[
https://issues.apache.org/jira/browse/LUCENENET-90?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12522399
]
Digy commented on LUCENENET-90:
-------------------------------
Hi George,
I looked at the java version of testIndexModifier and saw that the line causing
the test to fail is commented out.
// depend on merge policy - Assert.AreEqual(3, i.DocCount());
So with a minimal patch that test can succeed(I will attach the new patch for
TestIndexModifier).
But the real problem continues to exist. IndexWriter does not give the "real"
document number till segments are merged and DocCount is used everywhere
(including other tests) with that logic(Codes using the docCount know that
issue). But in this case, there are no way to get the "real DocCount" using
APIs.
Therefore i think java version should also be updated like in these patches.
DIGY
> Nunit test for TestIndexModifier.TestIndex
> ------------------------------------------
>
> Key: LUCENENET-90
> URL: https://issues.apache.org/jira/browse/LUCENENET-90
> Project: Lucene.Net
> Issue Type: Bug
> Reporter: Digy
> Priority: Minor
> Attachments: IndexModifier.patch, IndexWriter.patch,
> TestIndexModifier.patch, TestIndexModifier2.patch
>
>
> IndexModifier uses DocCount of IndexWriter which returns the count of all
> documents(including deleted ones) unless segments are merged(ex. when
> explicitely calling Optimize). Changing the DocCount of IndexWriter effects
> other parts of the project. So I added a new method
> "DocCountExcludingDeletedDocs" to IndexWriter and used it in IndexModifier.
> With this patch Nunit test succeeds.
> I will also add a patch for a new test method "TestIndexModifierDocCount"
> into the TestIndexModifier.cs
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
