src/data/caseinit.c

   1 /* PSPP - a program for statistical analysis.
   2    Copyright (C) 2007, 2009, 2010, 2011 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/caseinit.h"
  20
  21 #include <stdbool.h>
  22 #include <stdlib.h>
  23 #include <string.h>
  24
  25 #include "data/case.h"
  26 #include "data/dictionary.h"
  27 #include "data/value.h"
  28 #include "data/variable.h"
  29 #include "libpspp/array.h"
  30 #include "libpspp/assertion.h"
  31 #include "libpspp/compiler.h"
  32
  33 #include "gl/xalloc.h"
  34 \f
  35 /* Initializer list: a set of values to write to locations within
  36    a case. */
  37
  38 /* Binds a value with a place to put it. */
  39 struct init_value
  40   {
  41     size_t case_index;
  42     int width;
  43     union value value;
  44   };
  45
  46 /* A set of values to initialize in a case. */
  47 struct init_list
  48   {
  49     struct init_value *values;
  50     size_t cnt;
  51   };
  52
  53 /* A bitmap of the "left" status of variables. */
  54 enum leave_class
  55   {
  56     LEAVE_REINIT = 0x001,       /* Reinitialize for every case. */
  57     LEAVE_LEFT = 0x002          /* Keep the value from one case to the next. */
  58   };
  59
  60 /* Initializes LIST as an empty initializer list. */
  61 static void
  62 init_list_create (struct init_list *list)
  63 {
  64   list->values = NULL;
  65   list->cnt = 0;
  66 }
  67
  68 /* Initializes NEW as a copy of OLD. */
  69 static void
  70 init_list_clone (struct init_list *new, const struct init_list *old)
  71 {
  72   size_t i;
  73
  74   new->values = xmemdup (old->values, old->cnt * sizeof *old->values);
  75   new->cnt = old->cnt;
  76
  77   for (i = 0; i < new->cnt; i++)
  78     {
  79       struct init_value *iv = &new->values[i];
  80       value_clone (&iv->value, &iv->value, iv->width);
  81     }
  82 }
  83
  84 /* Frees the storage associated with LIST. */
  85 static void
  86 init_list_destroy (struct init_list *list)
  87 {
  88   struct init_value *iv;
  89
  90   for (iv = &list->values[0]; iv < &list->values[list->cnt]; iv++)
  91     value_destroy (&iv->value, iv->width);
  92   free (list->values);
  93 }
  94
  95 /* Clears LIST, making it an empty list. */
  96 static void
  97 init_list_clear (struct init_list *list)
  98 {
  99   init_list_destroy (list);
 100   init_list_create (list);
 101 }
 102
 103 /* Compares `struct init_value's A and B by case_index and
 104    returns a strcmp()-type result. */
 105 static int
 106 compare_init_values (const void *a_, const void *b_, const void *aux UNUSED)
 107 {
 108   const struct init_value *a = a_;
 109   const struct init_value *b = b_;
 110
 111   return a->case_index < b->case_index ? -1 : a->case_index > b->case_index;
 112 }
 113
 114 /* Returns true if LIST includes CASE_INDEX, false otherwise. */
 115 static bool
 116 init_list_includes (const struct init_list *list, size_t case_index)
 117 {
 118   struct init_value value;
 119   value.case_index = case_index;
 120   return binary_search (list->values, list->cnt, sizeof *list->values,
 121                         &value, compare_init_values, NULL) != NULL;
 122 }
 123
 124 /* Marks LIST to initialize the `union value's for the variables
 125    in dictionary D that both (1) fall in the leave class or
 126    classes designated by INCLUDE and (2) are not in EXCLUDE. */
 127 static void
 128 init_list_mark (struct init_list *list, const struct init_list *exclude,
 129                 enum leave_class include, const struct dictionary *d)
 130 {
 131   size_t n_vars = dict_get_n_vars (d);
 132
 133   assert (list != exclude);
 134   list->values = xnrealloc (list->values, list->cnt + dict_get_n_vars (d),
 135                             sizeof *list->values);
 136   for (size_t i = 0; i < n_vars; i++)
 137     {
 138       struct variable *v = dict_get_var (d, i);
 139       size_t case_index = var_get_case_index (v);
 140       struct init_value *iv;
 141
 142       /* Only include the correct class. */
 143       if (!(include & (var_get_leave (v) ? LEAVE_LEFT : LEAVE_REINIT)))
 144         continue;
 145
 146       /* Don't include those to be excluded. */
 147       if (exclude != NULL && init_list_includes (exclude, case_index))
 148         continue;
 149
 150       iv = &list->values[list->cnt++];
 151       iv->case_index = case_index;
 152       iv->width = var_get_width (v);
 153       value_init (&iv->value, iv->width);
 154       if (var_is_numeric (v) && var_get_leave (v))
