src/language/expressions/evaluate.c

   1 /* PSPP - a program for statistical analysis.
   2    Copyright (C) 1997-9, 2000, 2006, 2007, 2009, 2010, 2011, 2012 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 "language/expressions/private.h"
  20 #include "evaluate.h"
  21
  22 #include <ctype.h>
  23
  24 #include "libpspp/assertion.h"
  25 #include "libpspp/message.h"
  26 #include "language/expressions/helpers.h"
  27 #include "language/lexer/format-parser.h"
  28 #include "language/lexer/value-parser.h"
  29 #include "libpspp/pool.h"
  30 #include "output/driver.h"
  31
  32 #include "xalloc.h"
  33
  34 static void
  35 expr_evaluate (struct expression *e, const struct ccase *c, int case_idx,
  36                void *result)
  37 {
  38   struct dataset *ds = e->ds;
  39   union operation_data *op = e->ops;
  40
  41   double *ns = e->number_stack;
  42   struct substring *ss = e->string_stack;
  43
  44   /* Without a dictionary/dataset, the expression can't refer to variables,
  45      and you don't need to specify a case when you evaluate the
  46      expression.  With a dictionary/dataset, the expression can refer
  47      to variables, so you must specify a case when you evaluate the
  48      expression. */
  49   assert ((c != NULL) == (e->ds != NULL));
  50
  51   pool_clear (e->eval_pool);
  52
  53   for (;;)
  54     {
  55       assert (op < e->ops + e->n_ops);
  56       switch (op++->operation)
  57         {
  58         case OP_number:
  59         case OP_boolean:
  60           *ns++ = op++->number;
  61           break;
  62
  63         case OP_string:
  64           {
  65             const struct substring *s = &op++->string;
  66             *ss++ = copy_string (e, s->string, s->length);
  67           }
  68           break;
  69
  70         case OP_return_number:
  71           *(double *) result = isfinite (ns[-1]) ? ns[-1] : SYSMIS;
  72           return;
  73
  74         case OP_return_string:
  75           *(struct substring *) result = ss[-1];
  76           return;
  77
  78 #include "evaluate.inc"
  79
  80         default:
  81           NOT_REACHED ();
  82         }
  83     }
  84 }
  85
  86 double
  87 expr_evaluate_num (struct expression *e, const struct ccase *c, int case_idx)
  88 {
  89   double d;
  90
  91   assert (e->type == OP_number || e->type == OP_boolean);
  92   expr_evaluate (e, c, case_idx, &d);
  93   return d;
  94 }
  95
  96 void
  97 expr_evaluate_str (struct expression *e, const struct ccase *c, int case_idx,
  98                    char *dst, size_t dst_size)
  99 {
 100   struct substring s;
 101
 102   assert (e->type == OP_string);
 103   assert ((dst == NULL) == (dst_size == 0));
 104   expr_evaluate (e, c, case_idx, &s);
 105
 106   buf_copy_rpad (dst, dst_size, s.string, s.length, ' ');
 107 }
 108 \f
 109 #include "language/lexer/lexer.h"
 110 #include "language/command.h"
 111
 112 static bool default_optimize = true;
 113
 114 int
 115 cmd_debug_evaluate (struct lexer *lexer, struct dataset *dsother UNUSED)
 116 {
 117   bool optimize = default_optimize;
 118   int retval = CMD_FAILURE;
 119   bool dump_postfix = false;
 120   bool set_defaults = false;
 121
 122   struct ccase *c = NULL;
 123
 124   struct dataset *ds = NULL;
 125
 126   char *name = NULL;
 127   char *title = NULL;
 128   struct fmt_spec format;
 129   bool has_format = false;
 130
 131   struct expression *expr;
 132
 133   struct dictionary *d = NULL;
 134
 135   for (;;)
 136     {
 137       if (lex_match_id (lexer, "NOOPTIMIZE"))
 138         optimize = false;
 139       else if (lex_match_id (lexer, "OPTIMIZE"))
 140         optimize = true;
 141       else if (lex_match_id (lexer, "POSTFIX"))
 142         dump_postfix = 1;
 143       else if (lex_match_id (lexer, "SET"))
 144         set_defaults = true;
