1 /* PSPP - a program for statistical analysis.
2 Copyright (C) 2010, 2011, 2012, 2013, 2014 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/>. */
19 #include "data/csv-file-writer.h"
29 #include "data/calendar.h"
30 #include "data/case.h"
31 #include "data/casewriter-provider.h"
32 #include "data/casewriter.h"
33 #include "data/data-out.h"
34 #include "data/dictionary.h"
35 #include "data/file-handle-def.h"
36 #include "data/format.h"
37 #include "data/make-file.h"
38 #include "data/missing-values.h"
39 #include "data/settings.h"
40 #include "data/value-labels.h"
41 #include "data/variable.h"
42 #include "libpspp/assertion.h"
43 #include "libpspp/i18n.h"
44 #include "libpspp/message.h"
45 #include "libpspp/str.h"
47 #include "gl/ftoastr.h"
48 #include "gl/minmax.h"
49 #include "gl/unlocked-io.h"
50 #include "gl/xalloc.h"
53 #define _(msgid) gettext (msgid)
54 #define N_(msgid) (msgid)
56 /* A variable in a CSV file. */
59 int width; /* Variable width (0 to 32767). */
60 int case_index; /* Index into case. */
61 struct fmt_spec format; /* Print format. */
62 struct missing_values missing; /* User-missing values, if recoding. */
63 struct val_labs *val_labs; /* Value labels, if any and they are in use. */
66 /* Comma-separated value (CSV) file writer. */
69 struct file_handle *fh; /* File handle. */
70 struct fh_lock *lock; /* Mutual exclusion for file. */
71 FILE *file; /* File stream. */
72 struct replace_file *rf; /* Ticket for replacing output file. */
74 struct csv_writer_options opts;
76 char *encoding; /* Encoding used by variables. */
79 struct csv_var *csv_vars; /* Variables. */
80 size_t n_csv_vars; /* Number of variables. */
83 static const struct casewriter_class csv_file_casewriter_class;
85 static void write_var_names (struct csv_writer *, const struct dictionary *);
87 static bool write_error (const struct csv_writer *);
88 static bool close_writer (struct csv_writer *);
90 /* Initializes OPTS with default options for writing a CSV file. */
92 csv_writer_options_init (struct csv_writer_options *opts)
94 opts->recode_user_missing = false;
95 opts->include_var_names = false;
96 opts->use_value_labels = false;
97 opts->use_print_formats = false;
98 opts->decimal = settings_get_decimal_char (FMT_F);
99 opts->delimiter = ',';
100 opts->qualifier = '"';
103 /* Opens the CSV file designated by file handle FH for writing cases from
104 dictionary DICT according to the given OPTS.
106 No reference to D is retained, so it may be modified or
107 destroyed at will after this function returns. */
109 csv_writer_open (struct file_handle *fh, const struct dictionary *dict,
110 const struct csv_writer_options *opts)
112 struct csv_writer *w;
115 /* Create and initialize writer. */
116 w = xmalloc (sizeof *w);
124 w->encoding = xstrdup (dict_get_encoding (dict));
126 w->n_csv_vars = dict_get_var_cnt (dict);
127 w->csv_vars = xnmalloc (w->n_csv_vars, sizeof *w->csv_vars);
128 for (i = 0; i < w->n_csv_vars; i++)
130 const struct variable *var = dict_get_var (dict, i);
131 struct csv_var *cv = &w->csv_vars[i];
133 cv->width = var_get_width (var);
134 cv->case_index = var_get_case_index (var);
136 cv->format = *var_get_print_format (var);
137 if (opts->recode_user_missing)
138 mv_copy (&cv->missing, var_get_missing_values (var));
140 mv_init (&cv->missing, cv->width);
142 if (opts->use_value_labels)
143 cv->val_labs = val_labs_clone (var_get_value_labels (var));
148 /* Open file handle as an exclusive writer. */
149 /* TRANSLATORS: this fragment will be interpolated into messages in fh_lock()
150 that identify types of files. */
151 w->lock = fh_lock (fh, FH_REF_FILE, N_("CSV file"), FH_ACC_WRITE, true);
155 /* Create the file on disk. */
156 w->rf = replace_file_start (fh, "w", 0666, &w->file);
159 msg (ME, _("Error opening `%s' for writing as a system file: %s."),
160 fh_get_file_name (fh), strerror (errno));
164 if (opts->include_var_names)
165 write_var_names (w, dict);
170 return casewriter_create (dict_get_proto (dict),
171 &csv_file_casewriter_class, w);
179 csv_field_needs_quoting (struct csv_writer *w, const char *s, size_t len)
183 for (p = s; p < &s[len]; p++)
184 if (*p == w->opts.qualifier || *p == w->opts.delimiter
185 || *p == '\n' || *p == '\r')
192 csv_output_buffer (struct csv_writer *w, const char *s, size_t len)
194 if (csv_field_needs_quoting (w, s, len))
198 putc (w->opts.qualifier, w->file);
