1 /* PSPP - a program for statistical analysis.
2 Copyright (C) 1997-9, 2000, 2009, 2010, 2011, 2012, 2013, 2015 Free Software Foundation, Inc.
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.
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.
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/>. */
22 #include "data/case.h"
23 #include "data/caseinit.h"
24 #include "data/casereader-provider.h"
25 #include "data/control-stack.h"
26 #include "data/dataset.h"
27 #include "data/dictionary.h"
28 #include "data/session.h"
29 #include "data/transformations.h"
30 #include "data/variable.h"
31 #include "language/command.h"
32 #include "language/data-io/data-reader.h"
33 #include "language/data-io/file-handle.h"
34 #include "language/data-io/inpt-pgm.h"
35 #include "language/expressions/public.h"
36 #include "language/lexer/lexer.h"
37 #include "libpspp/assertion.h"
38 #include "libpspp/compiler.h"
39 #include "libpspp/message.h"
40 #include "libpspp/misc.h"
41 #include "libpspp/str.h"
43 #include "gl/xalloc.h"
46 #define _(msgid) gettext (msgid)
48 /* Indicates how a `union value' should be initialized. */
49 struct input_program_pgm
51 struct session *session;
53 struct trns_chain *trns_chain;
54 enum trns_result restart;
56 casenumber case_nr; /* Incremented by END CASE transformation. */
58 struct caseinit *init;
59 struct caseproto *proto;
62 static void destroy_input_program (struct input_program_pgm *);
63 static trns_proc_func end_case_trns_proc;
64 static trns_proc_func reread_trns_proc;
65 static trns_proc_func end_file_trns_proc;
66 static trns_free_func reread_trns_free;
68 static const struct casereader_class input_program_casereader_class;
70 static bool inside_input_program;
72 /* Returns true if we're parsing the inside of a INPUT
73 PROGRAM...END INPUT PROGRAM construct, false otherwise. */
75 in_input_program (void)
77 return inside_input_program;
80 /* Emits an END CASE transformation for INP. */
82 emit_END_CASE (struct dataset *ds, struct input_program_pgm *inp)
84 add_transformation (ds, end_case_trns_proc, NULL, inp);
88 cmd_input_program (struct lexer *lexer, struct dataset *ds)
90 struct input_program_pgm *inp;
91 bool saw_END_CASE = false;
92 bool saw_END_FILE = false;
93 bool saw_DATA_LIST = false;
95 if (!lex_match (lexer, T_ENDCMD))
96 return lex_end_of_command (lexer);
98 inp = xmalloc (sizeof *inp);
99 inp->session = session_create (dataset_session (ds));
100 inp->ds = dataset_create (inp->session, "INPUT PROGRAM");
101 inp->trns_chain = NULL;
105 inside_input_program = true;
106 while (!lex_match_phrase (lexer, "END INPUT PROGRAM"))
108 enum cmd_result result;
110 result = cmd_parse_in_state (lexer, inp->ds, CMD_STATE_INPUT_PROGRAM);
114 saw_DATA_LIST = true;
118 emit_END_CASE (inp->ds, inp);
130 if (cmd_result_is_failure (result)
131 && lex_get_error_mode (lexer) != LEX_ERROR_TERMINAL)
133 if (result == CMD_EOF)
134 msg (SE, _("Unexpected end-of-file within %s."), "INPUT PROGRAM");
135 inside_input_program = false;
136 destroy_input_program (inp);
142 emit_END_CASE (inp->ds, inp);
143 inside_input_program = false;
145 if (!saw_DATA_LIST && !saw_END_FILE)
147 msg (SE, _("Input program must contain %s or %s."), "DATA LIST", "END FILE");
148 destroy_input_program (inp);
151 if (dict_get_next_value_idx (dataset_dict (inp->ds)) == 0)
153 msg (SE, _("Input program did not create any variables."));
154 destroy_input_program (inp);
159 inp->trns_chain = proc_capture_transformations (ds);
161 inp->restart = TRNS_CONTINUE;
163 /* Figure out how to initialize each input case. */
164 inp->init = caseinit_create ();
165 caseinit_mark_for_init (inp->init, dataset_dict (inp->ds));
166 inp->proto = caseproto_ref (dict_get_proto (dataset_dict (inp->ds)));
168 dataset_set_dict (ds, dict_clone (dataset_dict (inp->ds)));
170 ds, casereader_create_sequential (NULL, inp->proto, CASENUMBER_MAX,
171 &input_program_casereader_class, inp));
177 cmd_end_input_program (struct lexer *lexer UNUSED, struct dataset *ds UNUSED)
179 /* Inside INPUT PROGRAM, this should get caught at the top of the loop in
182 Outside of INPUT PROGRAM, the command parser should reject this
187 /* Returns true if STATE is valid given the transformations that
188 are allowed within INPUT PROGRAM. */
190 is_valid_state (enum trns_result state)
192 return (state == TRNS_CONTINUE
193 || state == TRNS_ERROR
194 || state == TRNS_END_FILE
198 /* Reads and returns one case.
