1 /* PSPPIRE - a graphical user interface for PSPP.
2 Copyright (C) 2008, 2009, 2010, 2011, 2012, 2013 Free Software Foundation
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 "ui/gui/text-data-import-dialog.h"
23 #include <gtk-contrib/psppire-sheet.h>
29 #include "data/data-in.h"
30 #include "data/data-out.h"
31 #include "data/format-guesser.h"
32 #include "data/casereader.h"
33 #include "data/gnumeric-reader.h"
34 #include "data/ods-reader.h"
35 #include "data/spreadsheet-reader.h"
36 #include "data/value-labels.h"
37 #include "language/data-io/data-parser.h"
38 #include "language/lexer/lexer.h"
39 #include "libpspp/assertion.h"
40 #include "libpspp/i18n.h"
41 #include "libpspp/line-reader.h"
42 #include "libpspp/message.h"
43 #include "ui/gui/checkbox-treeview.h"
44 #include "ui/gui/dialog-common.h"
45 #include "ui/gui/executor.h"
46 #include "ui/gui/helper.h"
47 #include "ui/gui/builder-wrapper.h"
48 #include "ui/gui/psppire-data-window.h"
49 #include "ui/gui/psppire-dialog.h"
50 #include "ui/gui/psppire-encoding-selector.h"
51 #include "ui/gui/psppire-empty-list-store.h"
52 #include "ui/gui/psppire-var-sheet.h"
53 #include "ui/gui/psppire-var-store.h"
54 #include "ui/gui/psppire-scanf.h"
55 #include "ui/syntax-gen.h"
58 #include "gl/intprops.h"
59 #include "gl/xalloc.h"
62 #define _(msgid) gettext (msgid)
63 #define N_(msgid) msgid
65 struct import_assistant;
68 static char *choose_file (GtkWindow *parent_window, gchar **encodingp);
69 enum { MAX_PREVIEW_LINES = 1000 }; /* Max number of lines to read. */
73 Update IA according to the contents of DICT and CREADER.
74 CREADER will be destroyed by this function.
77 update_assistant (struct import_assistant *ia)
79 struct sheet_spec_page *ssp = ia->sheet_spec;
81 struct file *file = &ia->file;
82 struct separators_page *sepp = ia->separators;
90 sepp->column_cnt = dict_get_var_cnt (ssp->dict);
91 sepp->columns = xcalloc (sepp->column_cnt, sizeof (*sepp->columns));
92 for (col = 0; col < sepp->column_cnt ; ++col)
94 const struct variable *var = dict_get_var (ssp->dict, col);
95 sepp->columns[col].name = xstrdup (var_get_name (var));
96 sepp->columns[col].contents = NULL;
99 for (; (c = casereader_read (ssp->reader)) != NULL; case_unref (c))
102 for (col = 0; col < sepp->column_cnt ; ++col)
105 const struct variable *var = dict_get_var (ssp->dict, col);
107 sepp->columns[col].contents = xrealloc (sepp->columns[col].contents,
108 sizeof (struct substring) * rows);
110 ss = data_out (case_data (c, var), dict_get_encoding (ssp->dict),
111 var_get_print_format (var));
113 sepp->columns[col].contents[rows - 1] = ss_cstr (ss);
116 if (rows > MAX_PREVIEW_LINES)
124 file->line_cnt = rows;
128 /* Obtains the file to import from the user and initializes IA's
129 file substructure. PARENT_WINDOW must be the window to use
130 as the file chooser window's parent.
132 Returns true if successful, false if the file name could not
133 be obtained or the file could not be read. */
135 init_file (struct import_assistant *ia, GtkWindow *parent_window)
137 enum { MAX_LINE_LEN = 16384 }; /* Max length of an acceptable line. */
138 struct file *file = &ia->file;
139 struct sheet_spec_page *ssp = ia->sheet_spec;
140 struct spreadsheet_read_info sri;
141 struct spreadsheet_read_options opts;
144 file->file_name = choose_file (parent_window, &file->encoding);
145 if (file->file_name == NULL)
148 opts.sheet_name = NULL;
149 opts.cell_range = NULL;
150 opts.sheet_index = 1;
152 sri.file_name = file->file_name;
153 sri.read_names = true;
156 if (ssp->spreadsheet == NULL)
157 ssp->spreadsheet = gnumeric_probe (sri.file_name);
159 if (ssp->spreadsheet == NULL)
160 ssp->spreadsheet = ods_probe (sri.file_name);
162 if (ssp->spreadsheet)
165 // update_assistant (ia);
170 struct line_reader *reader = line_reader_for_file (file->encoding, file->file_name, O_RDONLY);
173 msg (ME, _("Could not open `%s': %s"),
174 file->file_name, strerror (errno));
178 ds_init_empty (&input);
179 file->lines = xnmalloc (MAX_PREVIEW_LINES, sizeof *file->lines);
180 for (; file->line_cnt < MAX_PREVIEW_LINES; file->line_cnt++)
183 if (!line_reader_read (reader, &input, MAX_LINE_LEN + 1)
184 || ds_length (&input) > MAX_LINE_LEN)
186 if (line_reader_eof (reader))
188 else if (line_reader_error (reader))
189 msg (ME, _("Error reading `%s': %s"),
190 file->file_name, strerror (line_reader_error (reader)));
192 msg (ME, _("Failed to read `%s', because it contains a line "
