1 /* PSPP - computes sample statistics.
2 Copyright (C) 1997-9, 2000 Free Software Foundation, Inc.
3 Written by Ben Pfaff <blp@gnu.org>.
5 This program is free software; you can redistribute it and/or
6 modify it under the terms of the GNU General Public License as
7 published by the Free Software Foundation; either version 2 of the
8 License, or (at your option) any later version.
10 This program is distributed in the hope that it will be useful, but
11 WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program; if not, write to the Free Software
17 Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
37 /* List of variable names. */
44 /* Represents a FLIP input program. */
47 struct variable **var; /* Variables to transpose. */
48 int var_cnt; /* Number of elements in `var'. */
49 int case_cnt; /* Pre-flip case count. */
50 size_t case_size; /* Post-flip bytes per case. */
52 struct variable *new_names; /* Variable containing new variable names. */
53 struct varname *new_names_head; /* First new variable. */
54 struct varname *new_names_tail; /* Last new variable. */
56 FILE *file; /* Temporary file containing data. */
59 static void destroy_flip_pgm (struct flip_pgm *);
60 static struct case_sink *flip_sink_create (struct flip_pgm *);
61 static struct case_source *flip_source_create (struct flip_pgm *);
62 static void flip_file (struct flip_pgm *);
63 static int build_dictionary (struct flip_pgm *);
65 static const struct case_source_class flip_source_class;
66 static const struct case_sink_class flip_sink_class;
68 /* Parses and executes FLIP. */
72 struct flip_pgm *flip;
76 msg (SM, _("FLIP ignores TEMPORARY. "
77 "Temporary transformations will be made permanent."));
81 flip = xmalloc (sizeof *flip);
85 flip->new_names = NULL;
86 flip->new_names_head = NULL;
87 flip->new_names_tail = NULL;
91 if (lex_match_id ("VARIABLES"))
94 if (!parse_variables (default_dict, &flip->var, &flip->var_cnt, PV_NO_DUPLICATE))
99 dict_get_vars (default_dict, &flip->var, &flip->var_cnt, 1u << DC_SYSTEM);
102 if (lex_match_id ("NEWNAMES"))
105 flip->new_names = parse_variable ();
106 if (!flip->new_names)
110 flip->new_names = dict_lookup_var (default_dict, "CASE_LBL");
116 for (i = 0; i < flip->var_cnt; i++)
117 if (flip->var[i] == flip->new_names)
119 memmove (&flip->var[i], &flip->var[i + 1], sizeof *flip->var * (flip->var_cnt - i - 1));
125 /* Read the active file into a flip_sink. */
127 temp_trns = temporary = 0;
128 vfm_sink = flip_sink_create (flip);
129 flip->new_names_tail = NULL;
130 procedure (NULL, NULL);
132 /* Flip the data we read. */
135 /* Flip the dictionary. */
136 dict_clear (default_dict);
137 if (!build_dictionary (flip))
139 discard_variables ();
142 flip->case_size = dict_get_case_size (default_dict);
144 /* Set up flipped data for reading. */
145 vfm_source = flip_source_create (flip);
147 return lex_end_of_command ();
150 destroy_flip_pgm (flip);
156 destroy_flip_pgm (struct flip_pgm *flip)
158 struct varname *iter, *next;
161 for (iter = flip->new_names_head; iter != NULL; iter = next)
166 if (flip->file != NULL)
171 /* Make a new variable with base name NAME, which is bowdlerized and
172 mangled until acceptable, and returns success. */
174 make_new_var (char name[])
176 /* Fix invalid characters. */
180 for (cp = name; *cp && !isspace (*cp); cp++)
182 *cp = toupper ((unsigned char) *cp);
183 if (!isalpha (*cp) && *cp != '@' && *cp != '#'
184 && (cp == name || (*cp != '.' && *cp != '$' && *cp != '_'
188 *cp = 'V'; /* _ not valid in first position. */
196 if (dict_create_var (default_dict, name, 0))
199 /* Add numeric extensions until acceptable. */
201 int len = (int) strlen (name);
205 for (i = 1; i < 10000000; i++)
207 int ofs = min (7 - intlog10 (i), len);
208 memcpy (n, name, ofs);
209 sprintf (&n[ofs], "%d", i);
211 if (dict_create_var (default_dict, n, 0))
216 msg (SE, _("Could not create acceptable variant for variable %s."), name);
220 /* Make a new dictionary for all the new variable names. */
222 build_dictionary (struct flip_pgm *flip)
224 dict_create_var_assert (default_dict, "CASE_LBL", 8);
226 if (flip->new_names_head == NULL)
230 if (flip->case_cnt > 99999)
232 msg (SE, _("Cannot create more than 99999 variable names."));
236 for (i = 0; i < flip->case_cnt; i++)
241 sprintf (s, "VAR%03d", i);
242 v = dict_create_var_assert (default_dict, s, 0);
249 for (v = flip->new_names_head; v; v = v->next)
250 if (!make_new_var (v->name))
257 /* Cases during transposition. */
258 struct flip_sink_info
260 struct flip_pgm *flip; /* FLIP program. */
261 union value *output_buf; /* Case output buffer. */
264 /* Creates a flip sink based on FLIP. */
265 static struct case_sink *
266 flip_sink_create (struct flip_pgm *flip)
268 struct flip_sink_info *info = xmalloc (sizeof *info);
272 info->output_buf = xmalloc (sizeof *info->output_buf * flip->var_cnt);
274 flip->file = tmpfile ();
276 msg (FE, _("Could not create temporary file for FLIP."));
