lucenenet-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chand2048 <...@git.apache.org>
Subject [GitHub] lucenenet pull request: Chand8
Date Mon, 16 Feb 2015 22:51:49 GMT
GitHub user Chand2048 opened a pull request:

    https://github.com/apache/lucenenet/pull/98

    Chand8

    

You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/guidotag/lucenenet Chand8

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/lucenenet/pull/98.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

    This closes #98
    
----
commit 7b2b53e40272140e488ebdfcb18b45b4a0f8d88d
Author: Chand2048 <chand@microsoft.com>
Date:   2015-02-11T18:14:46Z

    Fix for TestIndexableField.TestArbitraryFields - In java, the byte class is signed and
in C# it is unsigned. Changed the unit test to cast to a (byte) insetad of an (sbyte).
    
    An alternate much deeper fix would be to change the Lucene.Net.Util.BytesRef class to
use sbyte insetad of byte. Not sure if this class was intentionally setup this way or if it
is a conversion error. Comments please.

commit 31164e4bf5c671b10ad807f469e0a28926294e51
Author: Chand2048 <chand@microsoft.com>
Date:   2015-02-11T18:38:53Z

    The ParallelCompositComposit reader now has a ParallelCompositReaderAnonymousInnerClassHelper
for the inner instance instead of ParallelCompositReader. Looking back at the history, it
has been this way for a long time but this test does not exist in the java version. Is this
test in place to show there is something that needs to be fixed in ParallelCompositComposit
reader?

commit 01d719c135191947d62800422828c8fea95897a6
Author: Chand2048 <chand@microsoft.com>
Date:   2015-02-12T20:56:38Z

    Cast the float to a double before converting to a string to get extra precision. This
helps keep all the digits for comparison later.

commit e0805a9f8f7d27c587df232a39ab30545fab4ab3
Author: Chand2048 <chand@microsoft.com>
Date:   2015-02-14T00:47:37Z

    There was a race condition triggered by TestDocumentsWriterDeleteQueue.TestStressDeleteQueue.
The value of Next changed after the Node original = Next; statement. This would cause the
function to return false and end up corrupting the linked list. Comparing with cmp (the value
passed in) ensures that we properly return false if there is a change in Next.

commit c210126c24b1e95c0062f5fc7e1f0bf732320fa4
Author: Chand2048 <chand@microsoft.com>
Date:   2015-02-14T01:57:19Z

    Fix for TestDocumentsWriterDeleteQueue.TestUpdateDelteSlices
    
    The test was using an iterator to compare two HashSet<>. I created a function to
turn the hashSet<> into an array, sort and then compare.

commit 3c6d13009571ebd7bc513fc8625df72b8cc777e2
Author: Chand2048 <chand@microsoft.com>
Date:   2015-02-16T22:50:38Z

    Better but not a full fix for TestThreadInterruptDeadlock and TestTwoThreadsInterruptDeadlock.
    
    C# already has a ThreadingInterruptedException and core components like SyncTextWriter
throw this exception when they are interrupted. I updated the unit test to catch both Lucene.Net.Util.ThreadInterruptedException
(already did this) and System.Threading.ThreadInterruptedException.
    
    This change makes both of the unit tests fail during the first phase of the two phase
commit because they are interrupted. Looking into the IndexWriter class, this seems to be
the expected behavior but the unit test still says it is not good. This needs more research.

----


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---

Mime
View raw message