 145       else if (lex_match (lexer, T_LPAREN))
 146         {
 147           struct variable *v;
 148
 149           if (!lex_force_id (lexer))
 150             goto done;
 151           name = xstrdup (lex_tokcstr (lexer));
 152
 153           lex_get (lexer);
 154           if (!lex_force_match (lexer, T_EQUALS))
 155             goto done;
 156
 157           union value value;
 158           int width;
 159           if (lex_is_number (lexer))
 160             {
 161               width = 0;
 162               value.f = lex_number (lexer);
 163               lex_get (lexer);
 164             }
 165           else if (lex_match_id (lexer, "SYSMIS"))
 166             {
 167               width = 0;
 168               value.f = SYSMIS;
 169             }
 170           else if (lex_is_string (lexer))
 171             {
 172               width = ss_length (lex_tokss (lexer));
 173               value.s = CHAR_CAST (uint8_t *, ss_xstrdup (lex_tokss (lexer)));
 174               lex_get (lexer);
 175             }
 176           else
 177             {
 178               lex_error (lexer, _("expecting number or string"));
 179               goto done;
 180             }
 181
 182           if  (ds == NULL)
 183             {
 184               ds = dataset_create (NULL, "");
 185               d = dataset_dict (ds);
 186             }
 187
 188           v = dict_create_var (d, name, width);
 189           if (v == NULL)
 190             {
 191               msg (SE, _("Duplicate variable name %s."), name);
 192               value_destroy (&value, width);
 193               goto done;
 194             }
 195           free (name);
 196           name = NULL;
 197
 198           if (lex_match_id (lexer, "MISSING"))
 199             {
 200               struct missing_values mv;
 201               mv_init (&mv, width);
 202               mv_add_value (&mv, &value);
 203               var_set_missing_values (v, &mv);
 204               mv_destroy (&mv);
 205             }
 206
 207           if (c == NULL)
 208             c = case_create (dict_get_proto (d));
 209           else
 210             c = case_unshare_and_resize (c, dict_get_proto (d));
 211           value_swap (case_data_rw (c, v), &value);
 212           value_destroy (&value, width);
 213
 214           if (!lex_force_match (lexer, T_RPAREN))
 215             goto done;
 216         }
 217       else if (lex_match_id (lexer, "VECTOR"))
 218         {
 219           struct variable **vars;
 220           size_t n;
 221           dict_get_vars_mutable (d, &vars, &n, 0);
 222           dict_create_vector_assert (d, "V", vars, n);
 223           free (vars);
 224         }
 225       else if (lex_match_id (lexer, "FORMAT"))
 226         {
 227           lex_match (lexer, T_EQUALS);
 228           if (!parse_format_specifier (lexer, &format)
 229               || !fmt_check_output (&format)
 230               || !fmt_check_type_compat (&format, VAL_NUMERIC))
 231             goto done;
 232           has_format = true;
 233         }
 234       else
 235         break;
 236     }
 237
 238   if (set_defaults)
 239     {
 240       retval = CMD_SUCCESS;
 241       default_optimize = optimize;
 242       goto done;
 243     }
 244
 245   if (!lex_force_match (lexer, T_SLASH))
 246     goto done;
 247
 248   for (size_t i = 1; ; i++)
 249     if (lex_next_token (lexer, i) == T_ENDCMD)
 250       {
 251         title = lex_next_representation (lexer, 0, i - 1);
 252         break;
 253       }
 254
 255   expr = expr_parse_any (lexer, ds, optimize);
 256   if (!expr || lex_end_of_command (lexer) != CMD_SUCCESS)
 257     {
 258       if (expr != NULL)
 259         expr_free (expr);
 260       output_log ("%s => error", title);
 261       goto done;
 262     }
 263
 264   if (dump_postfix)