 155         iv->value.f = 0;
 156       else
 157         value_set_missing (&iv->value, iv->width);
 158     }
 159
 160   /* Drop duplicates. */
 161   list->cnt = sort_unique (list->values, list->cnt, sizeof *list->values,
 162                            compare_init_values, NULL);
 163 }
 164
 165 /* Initializes data in case C to the values in the initializer
 166    LIST. */
 167 static void
 168 init_list_init (const struct init_list *list, struct ccase *c)
 169 {
 170   const struct init_value *iv;
 171
 172   for (iv = &list->values[0]; iv < &list->values[list->cnt]; iv++)
 173     value_copy (case_data_rw_idx (c, iv->case_index), &iv->value, iv->width);
 174 }
 175
 176 /* Updates the values in the initializer LIST from the data in
 177    case C. */
 178 static void
 179 init_list_update (const struct init_list *list, const struct ccase *c)
 180 {
 181   struct init_value *iv;
 182
 183   for (iv = &list->values[0]; iv < &list->values[list->cnt]; iv++)
 184     value_copy (&iv->value, case_data_idx (c, iv->case_index), iv->width);
 185 }
 186 \f
 187 /* A case initializer. */
 188 struct caseinit
 189   {
 190     /* Values that do not need to be initialized by the
 191        procedure, because they are initialized by the data
 192        source. */
 193     struct init_list preinited_values;
 194
 195     /* Values that need to be initialized to SYSMIS or spaces in
 196        each case. */
 197     struct init_list reinit_values;
 198
 199     /* Values that need to be initialized to 0 or spaces in the
 200        first case and thereafter retain their values from case to
 201        case. */
 202     struct init_list left_values;
 203   };
 204
 205 /* Creates and returns a new case initializer. */
 206 struct caseinit *
 207 caseinit_create (void)
 208 {
 209   struct caseinit *ci = xmalloc (sizeof *ci);
 210   init_list_create (&ci->preinited_values);
 211   init_list_create (&ci->reinit_values);
 212   init_list_create (&ci->left_values);
 213   return ci;
 214 }
 215
 216 /* Creates and returns a copy of OLD. */
 217 struct caseinit *
 218 caseinit_clone (struct caseinit *old)
 219 {
 220   struct caseinit *new = xmalloc (sizeof *new);
 221   init_list_clone (&new->preinited_values, &old->preinited_values);
 222   init_list_clone (&new->reinit_values, &old->reinit_values);
 223   init_list_clone (&new->left_values, &old->left_values);
 224   return new;
 225 }
 226
 227 /* Clears the contents of case initializer CI. */
 228 void
 229 caseinit_clear (struct caseinit *ci)
 230 {
 231   init_list_clear (&ci->preinited_values);
 232   init_list_clear (&ci->reinit_values);
 233   init_list_clear (&ci->left_values);
 234 }
 235
 236 /* Destroys case initializer CI. */
 237 void
 238 caseinit_destroy (struct caseinit *ci)
 239 {
 240   if (ci != NULL)
 241     {
 242       init_list_destroy (&ci->preinited_values);
 243       init_list_destroy (&ci->reinit_values);
 244       init_list_destroy (&ci->left_values);
 245       free (ci);
 246     }
 247 }
 248
 249 /* Marks the variables from dictionary D in CI as being
 250    initialized by the data source, so that the case initializer
 251    need not initialize them itself. */
 252 void
 253 caseinit_mark_as_preinited (struct caseinit *ci, const struct dictionary *d)
 254 {
 255   init_list_mark (&ci->preinited_values, NULL, LEAVE_REINIT | LEAVE_LEFT, d);
 256 }
 257
 258 /* Marks in CI the variables from dictionary D, except for any
 259    variables that were already marked with
 260    caseinit_mark_as_preinited, as needing initialization
 261    according to their leave status. */
 262 void
 263 caseinit_mark_for_init (struct caseinit *ci, const struct dictionary *d)
 264 {
 265   init_list_mark (&ci->reinit_values, &ci->preinited_values, LEAVE_REINIT, d);
 266   init_list_mark (&ci->left_values, &ci->preinited_values, LEAVE_LEFT, d);
 267 }
 268
 269 /* Initializes variables in *C as described by CI.
 270    C must not be shared. */
 271 void
 272 caseinit_init_vars (const struct caseinit *ci, struct ccase *c)
 273 {
 274   init_list_init (&ci->reinit_values, c);
 275   init_list_init (&ci->left_values, c);
 276 }
 277
 278 /* Updates the left vars in CI from the data in C, so that the
 279    next call to caseinit_init_vars will store those values in the
 280    next case. */
 281 void
 282 caseinit_update_left_vars (struct caseinit *ci, const struct ccase *c)
 283 {
 284   init_list_update (&ci->left_values, c);
 285 }
 286