DisjunctionMaxScorer.skipTo has bug that keeps it from skipping
---
Key: LUCENE-912
URL: https://issues.apache.org/jira/browse/LUCENE-912
Project: Lucene - Java
Issue Type: Bug
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hoss Man updated LUCENE-912:
Attachment: dismax_skipto.patch
patch with test and fix ... it would be helpful if someone who understands
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hoss Man updated LUCENE-912:
Lucene Fields: [Patch Available] (was: [New])
DisjunctionMaxScorer.skipTo has bug that keeps it from
To whom it may engage...
This is an automated request, but not an unsolicited one. For
more information please visit http://gump.apache.org/nagged.html,
and/or contact the folk at [EMAIL PROTECTED]
Project lucene-java has an issue affecting its community integration.
This issue affects
[
https://issues.apache.org/jira/browse/LUCENE-908?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502040
]
Michael Busch commented on LUCENE-908:
--
* manifest file in any of gdata's jars/war (it doesn't use the
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502048
]
Doron Cohen commented on LUCENE-912:
dismax score calculation isn't incremental... it's all done during the
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502054
]
Yonik Seeley commented on LUCENE-912:
-
Yes, two calls to score() should yield the same score.
What scorer was
[
https://issues.apache.org/jira/browse/LUCENE-908?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael Busch updated LUCENE-908:
-
Attachment: lucene-908.patch
In addition to Hoss' great patch this one:
- changes the MANIFEST
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doron Cohen updated LUCENE-912:
---
Attachment: checkTwoCallsToScore.patch
Attached adds to QueryUtils.check(Query q1, Searcher s):
-
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doron Cohen updated LUCENE-912:
---
Attachment: checkTwoCallsToScore.patch
Updating checkTwoCallsToScore.patch (unintended comment out
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael Busch updated LUCENE-912:
-
Attachment: lucene-912.patch
In BooleanScorer2.score() the coordinator is initialized:
[
https://issues.apache.org/jira/browse/LUCENE-908?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502104
]
Hoss Man commented on LUCENE-908:
-
A couple of random thoughts...
1) macro's can take multiple optional named
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502111
]
Doron Cohen commented on LUCENE-912:
This would recompute, but correctly.
If this fix is just for sanity, ie
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502120
]
Michael Busch commented on LUCENE-912:
--
Yes I agree. Keeping the score in BooleanScorer until next() or
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502121
]
Doron Cohen commented on LUCENE-912:
Actually I changed my mind (and saw your reply just before clicking
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doron Cohen reassigned LUCENE-912:
--
Assignee: Doron Cohen
DisjunctionMaxScorer.skipTo has bug that keeps it from skipping
: I feel like I am forgetting something with all the emails flying
: around about packaging, but don't the contrib jars also need their
: libs packaged with them in the binary distribution? Or at least a
: way to specify what they are and where to get them? They are there
: in the source
Two consecutive score() calls return different scores for Bollean Queries.
--
Key: LUCENE-913
URL: https://issues.apache.org/jira/browse/LUCENE-913
Project: Lucene - Java
-- Feature freeze from Wednesday (06/06)
All features must be checked in by end of Tuesday. On Wednesday I will
Michael, is there a need to hold commits to trunk while
the new branch is created?
(This is with LUCENE-912 and LUCENE-913 in mind - I think
they can wait for 2.3, there would
Doron Cohen wrote:
Michael, is there a need to hold commits to trunk while
the new branch is created?
No, normal trunk development may continue as usual.
(This is with LUCENE-912 and LUCENE-913 in mind - I think
they can wait for 2.3, there would always be new issues.)
I haven't
Hmmm, it is all nicely packaged in source, which satisfies the
discussion you listed below. As for the binary distributions, they
are pretty much worthless unless you have some way of knowing what
the dependencies are, right? Or am I missing something?
