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"
28 #include "data/data-in.h"
29 #include "data/data-out.h"
30 #include "data/format-guesser.h"
31 #include "data/casereader.h"
32 #include "data/gnumeric-reader.h"
33 #include "data/ods-reader.h"
34 #include "data/spreadsheet-reader.h"
35 #include "data/value-labels.h"
36 #include "language/data-io/data-parser.h"
37 #include "language/lexer/lexer.h"
38 #include "libpspp/assertion.h"
39 #include "libpspp/i18n.h"
40 #include "libpspp/line-reader.h"
41 #include "libpspp/message.h"
42 #include "ui/gui/dialog-common.h"
43 #include "ui/gui/executor.h"
44 #include "ui/gui/helper.h"
45 #include "ui/gui/builder-wrapper.h"
46 #include "ui/gui/psppire-data-window.h"
47 #include "ui/gui/psppire-dialog.h"
48 #include "ui/gui/psppire-encoding-selector.h"
49 #include "ui/gui/psppire-empty-list-store.h"
50 #include "ui/gui/psppire-var-sheet.h"
51 #include "ui/gui/psppire-scanf.h"
52 #include "ui/syntax-gen.h"
54 #include "gl/intprops.h"
55 #include "gl/xalloc.h"
58 #define _(msgid) gettext (msgid)
59 #define N_(msgid) msgid
61 struct import_assistant;
64 static char *choose_file (GtkWindow *parent_window, gchar **encodingp);
69 /* Obtains the file to import from the user and initializes IA's
70 file substructure. PARENT_WINDOW must be the window to use
71 as the file chooser window's parent.
73 Returns true if successful, false if the file name could not
74 be obtained or the file could not be read. */
76 init_file (struct import_assistant *ia, GtkWindow *parent_window)
78 enum { MAX_LINE_LEN = 16384 }; /* Max length of an acceptable line. */
79 struct file *file = &ia->file;
82 file->file_name = choose_file (parent_window, &file->encoding);
83 if (file->file_name == NULL)
86 if (ia->spreadsheet == NULL)
87 ia->spreadsheet = gnumeric_probe (file->file_name, false);
89 if (ia->spreadsheet == NULL)
90 ia->spreadsheet = ods_probe (file->file_name, false);
92 if (ia->spreadsheet == NULL)
95 struct line_reader *reader = line_reader_for_file (file->encoding, file->file_name, O_RDONLY);
98 msg_error (errno, _("Could not open `%s'"),
103 ds_init_empty (&input);
104 file->lines = xnmalloc (MAX_PREVIEW_LINES, sizeof *file->lines);
105 for (; file->line_cnt < MAX_PREVIEW_LINES; file->line_cnt++)
108 if (!line_reader_read (reader, &input, MAX_LINE_LEN + 1)
109 || ds_length (&input) > MAX_LINE_LEN)
111 if (line_reader_eof (reader))
113 else if (line_reader_error (reader))
114 msg (ME, _("Error reading `%s': %s"),
115 file->file_name, strerror (line_reader_error (reader)));
117 msg (ME, _("Failed to read `%s', because it contains a line "
118 "over %d bytes long and therefore appears not to be "
120 file->file_name, MAX_LINE_LEN);
121 line_reader_close (reader);
127 ds_init_cstr (&file->lines[file->line_cnt],
128 recode_string ("UTF-8", line_reader_get_encoding (reader),
129 ds_cstr (&input), ds_length (&input)));
133 if (file->line_cnt == 0)
135 msg (ME, _("`%s' is empty."), file->file_name);
136 line_reader_close (reader);
141 /* Estimate the number of lines in the file. */
142 if (file->line_cnt < MAX_PREVIEW_LINES)
143 file->total_lines = file->line_cnt;
147 off_t position = line_reader_tell (reader);
148 if (fstat (line_reader_fileno (reader), &s) == 0 && position > 0)
149 file->total_lines = (double) file->line_cnt / position * s.st_size;
151 file->total_lines = 0;
154 line_reader_close (reader);
160 /* Frees IA's file substructure. */
162 destroy_file (struct import_assistant *ia)
164 struct file *f = &ia->file;
169 for (i = 0; i < f->line_cnt; i++)
170 ds_destroy (&f->lines[i]);
174 g_free (f->file_name);
175 g_free (f->encoding);
178 /* Obtains the file to read from the user. If successful, returns the name of
179 the file and stores the user's chosen encoding for the file into *ENCODINGP.
