[bt-devel] [ bibletime-Bugs-1468307 ] Searching in Commentaries has problems

SourceForge.net noreply at sourceforge.net
Sat Feb 10 10:34:43 MST 2007


Bugs item #1468307, was opened at 2006-04-11 11:41
Message generated for change (Comment added) made by mgruner
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=100954&aid=1468307&group_id=954

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Frontend / Search dialog
Group: in progress
Status: Open
>Resolution: Accepted
Priority: 5
Private: No
Submitted By: Wolfgang Stradner (ewst)
Assigned to: Martin Gruner (mgruner)
Summary: Searching in Commentaries has problems

Initial Comment:
Searching in the JFB commentary with Ctrl+O for
'elapse' gives:
- wrong results (e.g. Dan 8:27)
- not all results (e.g. Rev 1:1 missing)
When I started to search in JFB, indexing took place.
(so probably not any older indexing was kept)

----------------------------------------------------------------------

>Comment By: Martin Gruner (mgruner)
Date: 2007-02-10 18:34

Message:
Logged In: YES 
user_id=169722
Originator: NO

This seems to be a problem in clucene, possibly. The index contains rev
1:0 up to the point where IRENÆUS is mentioned. The Æ seems to be a
problem for clucene. Wonder why it worked before, though.... 

Joachim, do you have an idea?

God bless, mg



----------------------------------------------------------------------

Comment By: Wolfgang Stradner (ewst)
Date: 2007-02-06 10:45

Message:
Logged In: YES 
user_id=1385862
Originator: YES

Opened JFB in BT 1.6.3b and checked for the words mentioned below:
E.g. informant in REV 1:1:
- is found when using Ctrl-F (finds it within the page), but
- not found with the Search-tool on the toolbar of the JFB window;
Searching for informant only finds  
-- DAN 4:30 and
-- MRK 2:2
Same problem when using the Search Ctrl-O and the English WEB bible and
JFB open:
- it finds only two hits in JFB (same as before, but not the one in REV
1:1, which actuelly is in the last paragraph).

-> Check it out

----------------------------------------------------------------------

Comment By: Martin Gruner (mgruner)
Date: 2007-01-14 20:15

Message:
Logged In: YES 
user_id=169722
Originator: NO

Very good analysis. This was the problem, and I believe it is fixed in
CVS. You can test it whenever 1.6.3 is out.

----------------------------------------------------------------------

Comment By: Wolfgang Stradner (ewst)
Date: 2006-12-20 16:20

Message:
Logged In: YES 
user_id=1385862
Originator: YES

BT 1.6.2 with clucene 1.16a does not find'elapse' in JFB/Rev 1:1 yet.

Testing with other words in JFB/Rev 1:1:
- compuation: not found
- mutally: occurs three times (can be found with Ctrl+F):
-- being mutually connected (marked in yellow with the search: Ctr+O)
-- mutually agreeing; (marked in yellow with the search: Ctr+O)
-- mutually involving
- informant: not found
- hastened: not found.

JFB/Rev 1:1 is a very long verse/text unit, as it is composed of:
- comment to the whole book of revelation and
- comment to verse 1
I have the impression, that the works, which occur only once in Rev 1:1 at
the very end of this text unit are not found.

-> Could it be, that clucene limits the size of text units which it can
search through?  (hastened and informant, which cannot be found are both
close to the end of JFB/Rev 1:1?

----------------------------------------------------------------------

Comment By: Martin Gruner (mgruner)
Date: 2006-11-01 10:13

Message:
Logged In: YES 
user_id=169722

Ok. In CVS (will be in 1.6.2) the missing results are fixed. The
"superfluous" results are actually correct. For example in Dan 8,27, the
match is in text of Dan 9:0, but BibleTime does not display Dan 9:0. I
think there is already a bug report relating to this so that we can close
this one. Thanks.

----------------------------------------------------------------------

Comment By: Jim Campbell (jim-campbell)
Date: 2006-06-20 06:24

Message:
Logged In: YES 
user_id=689979

The Rev 1:1 problem if fixed for me.
I am using the current cvs build with clucene-0.9.12.

The wrong results are still a mystery.
I get wrong results searching for time in the KJV module. 
It will report Titus 3:15.  I have seen this behavior
before, but I cannot remember the searches.
Also, I get other false hits.  For example, JFB commentary:
hit on II Kings 6:25 for "entertain".

I will dig a little deeper but this appears to be out of my
league.  Although, my thoughts cause me to believe that
clucene is finding a word match between two words that are
loosely related.
For example searching for <root-word> will match
<pre><root-word><post>.
Or maybe its just a bug :)

If anyone has any clues, please let me know.

----------------------------------------------------------------------

Comment By: Martin Gruner (mgruner)
Date: 2006-06-17 20:40

Message:
Logged In: YES 
user_id=169722

Jim, since you are our search expert I'd like to assign this to you.
Please speak up if this is not ok for you. Thanks.

The bug seems to be fixed partly, perhaps by the clucene upgrade to
0.9.12.

----------------------------------------------------------------------

Comment By: Wolfgang Stradner (ewst)
Date: 2006-04-27 12:08

Message:
Logged In: YES 
user_id=1385862

I checked it again with 1.6beta2:
The wrong result for elapse in JFB/Dan 8:27 is still there,
(I did not find an occurence of elapse in Rev 1:1).
Checking for other words seem to work quite well, I did not
find other problems.
What is nice, that the order of bible books is correct now
in the list of results.




----------------------------------------------------------------------

Comment By: Joachim Ansorg (joachim)
Date: 2006-04-27 10:24

Message:
Logged In: YES 
user_id=1486

I did some fixes in beta2 to solve this problem. Let us know
whether it still is buggy.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=100954&aid=1468307&group_id=954



More information about the bt-devel mailing list