 265     expr_debug_print_postfix (expr);
 266   else
 267     switch (expr->type)
 268       {
 269       case OP_number:
 270       case OP_num_vec_elem:
 271         {
 272           double d = expr_evaluate_num (expr, c, 0);
 273           if (has_format)
 274             {
 275               char *output = data_out (&(const union value) { .f = d },
 276                                        NULL, &format,
 277                                        settings_get_fmt_settings ());
 278               output_log ("%s => %s", title, output);
 279               free (output);
 280             }
 281           else if (d == SYSMIS)
 282             output_log ("%s => sysmis", title);
 283           else
 284             output_log ("%s => %.2f", title, d);
 285         }
 286         break;
 287
 288       case OP_boolean:
 289         {
 290           double b = expr_evaluate_num (expr, c, 0);
 291           output_log ("%s => %s", title,
 292                       b == SYSMIS ? "sysmis" : b == 0.0 ? "false" : "true");
 293         }
 294         break;
 295
 296       case OP_string:
 297         {
 298           struct substring out;
 299           expr_evaluate (expr, c, 0, &out);
 300           output_log ("%s => \"%.*s\"", title, (int) out.length, out.string);
 301           break;
 302         }
 303
 304       default:
 305         NOT_REACHED ();
 306       }
 307
 308   expr_free (expr);
 309   retval = CMD_SUCCESS;
 310
 311  done:
 312   dataset_destroy (ds);
 313
 314   case_unref (c);
 315
 316   free (name);
 317   free (title);
 318
 319   return retval;
 320 }
 321
 322 void
 323 expr_debug_print_postfix (const struct expression *e)
 324 {
 325   struct string s = DS_EMPTY_INITIALIZER;
 326
 327   for (size_t i = 0; i < e->n_ops; i++)
 328     {
 329       union operation_data *op = &e->ops[i];
 330       if (i > 0)
 331         ds_put_byte (&s, ' ');
 332       switch (e->op_types[i])
 333         {
 334         case OP_operation:
 335           if (op->operation == OP_return_number)
 336             ds_put_cstr (&s, "return_number");
 337           else if (op->operation == OP_return_string)
 338             ds_put_cstr (&s, "return_string");
 339           else if (is_function (op->operation))
 340             ds_put_format (&s, "%s", operations[op->operation].prototype);
 341           else if (is_composite (op->operation))
 342             ds_put_format (&s, "%s", operations[op->operation].name);
 343           else
 344             ds_put_format (&s, "%s:", operations[op->operation].name);
 345           break;
 346         case OP_number:
 347           if (op->number != SYSMIS)
 348             ds_put_format (&s, "n<%g>", op->number);
 349           else
 350             ds_put_cstr (&s, "n<SYSMIS>");
 351           break;
 352         case OP_string:
 353           ds_put_cstr (&s, "s<");
 354           ds_put_substring (&s, op->string);
 355           ds_put_byte (&s, '>');
 356           break;
 357         case OP_format:
 358           {
 359             char str[FMT_STRING_LEN_MAX + 1];
 360             fmt_to_string (op->format, str);
 361             ds_put_format (&s, "f<%s>", str);
 362           }
 363           break;
 364         case OP_variable:
 365           ds_put_format (&s, "v<%s>", var_get_name (op->variable));
 366           break;
 367         case OP_vector:
 368           ds_put_format (&s, "vec<%s>", vector_get_name (op->vector));
 369           break;
 370         case OP_integer:
 371           ds_put_format (&s, "i<%d>", op->integer);
 372           break;
 373         case OP_expr_node:
 374           ds_put_cstr (&s, "expr_node");
 375           break;
 376         default:
 377           NOT_REACHED ();
 378         }
 379     }
 380   output_log_nocopy (ds_steal_cstr (&s));
 381 }