On Jun 6, 2007, at 5:45 PM, Chris
: Hmmm, it is all nicely packaged in source, which satisfies the
: discussion you listed below. As for the binary distributions, they
: are pretty much worthless unless you have some way of knowing what
: the dependencies are, right? Or am I missing something?
I would assume so, but i don't use
On 6/6/07, Michael Busch [EMAIL PROTECTED] wrote:
I haven't branched the trunk yet (I will send a different mail to this
list afterwards).
Since the patches are in place we might want to commit LUCENE-912 and
LUCENE-913 before?
+1, these are medium-serious bugs with fixes that look unlikely to
Michael Busch wrote:
I haven't branched the trunk yet (I will send a different mail to this
list afterwards).
Since the patches are in place we might want to commit LUCENE-912 and
LUCENE-913 before?
If there are no objections I plan to commit it later today.
[
https://issues.apache.org/jira/browse/LUCENE-913?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael Busch updated LUCENE-913:
-
Fix Version/s: 2.2
Two consecutive score() calls return different scores for Bollean Queries.
Doron Cohen wrote:
If there are no objections I plan to commit it later today.
+1. Thanks for taking care of these, Doron!
I'm working on LUCENE-908 in parallel. After all three are committed
(908, 912, 913)
I will make the branch.
Hi, following LUCENE-912, looking closer at skipTo() in
DisjunctionMaxScorer, it seems not to follow Scorer.skipTo() contract:
Skips to the first match beyond the current whose document number is
greater than or equal to a given target.
It is the __beyond__ the current requirement that is
The method states 'greater than OR EQUAL TO' so your d1 != d2 test is
invalid.
It should be assert (d2=d1)
On Jun 6, 2007, at 6:30 PM, Doron Cohen wrote:
Hi, following LUCENE-912, looking closer at skipTo() in
DisjunctionMaxScorer, it seems not to follow Scorer.skipTo()
contract:
: Hi, following LUCENE-912, looking closer at skipTo() in
: DisjunctionMaxScorer, it seems not to follow Scorer.skipTo() contract:
:
: Skips to the first match beyond the current whose document number is
: greater than or equal to a given target.
:
: It is the __beyond__ the current
Chris Hostetter wrote:
skipTo semantics have always made my head hurt, but it may be that the
contract is expressed too simplisticly ... i believe the expectation is
that while skipTo should always skip to the first match beyond the
current whose document number is greater than or equal to a
The method states 'greater than OR EQUAL TO' so
your d1 != d2 test is invalid.
It should be assert (d2=d1)
Yes, the '=' is regarding to the skip target, I am okay with this.
But beyond the current regards to where the scorer is currently
located, ie what doc() would have returned just
robert engels wrote:
The method states 'greater than OR EQUAL TO' so your d1 != d2 test is
invalid.
It should be assert (d2=d1)
Well, but the javadoc says BEYOND the current.
But I think it should be the desired behavior for skipTo() to not skip
at all
if curDoc==target already? Which
: The method states 'greater than OR EQUAL TO' so your d1 != d2 test is
: invalid.
:
: It should be assert (d2=d1)
I think you are missreading Doron's spuedo code (either that or i am) ...
the 'greater than OR EQUAL TO' statement is regarding the current match
and the target (input to skipTo)
Scorer.skipTo(current) remains on current for some scorers
--
Key: LUCENE-914
URL: https://issues.apache.org/jira/browse/LUCENE-914
Project: Lucene - Java
Issue Type: Bug
[
https://issues.apache.org/jira/browse/LUCENE-914?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502153
]
Hoss Man commented on LUCENE-914:
-
my gut says we should fix the Impls and leave the spec the way it is ... less
Chris Hostetter wrote:
: But I think it should be the desired behavior for skipTo() to not skip
: at all
: if curDoc==target already? Which means we should clearify the javadocs.
