Bug 143368

Summary: abiword - various CRITICAL messages
Product: [Fedora] Fedora Reporter: Michal Jaegermann <michal>
Component: abiwordAssignee: Caolan McNamara <caolanm>
Status: CLOSED RAWHIDE QA Contact: Mike McLean <mikem>
Severity: medium Docs Contact:
Priority: medium    
Version: rawhide   
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2005-01-14 08:49:00 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Michal Jaegermann 2004-12-20 01:06:08 UTC
Description of problem:

When trying to open various documents (RTF or M$-word produced
files) with abiword-2.2.2-1 I am getting alerts like that:

Could not load the dictionary for the English (Canada)[en-CA] language.
Could not load the dictionary for the English (US)[en-US] language.

Sometimes even both for the same document.  aspell-en-0.51-11.x86_64
is installed on the test system.  What abiword would like to
open it is somewhat hard to find out as

   strace abiword <some_document>

just sits at the initial 'execve("/usr/bin/abiword",  .... ) = 0'
and 'gdb abiword' never gets to a '(gdb)' prompt.  This with
"SELINUX=disabled" in /etc/sysconfig/selinux before will anybody ask.

OTOH on a console abiword is all the time producing messages like
these:

I've been notified!! Text is |Introduction|
CP !<CluePacket>
    <Frontend>Abiword</Frontend>
    <Context>Abiword</Context>
    <Focused>true</Focused>
    <Clue Type="textblock" Relevance="10">Introduction</Clue>
</CluePacket>
!
Dashboard: Sending cluepacket...
Dashboard: connect: Connection refused
Illegal format hash table
1
(AbiWord-2.2:5205): Gtk-CRITICAL **: file gtkfilesystemunix.c: line
793 (gtk_file_system_unix_get_parent): assertion `g_path_is_absolute
(filename)' failed

(AbiWord-2.2:5205): libgsf:msole-CRITICAL **: file gsf-infile-msole.c:
line 113 (ole_get_block): assertion `block < ole->info->max_block'
failed

On occasions in these <Clue ...> ... </Clue> brackets one gets
quite sizeable chunks of a text.

Once past all these hurdles a text in an editor window seems
to look as usual.

This is marked 'x86_64' for now as I simply do not know if
this happens on x86 too or not.

Version-Release number of selected component (if applicable):
abiword-2.2.2-1

How reproducible:
Every time

Comment 1 Michal Jaegermann 2004-12-20 04:46:29 UTC
strace and gdb bit looks like a kernel problem (see bug #143350).
With that resolved from traces follows that AbiWord looks at the 
content of
/usr/share/AbiSuite-2.2/dictionary/ispell_dictionary_list.xml
and from that it seems that indeed it expects dictionaries in
a format used by ispell.  Small wonder that it cannot find anything
as only aspell dictionaries, in a diffrent format, are
provided on a system; with an apparent exception of a single
/usr/share/AbiSuite-2.2/dictionary/american.hash  and
which is not found anyway.  Wrong encoding?

It appears that an adjustment to use aspell dictionaries is necessary.
It is also possibly unrelated to other problems.

Comment 2 Caolan McNamara 2004-12-20 13:02:13 UTC
Hmm, abiword really wants enchant to be installed thesedays... Might
package enchant and configure enchant to use aspell and then problem
should go away.

Comment 3 Caolan McNamara 2005-01-11 11:49:50 UTC
Should work in 2.2.2-2 coming soon

Comment 4 Caolan McNamara 2005-01-13 10:07:27 UTC
dictionary warning should be solved in available rawhide abiword-2.2.2-2

Comment 5 Michal Jaegermann 2005-01-13 18:46:22 UTC
Dictionary warnings are indeed gone but all these 

CP !<CluePacket>
.....
Dashboard: Sending cluepacket...
Dashboard: connect: Connection refused

with sizeable chunks of text in them, are all over the place.
abiword-2.2.2-3.

If the above is indeed a garbage (but what this "Connection refused"
is supposed to mean?) then why not turn that off?


Comment 6 Caolan McNamara 2005-01-14 08:49:00 UTC
I'll handle the cluepacket problems as RH#145085# as its unrelated to
the spell checking error messages.