }
 126         }
 127       else if (lex_match_id (lexer, "REPLACE"))
 128         replace = true;
 129       else if (lex_match_id (lexer, "FIELDNAMES"))
 130         include_var_names = true;
 131       else if (lex_match_id (lexer, "MISSING"))
 132         {
 133           lex_match (lexer, T_EQUALS);
 134           if (lex_match_id (lexer, "IGNORE"))
 135             recode_user_missing = false;
 136           else if (lex_match_id (lexer, "RECODE"))
 137             recode_user_missing = true;
 138           else
 139             {
 140               lex_error_expecting (lexer, "IGNORE", "RECODE", NULL_SENTINEL);
 141               goto error;
 142             }
 143         }
 144       else if (lex_match_id (lexer, "CELLS"))
 145         {
 146           lex_match (lexer, T_EQUALS);
 147           if (lex_match_id (lexer, "VALUES"))
 148             use_value_labels = false;
 149           else if (lex_match_id (lexer, "LABELS"))
 150             use_value_labels = true;
 151           else
 152             {
 153               lex_error_expecting (lexer, "VALUES", "LABELS", NULL_SENTINEL);
 154               goto error;
 155             }
 156         }
 157       else if (lex_match_id (lexer, "TEXTOPTIONS"))
 158         {
 159           lex_match (lexer, T_EQUALS);
 160           for (;;)
 161             {
 162               if (lex_match_id (lexer, "DELIMITER"))
 163                 {
 164                   lex_match (lexer, T_EQUALS);
 165                   if (!lex_force_string (lexer))
 166                     goto error;
 167                   /* XXX should support multibyte UTF-8 delimiters */
 168                   if (ss_length (lex_tokss (lexer)) != 1)
 169                     {
 170                       msg (SE, _("The %s string must contain exactly one "
 171                                  "character."), "DELIMITER");
 172                       goto error;
 173                     }
 174                   delimiter = ss_first (lex_tokss (lexer));
 175                   lex_get (lexer);
 176                 }
 177               else if (lex_match_id (lexer, "QUALIFIER"))
 178                 {
 179                   lex_match (lexer, T_EQUALS);
 180                   if (!lex_force_string (lexer))
 181                     goto error;
 182                   /* XXX should support multibyte UTF-8 qualifiers */
 183                   if (ss_length (lex_tokss (lexer)) != 1)
 184                     {
 185                       msg (SE, _("The %s string must contain exactly one "
 186                                  "character."), "QUALIFIER");
 187                       goto error;
 188                     }
 189                   qualifier = ss_first (lex_tokss (lexer));
 190                   lex_get (lexer);
 191                 }
 192               else if (lex_match_id (lexer, "DECIMAL"))
 193                 {
 194                   lex_match (lexer, T_EQUALS);
 195                   if (lex_match_id (lexer, "DOT"))
 196                     decimal = '.';
 197                   else if (lex_match_id (lexer, "COMMA"))
 198                     decimal = ',';
 199                   else
 200                     {
 201                       lex_error_expecting (lexer, "DOT", "COMMA",
 202                                            NULL_SENTINEL);
 203                       goto error;
 204                     }
 205                 }
 206               else if (lex_match_id (lexer, "FORMAT"))
 207                 {
 208                   lex_match (lexer, T_EQUALS);
 209                   if (lex_match_id (lexer, "PLAIN"))
 210                     use_print_formats = false;
 211                   else if (lex_match_id (lexer, "VARIABLE"))
 212                     use_print_formats = true;
 213                   else
 214                     {
 215                       lex_error_expecting (lexer, "PLAIN", "VARIABLE",
 216                                            NULL_SENTINEL);
 217                       goto error;
 218                     }
 219                 }
 220               else
 221                 break;
 222             }
 223         }
 224       else if (lex_match_id (lexer, "UNSELECTED"))
 225         {
 226           lex_match (lexer, T_EQUALS);
 227           if (lex_match_id (lexer, "RETAIN"))
 228             retain_unselected = true;
 229           else if (lex_match_id (lexer, "DELETE"))
 230             retain_unselected = false;
 231           else
 232             {
 233               lex_error_expecting (lexer, "RETAIN", "DELETE", NULL_SENTINEL);
 234               goto error;
 235             }
 236         }
 237       else if (!parse_dict_trim (lexer, dict))
 238         goto error;
 239     }
 240
 241   if (type == 0)
 242     {
 243       lex_sbc_missing ("TYPE");
 244       goto error;
 245     }
 246   else if (handle == NULL)
 247     {
 248       lex_sbc_missing ("OUTFILE");
 249       goto error;
 250     }
 251   else if (!replace && fn_exists (handle))
 252     {
 253       msg (SE, _("Output file `%s' exists but %s was not specified."),
 254            fh_get_file_name (handle), "REPLACE");
 255       goto error;
 256     }
 257
 258   dict_delete_scratch_vars (dict);
 259   dict_compact_values (dict);
 260
 261   csv_opts.recode_user_missing = recode_user_missing;
 262   csv_opts.include_var_names = include_var_names;
 263   csv_opts.use_value_labels = use_value_labels;
 264   csv_opts.use_print_formats = use_print_formats;
 265   csv_opts.decimal = decimal;
 266   csv_opts.delimiter = (delimiter ? delimiter
 267                         : type == TAB_FILE ? '\t'
 268                         : decimal == '.' ? ','
 269                         : ';');
 270   csv_opts.qualifier = qualifier;
 271
 272   writer = csv_writer_open (handle, dict, &csv_opts);
 273   if (writer == NULL)
 274     goto error;
 275   fh_unref (handle);
 276
 277   map = case_map_stage_get_case_map (stage);
 278   case_map_stage_destroy (stage);
 279   if (map != NULL)
 280     writer = case_map_create_output_translator (map, writer);
 281   dict_destroy (dict);
 282
 283   casereader_transfer (proc_open_filtering (ds, !retain_unselected), writer);
 284   ok = casewriter_destroy (writer);
 285   ok = proc_commit (ds) && ok;
 286
 287   return ok ? CMD_SUCCESS : CMD_CASCADING_FAILURE;
 288
 289 error:
 290   case_map_stage_destroy (stage);
 291   fh_unref (handle);
 292   dict_destroy (dict);
 293   case_map_destroy (map);
 294   return CMD_FAILURE;
 295 }