src/data/any-reader.c

   1 /* PSPP - a program for statistical analysis.
   2    Copyright (C) 2006, 2010, 2011, 2012, 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 "data/any-reader.h"
  20
  21 #include <assert.h>
  22 #include <errno.h>
  23 #include <stdbool.h>
  24 #include <stdio.h>
  25 #include <stdlib.h>
  26
  27 #include "data/casereader.h"
  28 #include "data/dataset.h"
  29 #include "data/dictionary.h"
  30 #include "data/file-handle-def.h"
  31 #include "data/file-name.h"
  32 #include "libpspp/assertion.h"
  33 #include "libpspp/cast.h"
  34 #include "libpspp/message.h"
  35 #include "libpspp/str.h"
  36
  37 #include "gl/xalloc.h"
  38
  39 #include "gettext.h"
  40 #define _(msgid) gettext (msgid)
  41 #define N_(msgid) (msgid)
  42
  43 static const struct any_reader_class dataset_reader_class;
  44
  45 static const struct any_reader_class *classes[] =
  46   {
  47     &sys_file_reader_class,
  48     &por_file_reader_class,
  49     &pcp_file_reader_class,
  50   };
  51 enum { N_CLASSES = sizeof classes / sizeof *classes };
  52
  53 int
  54 any_reader_detect (const struct file_handle *file_handle,
  55                    const struct any_reader_class **classp)
  56 {
  57   struct detector
  58     {
  59       enum any_type type;
  60       int (*detect) (FILE *);
  61     };
  62
  63   FILE *file;
  64   int retval;
  65
  66   if (classp)
  67     *classp = NULL;
  68
  69   file = fn_open (file_handle, "rb");
  70   if (file == NULL)
  71     {
  72       msg (ME, _("An error occurred while opening `%s': %s."),
  73            fh_get_file_name (file_handle), strerror (errno));
  74       return -errno;
  75     }
  76
  77   retval = 0;
  78   for (int i = 0; i < N_CLASSES; i++)
  79     {
  80       int rc = classes[i]->detect (file);
  81       if (rc == 1)
  82         {
  83           retval = 1;
  84           if (classp)
  85             *classp = classes[i];
  86           break;
  87         }
  88       else if (rc < 0)
  89         retval = rc;
  90     }
  91
  92   if (retval < 0)
  93     msg (ME, _("Error reading `%s': %s."), fh_get_file_name (file_handle), strerror (-retval));
  94
  95   fn_close (file_handle, file);
  96
  97   return retval;
  98 }
  99
 100 struct any_reader *
 101 any_reader_open (struct file_handle *handle)
 102 {
 103   switch (fh_get_referent (handle))
 104     {
 105     case FH_REF_FILE:
 106       {
 107         const struct any_reader_class *class;
 108         int retval;
 109
 110         retval = any_reader_detect (handle, &class);
 111         if (retval <= 0)
 112           {
 113             if (retval == 0)
 114               msg (SE, _("`%s' is not a system or portable file."),
 115                    fh_get_file_name (handle));
 116             return NULL;
 117           }
 118
 119         return class->open (handle);
 120       }
 121
 122     case FH_REF_INLINE:
 123       msg (SE, _("The inline file is not allowed here."));
 124       return NULL;
 125
 126     case FH_REF_DATASET:
 127       return dataset_reader_class.open (handle);
 128     }
 129   NOT_REACHED ();
 130 }
 131
 132 /* gnulib on some systems defines "close" as something else,
 133    which causes problems for this code.  So undefine it here.  */
 134 #undef close
 135
 136 bool
 137 any_reader_close (struct any_reader *any_reader)
 138 {
 139   return any_reader ? any_reader->klass->close (any_reader) : true;
 140 }
 141
 142 struct casereader *
 143 any_reader_decode (struct any_reader *any_reader,
 144                    const char *encoding,
 145                    struct dictionary **dictp,
 146                    struct any_read_info *info)
 147 {
 148   const struct any_reader_class *class = any_reader->klass;
 149   struct casereader *reader;
 150
 151   reader = any_reader->klass->decode (any_reader, encoding, dictp, info);