199 Returns the case if successful, null at end of file or if an
200 I/O error occurred. */
201 static struct ccase *
202 input_program_casereader_read (struct casereader *reader UNUSED, void *inp_)
204 struct input_program_pgm *inp = inp_;
205 struct ccase *c = case_create (inp->proto);
209 assert (is_valid_state (inp->restart));
210 if (inp->restart == TRNS_ERROR || inp->restart == TRNS_END_FILE)
216 c = case_unshare (c);
217 caseinit_init_vars (inp->init, c);
218 inp->restart = trns_chain_execute (inp->trns_chain, inp->restart,
220 assert (is_valid_state (inp->restart));
221 caseinit_update_left_vars (inp->init, c);
223 while (inp->restart < 0);
229 destroy_input_program (struct input_program_pgm *pgm)
233 session_destroy (pgm->session);
235 trns_chain_destroy (pgm->trns_chain);
236 caseinit_destroy (pgm->init);
237 caseproto_unref (pgm->proto);
242 /* Destroys the casereader. */
244 input_program_casereader_destroy (struct casereader *reader UNUSED, void *inp_)
246 struct input_program_pgm *inp = inp_;
247 if (inp->restart == TRNS_ERROR)
248 casereader_force_error (reader);
249 destroy_input_program (inp);
252 static const struct casereader_class input_program_casereader_class =
254 input_program_casereader_read,
255 input_program_casereader_destroy,
261 cmd_end_case (struct lexer *lexer, struct dataset *ds UNUSED)
263 assert (in_input_program ());
264 if (lex_token (lexer) == T_ENDCMD)
269 /* Outputs the current case */
271 end_case_trns_proc (void *inp_, struct ccase **c UNUSED,
272 casenumber case_nr UNUSED)
274 struct input_program_pgm *inp = inp_;
276 return TRNS_END_CASE;
279 /* REREAD transformation. */
282 struct dfm_reader *reader; /* File to move file pointer back on. */
283 struct expression *column; /* Column to reset file pointer to. */
286 /* Parses REREAD command. */
288 cmd_reread (struct lexer *lexer, struct dataset *ds)
290 struct file_handle *fh; /* File to be re-read. */
291 struct expression *e; /* Expression for column to set. */
292 struct reread_trns *t; /* Created transformation. */
293 char *encoding = NULL;
295 fh = fh_get_default_handle ();
297 while (lex_token (lexer) != T_ENDCMD)
299 if (lex_match_id (lexer, "COLUMN"))
301 lex_match (lexer, T_EQUALS);
305 lex_sbc_only_once ("COLUMN");
309 e = expr_parse (lexer, ds, EXPR_NUMBER);
313 else if (lex_match_id (lexer, "FILE"))
315 lex_match (lexer, T_EQUALS);
317 fh = fh_parse (lexer, FH_REF_FILE | FH_REF_INLINE, NULL);
321 else if (lex_match_id (lexer, "ENCODING"))
323 lex_match (lexer, T_EQUALS);
324 if (!lex_force_string (lexer))
328 encoding = ss_xstrdup (lex_tokss (lexer));
334 lex_error (lexer, NULL);
339 t = xmalloc (sizeof *t);
340 t->reader = dfm_open_reader (fh, lexer, encoding);
342 add_transformation (ds, reread_trns_proc, reread_trns_free, t);
351 return CMD_CASCADING_FAILURE;
354 /* Executes a REREAD transformation. */
356 reread_trns_proc (void *t_, struct ccase **c, casenumber case_num)
358 struct reread_trns *t = t_;
360 if (t->column == NULL)
361 dfm_reread_record (t->reader, 1);
364 double column = expr_evaluate_num (t->column, *c, case_num);
365 if (!isfinite (column) || column < 1)
367 msg (SE, _("REREAD: Column numbers must be positive finite "
368 "numbers. Column set to 1."));
369 dfm_reread_record (t->reader, 1);
372 dfm_reread_record (t->reader, column);
374 return TRNS_CONTINUE;
377 /* Frees a REREAD transformation.
378 Returns true if successful, false if an I/O error occurred. */
380 reread_trns_free (void *t_)
382 struct reread_trns *t = t_;
383 expr_free (t->column);
384 dfm_close_reader (t->reader);
388 /* Parses END FILE command. */
390 cmd_end_file (struct lexer *lexer UNUSED, struct dataset *ds)
392 assert (in_input_program ());
394 add_transformation (ds, end_file_trns_proc, NULL, NULL);
399 /* Executes an END FILE transformation. */
401 end_file_trns_proc (void *trns_ UNUSED, struct ccase **c UNUSED,
402 casenumber case_num UNUSED)
404 return TRNS_END_FILE;