i'm not certain about that ... in theory (given the way the
javadocs are currently written) shouldn't s.skipTo(0)
[
https://issues.apache.org/jira/browse/LUCENE-908?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502156
]
Michael Busch commented on LUCENE-908:
--
As always these are very good recommendations Hoss! I think I will
On 6/6/07, Michael Busch [EMAIL PROTECTED] wrote:
I was just wondering if this behavior is what the average user would
expect from a skipTo() method
without reading the javadocs carefully enough.
skipTo() use isn't for the average user ;-)
This is very inner-loop stuff and should be primarily
[
https://issues.apache.org/jira/browse/LUCENE-908?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael Busch updated LUCENE-908:
-
Fix Version/s: (was: 2.2)
Patch committed. Leaving this issue open for the simplifications
[
https://issues.apache.org/jira/browse/LUCENE-914?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502162
]
Yonik Seeley commented on LUCENE-914:
-
The spec text is ambiguous.
The biggest issue is to look if any scorers
[
https://issues.apache.org/jira/browse/LUCENE-913?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Timm updated LUCENE-913:
-
Summary: Two consecutive score() calls return different scores for Boolean
Queries. (was: Two
[
https://issues.apache.org/jira/browse/LUCENE-912?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doron Cohen resolved LUCENE-912.
Resolution: Fixed
Committed the original lucene-912.patch by Hoss.
Thanks Hoss and Sudaakeran B.
[
https://issues.apache.org/jira/browse/LUCENE-913?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doron Cohen updated LUCENE-913:
---
Attachment: lucene-913.patch
patch attached for future reference, contains:
- Michael's fix
- new
[
https://issues.apache.org/jira/browse/LUCENE-913?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doron Cohen resolved LUCENE-913.
Resolution: Fixed
Lucene Fields: [Patch Available] (was: [New])
Committed, thanks Michael!
[
https://issues.apache.org/jira/browse/LUCENE-914?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doron Cohen updated LUCENE-914:
---
Attachment: lucene-914.patch
patch adds a test to QueryUtils that exposes this issue.
50 test cases
Doron Cohen wrote:
Since the patches are in place we might want to commit LUCENE-912 and
LUCENE-913 before?
If there are no objections I plan to commit it later today.
912 and 913 are committed. Great job, Doron! Thank you!
Alright, everything seems to be in place. Good timing! I
[
https://issues.apache.org/jira/browse/LUCENE-914?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502186
]
Yonik Seeley commented on LUCENE-914:
-
ConstantScorer is implemented like:
public boolean skipTo(int
[
https://issues.apache.org/jira/browse/LUCENE-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502213
]
Hoss Man commented on LUCENE-902:
-
A few comments in no particular order...
1) in future patches, could you please
In hte spirit of improving documentation, i've started a document
summarizing the various comments/ideas about improving out
documentation...
http://wiki.apache.org/lucene-java/Documentation_Improvements
: 3. There is a whole lot of knowledge stored in the email archives,
: how can we leverage
[
https://issues.apache.org/jira/browse/LUCENE-914?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502219
]
Doug Cutting commented on LUCENE-914:
-
The text was mostly trying to describe what the specified implementation
PorterStemmer is incorrectly truncating words ending in e
-
Key: LUCENE-915
URL: https://issues.apache.org/jira/browse/LUCENE-915
Project: Lucene - Java
Issue Type: Bug
[
https://issues.apache.org/jira/browse/LUCENE-915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12502221
]
Hoss Man commented on LUCENE-915:
-
can you elaborate on why you think this is a bug?
This is a fairly basic
[
https://issues.apache.org/jira/browse/LUCENE-915?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hoss Man resolved LUCENE-915.
-
Resolution: Invalid
I'd imagine you aren't going to fix it since it would require explicit
'exception
Hello Team,
well, first of all, let's take a deep breath! Behind us are a couple of
busy weeks. I would like to take this chance to thank everyone very much
for the great work! We're on track for our 2.2 release on the 19th of June.
As announced I created a Lucene 2.2 branch today from trunk
54 matches
Mail list logo
