Bug 332878 – DBF import creates non-UTF8 strings

After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.

Bug 332878 - DBF import creates non-UTF8 strings


Summary:	DBF import creates non-UTF8 strings


Status:	RESOLVED FIXED

Product:	Gnumeric
Classification:	Applications
Component:	General
Version:	1.6.x
Hardware:	Other All

Importance:	Normal critical
Target Milestone:	---
Assigned To:	Jody Goldberg
QA Contact:	Jody Goldberg

URL:
Whiteboard:

Depends on:
Blocks:

Reported:	2006-02-28 18:43 UTC by Karsten Bräckelmann
Modified:	2006-03-02 02:44 UTC

See Also:
GNOME target:	---
GNOME version:	---

Description Karsten Bräckelmann 2006-02-28 18:43:28 UTC

Gnumeric just crashed for me on searching a dBase dbf table for a given search string.

In case it matters, there is an associated dbt file as well. Don't know, if it is used at all here.

gnumeric 1.6.2, GNOME 1.13.91+
Including the crashing top-most thread only.


Backtrace was generated from '/opt/gnome-2.13/bin/gnumeric'

Using host libthread_db library "/lib/tls/libthread_db.so.1".
[Thread debugging using libthread_db enabled]
[New Thread -1228130624 (LWP 9910)]
0xffffe410 in ?? ()

+ Trace 66577

#0 ??
#1 ??
#2 ??
#3 ??
#4 waitpid
from /lib/tls/libc.so.6
#5 libgnomeui_segv_handle
at gnome-ui-init.c line 792
#6 <signal handler called>
#7 IA__g_logv
at gmessages.c line 502
#8 IA__g_log
at gmessages.c line 517
#9 IA__g_malloc
at gmem.c line 135
#10 compute_value
at number-match.c line 155
#11 format_match
at number-match.c line 683
#12 format_match_number
at number-match.c line 718
#13 cell_get_entered_text
at rendered-value.c line 605
#14 gnm_search_replace_cell
at search.c line 312
#15 search_filter_matching
at search.c line 208
#16 search_clicked
at dialog-search.c line 360
#17 IA__g_cclosure_marshal_VOID__VOID
at gmarshal.c line 77
#18 IA__g_closure_invoke
at gclosure.c line 490
#19 signal_emit_unlocked_R
at gsignal.c line 2438
#20 IA__g_signal_emit_valist
at gsignal.c line 2197
#21 IA__g_signal_emit
at gsignal.c line 2241
#22 IA__gtk_button_clicked
at gtkbutton.c line 834
#23 gtk_button_finish_activate
at gtkbutton.c line 1442
#24 gtk_button_key_release
at gtkbutton.c line 1299
#25 _gtk_marshal_BOOLEAN__BOXED
at gtkmarshalers.c line 83
#26 g_type_class_meta_marshal
at gclosure.c line 567
#27 IA__g_closure_invoke
at gclosure.c line 490
#28 signal_emit_unlocked_R
at gsignal.c line 2476
#29 IA__g_signal_emit_valist
at gsignal.c line 2207
#30 IA__g_signal_emit
at gsignal.c line 2241
#31 gtk_widget_event_internal
at gtkwidget.c line 3735
#32 IA__gtk_propagate_event
at gtkmain.c line 2141
#33 IA__gtk_main_do_event
at gtkmain.c line 1412
#34 gdk_event_dispatch
at gdkevents-x11.c line 2291
#35 IA__g_main_context_dispatch
at gmain.c line 1916
#36 g_main_context_iterate
at gmain.c line 2547
#37 IA__g_main_loop_run
at gmain.c line 2751
#38 bonobo_main
at bonobo-main.c line 312
#39 main
at main-application.c line 483

Comment 1 Karsten Bräckelmann 2006-02-28 18:45:23 UTC

Oh, before anyone asks:  No, I can not attach this database. But I may be able to come up with a stripped down test case, if required and I find the time to do so... ;)

Comment 2 Morten Welinder 2006-02-28 20:45:13 UTC

This looks like a search/replace, not a search.  Right?

The cell that contains problems contains "Lindenstra[ss]e 3 A".  It would be
helpful if you could create a dbf file with just that in it and attach it here.

Current guessesw as to the cause:

1. dbf input creates non-UTF8 strings.
-or-
2. A bug in libpcre.

Comment 3 Morten Welinder 2006-03-01 17:43:17 UTC

Also, if you run from the command line, do you get any error messages?

Comment 4 Morten Welinder 2006-03-01 18:48:54 UTC

Never mind, this is surely an import issue and I see problems with files
like http://www.preagro.de/Ergebnisse/codepage/Gemarkungen_after.dbf

Fixed in cvs.

Comment 5 Karsten Bräckelmann 2006-03-02 01:43:47 UTC

May I ask which release this will be in?

I don't have a lot of time to create a test case with stripped data, but I sure can easily test this with the same data again.

FWIW, since this database originally was created years ago using Win98 or maybe even the ancient DOS app, char encoding issues seem very likely. :)

Comment 6 Morten Welinder 2006-03-02 02:44:19 UTC

It'll be in 1.6.3 if we ever make one.
It'll be in 1.7.0 [unstable].