 152   if (reader && info)
 153     info->klass = class;
 154   return reader;
 155 }
 156
 157 size_t
 158 any_reader_get_strings (const struct any_reader *any_reader, struct pool *pool,
 159                         char ***labels, bool **ids, char ***values)
 160 {
 161   return (any_reader->klass->get_strings
 162           ? any_reader->klass->get_strings (any_reader, pool, labels, ids,
 163                                             values)
 164           : 0);
 165 }
 166
 167 struct casereader *
 168 any_reader_open_and_decode (struct file_handle *handle,
 169                             const char *encoding,
 170                             struct dictionary **dictp,
 171                             struct any_read_info *info)
 172 {
 173   struct any_reader *any_reader = any_reader_open (handle);
 174   return (any_reader
 175           ? any_reader_decode (any_reader, encoding, dictp, info)
 176           : NULL);
 177 }
 178 \f
 179 struct dataset_reader
 180   {
 181     struct any_reader any_reader;
 182     struct dictionary *dict;
 183     struct casereader *reader;
 184   };
 185
 186 /* Opens FH, which must have referent type FH_REF_DATASET, and returns a
 187    dataset_reader for it, or a null pointer on failure.  Stores a copy of the
 188    dictionary for the dataset file into *DICT.  The caller takes ownership of
 189    the casereader and the dictionary.  */
 190 static struct any_reader *
 191 dataset_reader_open (struct file_handle *fh)
 192 {
 193   struct dataset_reader *reader;
 194   struct dataset *ds;
 195
 196   /* We don't bother doing fh_lock or fh_ref on the file handle,
 197      as there's no advantage in this case, and doing these would
 198      require us to keep track of the "struct file_handle" and
 199      "struct fh_lock" and undo our work later. */
 200   assert (fh_get_referent (fh) == FH_REF_DATASET);
 201
 202   ds = fh_get_dataset (fh);
 203   if (ds == NULL || !dataset_has_source (ds))
 204     {
 205       msg (SE, _("Cannot read from dataset %s because no dictionary or data "
 206                  "has been written to it yet."),
 207            fh_get_name (fh));
 208       return NULL;
 209     }
 210
 211   reader = xmalloc (sizeof *reader);
 212   reader->any_reader.klass = &dataset_reader_class;
 213   reader->dict = dict_clone (dataset_dict (ds));
 214   reader->reader = casereader_clone (dataset_source (ds));
 215   return &reader->any_reader;
 216 }
 217
 218 static struct dataset_reader *
 219 dataset_reader_cast (const struct any_reader *r_)
 220 {
 221   assert (r_->klass == &dataset_reader_class);
 222   return UP_CAST (r_, struct dataset_reader, any_reader);
 223 }
 224
 225 static bool
 226 dataset_reader_close (struct any_reader *r_)
 227 {
 228   struct dataset_reader *r = dataset_reader_cast (r_);
 229   dict_unref (r->dict);
 230   casereader_destroy (r->reader);
 231   free (r);
 232
 233   return true;
 234 }
 235
 236 static struct casereader *
 237 dataset_reader_decode (struct any_reader *r_, const char *encoding UNUSED,
 238                        struct dictionary **dictp, struct any_read_info *info)
 239 {
 240   struct dataset_reader *r = dataset_reader_cast (r_);
 241   struct casereader *reader;
 242
 243   *dictp = r->dict;
 244   reader = r->reader;
 245   if (info)
 246     {
 247       memset (info, 0, sizeof *info);
 248       info->integer_format = INTEGER_NATIVE;
 249       info->float_format = FLOAT_NATIVE_DOUBLE;
 250       info->compression = ANY_COMP_NONE;
 251       info->case_cnt = casereader_get_case_cnt (reader);
 252     }
 253   free (r);
 254
 255   return reader;
 256 }
 257
 258 static const struct any_reader_class dataset_reader_class =
 259   {
 260     N_("Dataset"),
 261     NULL,
 262     dataset_reader_open,
 263     dataset_reader_close,
 264     dataset_reader_decode,
 265     NULL,
 266   };