193 "over %d bytes long and therefore appears not to be "
195 file->file_name, MAX_LINE_LEN);
196 line_reader_close (reader);
202 ds_init_cstr (&file->lines[file->line_cnt],
203 recode_string ("UTF-8", line_reader_get_encoding (reader),
204 ds_cstr (&input), ds_length (&input)));
208 if (file->line_cnt == 0)
210 msg (ME, _("`%s' is empty."), file->file_name);
211 line_reader_close (reader);
216 /* Estimate the number of lines in the file. */
217 if (file->line_cnt < MAX_PREVIEW_LINES)
218 file->total_lines = file->line_cnt;
222 off_t position = line_reader_tell (reader);
223 if (fstat (line_reader_fileno (reader), &s) == 0 && position > 0)
224 file->total_lines = (double) file->line_cnt / position * s.st_size;
226 file->total_lines = 0;
229 line_reader_close (reader);
235 /* Frees IA's file substructure. */
237 destroy_file (struct import_assistant *ia)
239 struct file *f = &ia->file;
244 for (i = 0; i < f->line_cnt; i++)
245 ds_destroy (&f->lines[i]);
249 g_free (f->file_name);
250 g_free (f->encoding);
253 /* Obtains the file to read from the user. If successful, returns the name of
254 the file and stores the user's chosen encoding for the file into *ENCODINGP.
255 The caller must free each of these strings with g_free().
257 On failure, stores a null pointer and stores NULL in *ENCODINGP.
259 PARENT_WINDOW must be the window to use as the file chooser window's
262 choose_file (GtkWindow *parent_window, gchar **encodingp)
265 GtkFileFilter *filter = NULL;
267 GtkWidget *dialog = gtk_file_chooser_dialog_new (_("Import Delimited Text Data"),
269 GTK_FILE_CHOOSER_ACTION_OPEN,
270 GTK_STOCK_CANCEL, GTK_RESPONSE_CANCEL,
271 GTK_STOCK_OPEN, GTK_RESPONSE_ACCEPT,
274 g_object_set (dialog, "local-only", FALSE, NULL);
276 filter = gtk_file_filter_new ();
277 gtk_file_filter_set_name (filter, _("Text files"));
278 gtk_file_filter_add_mime_type (filter, "text/*");
279 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
281 filter = gtk_file_filter_new ();
282 gtk_file_filter_set_name (filter, _("Text (*.txt) Files"));
283 gtk_file_filter_add_pattern (filter, "*.txt");
284 gtk_file_filter_add_pattern (filter, "*.TXT");
285 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
287 filter = gtk_file_filter_new ();
288 gtk_file_filter_set_name (filter, _("Plain Text (ASCII) Files"));
289 gtk_file_filter_add_mime_type (filter, "text/plain");
290 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
292 filter = gtk_file_filter_new ();
293 gtk_file_filter_set_name (filter, _("Comma Separated Value Files"));
294 gtk_file_filter_add_mime_type (filter, "text/csv");
295 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
297 /* I've never encountered one of these, but it's listed here:
298 http://www.iana.org/assignments/media-types/text/tab-separated-values */
299 filter = gtk_file_filter_new ();
300 gtk_file_filter_set_name (filter, _("Tab Separated Value Files"));
301 gtk_file_filter_add_mime_type (filter, "text/tab-separated-values");
302 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
304 filter = gtk_file_filter_new ();
305 gtk_file_filter_set_name (filter, _("Gnumeric Spreadsheet Files"));
306 gtk_file_filter_add_mime_type (filter, "application/x-gnumeric");
307 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
309 filter = gtk_file_filter_new ();
310 gtk_file_filter_set_name (filter, _("OpenOffice.Org Spreadsheet Files"));
311 gtk_file_filter_add_mime_type (filter, "application/vnd.oasis.opendocument.spreadsheet");
312 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
314 filter = gtk_file_filter_new ();
315 gtk_file_filter_set_name (filter, _("All Spreadsheet Files"));
316 gtk_file_filter_add_mime_type (filter, "application/x-gnumeric");
317 gtk_file_filter_add_mime_type (filter, "application/vnd.oasis.opendocument.spreadsheet");
318 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
321 gtk_file_chooser_set_extra_widget (
322 GTK_FILE_CHOOSER (dialog), psppire_encoding_selector_new ("Auto", true));
324 filter = gtk_file_filter_new ();
325 gtk_file_filter_set_name (filter, _("All Files"));
326 gtk_file_filter_add_pattern (filter, "*");
327 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
329 switch (gtk_dialog_run (GTK_DIALOG (dialog)))
331 case GTK_RESPONSE_ACCEPT:
332 file_name = gtk_file_chooser_get_filename (GTK_FILE_CHOOSER (dialog));
333 *encodingp = psppire_encoding_selector_get_encoding (
334 gtk_file_chooser_get_extra_widget (GTK_FILE_CHOOSER (dialog)));
341 gtk_widget_destroy (dialog);