278 /* Write variable names as first case. */
279 for (i = 0; i < flip->var_cnt; i++)
280 st_bare_pad_copy (info->output_buf[i].s, flip->var[i]->name, 8);
281 if (fwrite (info->output_buf, sizeof *info->output_buf,
282 flip->var_cnt, flip->file) != (size_t) flip->var_cnt)
283 msg (FE, _("Error writing FLIP file: %s."), strerror (errno));
287 return create_case_sink (&flip_sink_class, default_dict, info);
290 /* Writes case C to the FLIP sink. */
292 flip_sink_write (struct case_sink *sink, const struct ccase *c)
294 struct flip_sink_info *info = sink->aux;
295 struct flip_pgm *flip = info->flip;
300 if (flip->new_names != NULL)
302 struct varname *v = xmalloc (sizeof (struct varname));
304 if (flip->new_names->type == NUMERIC)
306 double f = c->data[sink->idx_to_fv[flip->new_names->index]].f;
309 strcpy (v->name, "VSYSMIS");
310 else if (f < INT_MIN)
311 strcpy (v->name, "VNEGINF");
312 else if (f > INT_MAX)
313 strcpy (v->name, "VPOSINF");
316 char name[INT_DIGITS + 2];
317 sprintf (name, "V%d", (int) f);
318 strncpy (v->name, name, 8);
324 int width = min (flip->new_names->width, 8);
325 memcpy (v->name, c->data[sink->idx_to_fv[flip->new_names->index]].s,
330 if (flip->new_names_head == NULL)
331 flip->new_names_head = v;
333 flip->new_names_tail->next = v;
334 flip->new_names_tail = v;
337 /* Write to external file. */
338 for (i = 0; i < flip->var_cnt; i++)
339 if (flip->var[i]->type == NUMERIC)
340 info->output_buf[i].f = c->data[sink->idx_to_fv[flip->var[i]->index]].f;
342 info->output_buf[i].f = SYSMIS;
344 if (fwrite (info->output_buf, sizeof *info->output_buf,
345 flip->var_cnt, flip->file) != (size_t) flip->var_cnt)
346 msg (FE, _("Error writing FLIP file: %s."), strerror (errno));
349 /* Transposes the external file into a new file. */
351 flip_file (struct flip_pgm *flip)
354 size_t case_capacity;
356 union value *input_buf, *output_buf;
357 FILE *input_file, *output_file;
359 /* Allocate memory for many cases. */
360 case_bytes = flip->var_cnt * sizeof *input_buf;
361 case_capacity = get_max_workspace() / case_bytes;
362 if (case_capacity > flip->case_cnt * 2)
363 case_capacity = flip->case_cnt * 2;
364 if (case_capacity < 2)
368 size_t bytes = case_bytes * case_capacity;
369 if (case_capacity > 2)
370 input_buf = malloc (bytes);
372 input_buf = xmalloc (bytes);
373 if (input_buf != NULL)
377 if (case_capacity < 2)
381 /* Use half the allocated memory for input_buf, half for
384 output_buf = input_buf + flip->var_cnt * case_capacity;
386 input_file = flip->file;
387 if (fseek (input_file, 0, SEEK_SET) != 0)
388 msg (FE, _("Error rewinding FLIP file: %s."), strerror (errno));
390 output_file = tmpfile ();
391 if (output_file == NULL)
392 msg (FE, _("Error creating FLIP source file."));
394 for (case_idx = 0; case_idx < flip->case_cnt; )
396 unsigned long read_cases = min (flip->case_cnt - case_idx,
400 if (read_cases != fread (input_buf, case_bytes, read_cases, input_file))
401 msg (FE, _("Error reading FLIP file: %s."), strerror (errno));
403 for (i = 0; i < flip->var_cnt; i++)
407 for (j = 0; j < read_cases; j++)
408 output_buf[j] = input_buf[i + j * flip->var_cnt];
410 if (fseek (output_file,
411 sizeof *input_buf * (case_idx + i * flip->case_cnt),
413 msg (FE, _("Error seeking FLIP source file: %s."),
416 if (fwrite (output_buf, sizeof *output_buf, read_cases, output_file)
418 msg (FE, _("Error writing FLIP source file: %s."),
422 case_idx += read_cases;
428 if (fseek (output_file, 0, SEEK_SET) != 0)
429 msg (FE, _("Error rewind FLIP source file: %s."), strerror (errno));
430 flip->file = output_file;
433 /* Destroy sink's internal data. */
435 flip_sink_destroy (struct case_sink *sink)
437 struct flip_sink_info *info = sink->aux;
439 free (info->output_buf);
443 /* FLIP sink class. */
444 static const struct case_sink_class flip_sink_class =
453 /* Creates and returns a FLIP source based on PGM,
454 which should have already been used as a sink. */
455 static struct case_source *
456 flip_source_create (struct flip_pgm *pgm)
458 return create_case_source (&flip_source_class, default_dict, pgm);
461 /* Reads the FLIP stream. Copies each case into C and calls
462 WRITE_CASE passing WC_DATA. */
464 flip_source_read (struct case_source *source,
466 write_case_func *write_case, write_case_data wc_data)
468 struct flip_pgm *flip = source->aux;
471 for (i = 0; i < flip->var_cnt; i++)
473 if (fread (c->data, sizeof *c->data, flip->case_cnt,
474 flip->file) != flip->case_cnt)
476 if (ferror (flip->file))
477 msg (SE, _("Error reading FLIP temporary file: %s."),
479 else if (feof (flip->file))
480 msg (SE, _("Unexpected end of file reading FLIP temporary file."));
486 if (!write_case (wc_data))
491 /* Destroy internal data in SOURCE. */
493 flip_source_destroy (struct case_source *source)
495 struct flip_pgm *flip = source->aux;
497 destroy_flip_pgm (flip);
500 static const struct case_source_class flip_source_class =