180 The caller must free each of these strings with g_free().
182 On failure, stores a null pointer and stores NULL in *ENCODINGP.
184 PARENT_WINDOW must be the window to use as the file chooser window's
187 choose_file (GtkWindow *parent_window, gchar **encodingp)
190 GtkFileFilter *filter = NULL;
192 GtkWidget *dialog = gtk_file_chooser_dialog_new (_("Import Delimited Text Data"),
194 GTK_FILE_CHOOSER_ACTION_OPEN,
195 GTK_STOCK_CANCEL, GTK_RESPONSE_CANCEL,
196 GTK_STOCK_OPEN, GTK_RESPONSE_ACCEPT,
199 g_object_set (dialog, "local-only", FALSE, NULL);
201 filter = gtk_file_filter_new ();
202 gtk_file_filter_set_name (filter, _("Text Files"));
203 gtk_file_filter_add_mime_type (filter, "text/*");
204 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
206 filter = gtk_file_filter_new ();
207 gtk_file_filter_set_name (filter, _("Text (*.txt) Files"));
208 gtk_file_filter_add_pattern (filter, "*.txt");
209 gtk_file_filter_add_pattern (filter, "*.TXT");
210 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
212 filter = gtk_file_filter_new ();
213 gtk_file_filter_set_name (filter, _("Plain Text (ASCII) Files"));
214 gtk_file_filter_add_mime_type (filter, "text/plain");
215 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
217 filter = gtk_file_filter_new ();
218 gtk_file_filter_set_name (filter, _("Comma Separated Value Files"));
219 gtk_file_filter_add_mime_type (filter, "text/csv");
220 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
222 /* I've never encountered one of these, but it's listed here:
223 http://www.iana.org/assignments/media-types/text/tab-separated-values */
224 filter = gtk_file_filter_new ();
225 gtk_file_filter_set_name (filter, _("Tab Separated Value Files"));
226 gtk_file_filter_add_mime_type (filter, "text/tab-separated-values");
227 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
229 filter = gtk_file_filter_new ();
230 gtk_file_filter_set_name (filter, _("Gnumeric Spreadsheet Files"));
231 gtk_file_filter_add_mime_type (filter, "application/x-gnumeric");
232 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
234 filter = gtk_file_filter_new ();
235 gtk_file_filter_set_name (filter, _("OpenDocument Spreadsheet Files"));
236 gtk_file_filter_add_mime_type (filter, "application/vnd.oasis.opendocument.spreadsheet");
237 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
239 filter = gtk_file_filter_new ();
240 gtk_file_filter_set_name (filter, _("All Spreadsheet Files"));
241 gtk_file_filter_add_mime_type (filter, "application/x-gnumeric");
242 gtk_file_filter_add_mime_type (filter, "application/vnd.oasis.opendocument.spreadsheet");
243 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
245 gtk_file_chooser_set_extra_widget (
246 GTK_FILE_CHOOSER (dialog), psppire_encoding_selector_new ("Auto", true));
248 filter = gtk_file_filter_new ();
249 gtk_file_filter_set_name (filter, _("All Files"));
250 gtk_file_filter_add_pattern (filter, "*");
251 gtk_file_chooser_add_filter (GTK_FILE_CHOOSER (dialog), filter);
253 switch (gtk_dialog_run (GTK_DIALOG (dialog)))
255 case GTK_RESPONSE_ACCEPT:
256 file_name = gtk_file_chooser_get_filename (GTK_FILE_CHOOSER (dialog));
257 *encodingp = psppire_encoding_selector_get_encoding (
258 gtk_file_chooser_get_extra_widget (GTK_FILE_CHOOSER (dialog)));
265 gtk_widget_destroy (dialog);