From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.0.6) Gecko/20060808 Fedora/1.5.0.6-3 Firefox/1.5.0.6 pango-text Description of problem: The beagle-search GUI crashes, when words are typed into the search field. Version-Release number of selected component (if applicable): beagle-0.2.8-1.fc6 How reproducible: Always Steps to Reproduce: 1. Run beagle-search from a terminal as a normal user. 2. Begin typing in the search field. Actual Results: ]$ beagle-search Unhandled Exception: System.ArgumentNullException: null key Parameter name: key at System.Collections.Hashtable.get_Item (object) <0x0017a> at Search.Tiles.IMLog.LoadIcon (Gtk.Image,int) <0x000cf> at Search.Tiles.TileFlat.OnRealized () <0x00071> at Gtk.Widget.realized_cb (intptr) <0x00036> at (wrapper native-to-managed) Gtk.Widget.realized_cb (intptr) <0x00032> in (unmanaged) 0x4331b138 at (wrapper managed-to-native) Gtk.Widget.gtk_widget_set_parent (intptr,intptr) <0x00004> at Gtk.Widget.set_Parent (Gtk.Widget) <0x00037> at Search.Category.OnAdded (Gtk.Widget) <0x00068> at Search.ConversationCategory.OnAdded (Gtk.Widget) <0x00017> at Gtk.Container.added_cb (intptr,intptr) <0x00063> at (wrapper native-to-managed) Gtk.Container.added_cb (intptr,intptr) <0x00035> in (unmanaged) 0x4331a648 at (wrapper managed-to-native) Gtk.Container.gtk_container_add (intptr,intptr) <0x00004> at Gtk.Container.Add (Gtk.Widget) <0x00037> at Search.GroupView.AddHit (Search.Tiles.Tile) <0x000d2> at Search.MainWindow.OnHitsAdded (Beagle.HitsAddedResponse) <0x000b7> at (wrapper delegate-invoke) System.MulticastDelegate.invoke_void_HitsAddedResponse (Beagle.HitsAddedResponse) <0x0003b> at Beagle.Query.OnHitsAdded (Beagle.ResponseMessage) <0x0003a> at (wrapper delegate-invoke) System.MulticastDelegate.invoke_void_ResponseMessage (Beagle.ResponseMessage) <0x0003b> at Beagle.RequestMessage.OnAsyncResponse (Beagle.ResponseMessage) <0x00042> at (wrapper delegate-invoke) System.MulticastDelegate.invoke_void_ResponseMessage (Beagle.ResponseMessage) <0x0003b> at EventThrowingClosure.ThrowEvent () <0x00023> at (wrapper delegate-invoke) System.MulticastDelegate.invoke_bool () <0x00037> at IdleProxy.Handler () <0x0002a> at (wrapper native-to-managed) IdleProxy.Handler () <0x00036> in (unmanaged) 0x43197610 at (wrapper managed-to-native) Gtk.Application.gtk_main () <0x00004> at Gtk.Application.Run () <0x00007> at Gnome.Program.Run () <0x00007> at Search.MainWindow.Main (string[]) <0x000f8> Expected Results: Beagle should show the search results. Additional info:
Very strange. I don't see this in 0.2.8-2. It seems like a gtk related crash. Could you try testing things like different themes to see if that makes it go away?
I tried different themes, but the crash still happened in the same way.
Does it crash whatever you type, or does it depend? The crash seems to be when loading an icon for an IM conversation hit, so maybe it only crashes when you hit something like that?
Infact it only crashes sometimes, e.g. I can get away with typing just 'a', and then it finds just one folder. It is able to find some things before crashing, so your suggestion sound quite plausible.
Could you verify that its in fact the IM log that is the problem. I.E. search for something that you know won't be in the gaim logs, and something that you know will only be there.
Hmm. Getting gaim conversation hits works for me.
(In reply to comment #6) > Hmm. Getting gaim conversation hits works for me. > Yes, it works for me too. I typed 'æ' in beagle-search, and an IM showed up. When I clicked it, it correctly opened Gaim's log to the right place. So it is not Gaim in particular.
Can you run "beagle-search --verbose blah" on the command line, replacing blah with a query that crashes, and post the output here.
beagle-query --verbose, not beagle-search.
Created attachment 134716 [details] beagle-query --verbose abe
(In reply to comment #10) > Created an attachment (id=134716) [edit] > beagle-query --verbose abe > I have obscured a couple of email addresses and aliases for privacy reasons.
Uri: file:///home/peter/.gaim/logs/irc/oejet.net/%23haskell.chat/2006-01-03.132501.txt PaUri: (null) Snip: (19:04:33) datenmaulwurf: @index <font color="red"><b>abs</b></font> ... (19:06:01) Lemmih: @. docs index <font color="red"><b>abs</b></font> ... (19:06:33) xerox: @compose docs index <font color="red"><b>abs</b></font> ... datenmaulwurf: but now i know what's <font color="red"><b>abs</b></font> ;) ... (19:06:58) Heffalump: @. (docs,index) <font color="red"><b>abs</b></font> ... (19:07:48) Cale: > map <font color="red"><b>abs</b></font> [-10..10] ... Type: IMLog MimeT: beagle/x-gaim-log Src: GaimLog Score: 15.6087693674607 Time: 1/4/2006 7:23:16 PM beagle:HitType = 'IMLog' beagle:MimeType = 'beagle/x-gaim-log' beagle:Source = 'GaimLog' This doesn't have fixme:protocol set...
Nor does it have any of the other required IM log properties. Peter, can you do the following? * shut down beagle (beagle-shutdown) * nuke your ~/.beagle/Indexes/GaimLogIndex directory * restart beagle with only the gaim backend enabled (BEAGLE_EXERCISE_THE_DOG=1 beagled --fg --debug --allow-backend GaimLog) * wait for it to index all your logs; it should go quickly since the scheduler throttling is turned off * do the search again and wait for it to crash Assuming it does crash, can you attach your ~/.beagle/current-* logs to the bug so we can hopefully see if there is an error there that's preventing the log from being detected correctly?
Created attachment 134735 [details] ~/.beagle/Log/current-Beagle
Created attachment 134737 [details] .beagle/Log/current-IndexHelperExceptions
Ok, this helps a lot. Would you mind attaching the relevant log to the bug? If you don't want to attach it, you can send it to me at joeshaw.
(In reply to comment #16) > Would you mind attaching the relevant log to the bug? Er, by which I mean the IRC log. /home/peter/.gaim/logs/irc/oejet.net/#haskell.chat/2006-01-03.132501.txt
Created attachment 134738 [details] Log that seems to touch the bug
Created attachment 134744 [details] Patch which fixes this and adds some sanity checks I just checked this patch into Beagle CVS. I'll probably do a 0.2.9 with it this weekend or early next week.
Thanks Joe, i've put that patch into 0.2.8-3.fc6