199 for (p = s; p < &s[len]; p++)
201 /* We are writing the output file in text mode, so transform any
202 explicit CR-LF line breaks into LF only, to allow the C library to
203 use correct system-specific new-lines. */
204 if (*p == '\r' && p[1] == '\n')
207 if (*p == w->opts.qualifier)
208 putc (w->opts.qualifier, w->file);
211 putc (w->opts.qualifier, w->file);
214 fwrite (s, 1, len, w->file);
218 csv_output_string (struct csv_writer *w, const char *s)
220 csv_output_buffer (w, s, strlen (s));
224 write_var_names (struct csv_writer *w, const struct dictionary *d)
228 for (i = 0; i < w->n_csv_vars; i++)
231 putc (w->opts.delimiter, w->file);
232 csv_output_string (w, var_get_name (dict_get_var (d, i)));
234 putc ('\n', w->file);
238 csv_output_format (struct csv_writer *w, const struct csv_var *cv,
239 const union value *value)
241 char *s = data_out (value, w->encoding, &cv->format);
242 struct substring ss = ss_cstr (s);
243 if (cv->format.type != FMT_A)
244 ss_trim (&ss, ss_cstr (" "));
246 ss_rtrim (&ss, ss_cstr (" "));
247 csv_output_buffer (w, ss.string, ss.length);
252 extract_date (double number, int *y, int *m, int *d)
256 calendar_offset_to_gregorian (number / 60. / 60. / 24., y, m, d, &yd);
257 return fmod (number, 60. * 60. * 24.);
261 extract_time (double number, double *H, int *M, int *S)
263 *H = floor (number / 60. / 60.);
264 number = fmod (number, 60. * 60.);
266 *M = floor (number / 60.);
267 number = fmod (number, 60.);
273 csv_write_var__ (struct csv_writer *w, const struct csv_var *cv,
274 const union value *value)
278 label = val_labs_find (cv->val_labs, value);
280 csv_output_string (w, label);
281 else if (cv->width == 0 && value->f == SYSMIS)
282 csv_output_buffer (w, " ", 1);
283 else if (w->opts.use_print_formats)
284 csv_output_format (w, cv, value);
287 char s[MAX (DBL_STRLEN_BOUND, 128)];
290 switch (cv->format.type)
314 dtoastr (s, sizeof s, 0, 0, value->f);
315 cp = strpbrk (s, ".,");
317 *cp = w->opts.decimal;
334 extract_date (value->f, &y, &m, &d);
335 snprintf (s, sizeof s, "%02d/%02d/%04d", m, d, y);
348 extract_time (extract_date (value->f, &y, &m, &d), &H, &M, &S);
349 snprintf (s, sizeof s, "%02d/%02d/%04d %02.0f:%02d:%02d",
361 extract_time (fabs (value->f), &H, &M, &S);
362 snprintf (s, sizeof s, "%s%02.0f:%02d:%02d",
363 value->f < 0 ? "-" : "", H, M, S);
369 csv_output_format (w, cv, value);
372 case FMT_NUMBER_OF_FORMATS:
375 csv_output_string (w, s);
380 csv_write_var (struct csv_writer *w, const struct csv_var *cv,
381 const union value *value)
383 if (mv_is_value_missing (&cv->missing, value, MV_USER))
387 value_init (&missing, cv->width);
388 value_set_missing (&missing, cv->width);
389 csv_write_var__ (w, cv, &missing);
390 value_destroy (&missing, cv->width);
393 csv_write_var__ (w, cv, value);
397 csv_write_case (struct csv_writer *w, const struct ccase *c)
401 for (i = 0; i < w->n_csv_vars; i++)
403 const struct csv_var *cv = &w->csv_vars[i];
406 putc (w->opts.delimiter, w->file);
407 csv_write_var (w, cv, case_data_idx (c, cv->case_index));
409 putc ('\n', w->file);
412 /* Writes case C to CSV file W. */
414 csv_file_casewriter_write (struct casewriter *writer, void *w_,
417 struct csv_writer *w = w_;
419 if (ferror (w->file))
421 casewriter_force_error (writer);
426 csv_write_case (w, c);
430 /* Destroys CSV file writer W. */
432 csv_file_casewriter_destroy (struct casewriter *writer, void *w_)
434 struct csv_writer *w = w_;
435 if (!close_writer (w))
436 casewriter_force_error (writer);
439 /* Returns true if an I/O error has occurred on WRITER, false otherwise. */
441 write_error (const struct csv_writer *writer)
443 return ferror (writer->file);
446 /* Closes a CSV file after we're done with it.
447 Returns true if successful, false if an I/O error occurred. */
449 close_writer (struct csv_writer *w)
462 if (fclose (w->file) == EOF)
466 msg (ME, _("An I/O error occurred writing CSV file `%s'."),
467 fh_get_file_name (w->fh));
469 if (ok ? !replace_file_commit (w->rf) : !replace_file_abort (w->rf))
478 for (i = 0; i < w->n_csv_vars; i++)
480 struct csv_var *cv = &w->csv_vars[i];
481 mv_destroy (&cv->missing);
482 val_labs_destroy (cv->val_labs);
491 /* CSV file writer casewriter class. */
492 static const struct casewriter_class csv_file_casewriter_class =
494 csv_file_casewriter_write,
495 csv_file_casewriter_destroy,