[
https://issues.apache.org/jira/browse/LUCENENET-143?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doug Sale updated LUCENENET-143:
--------------------------------
Attachment: TestStressIndexing2.patch
Resolves outstanding issues. Incorporates fixes from earlier
"TestStressIndexing2.patch" patches.
These tests create docs composed of random fields and add them to an index.
The index documents are then copied, field-by-field to another index. The
indexes are then compared for equality. (This is all done with multiple
threads and myriad different knobs twisted this way and that...)
In the Java code, there were "random" orderings of the fields, which were
sometimes sorted at different points depending on chance.
However, the randomizations in the Java code are providing the completely
consistent results, as the logic of the original tests could not work otherwise.
The fixes incorporated in this patch include (introduced prior):
- a fix to the random ordering of the fields (not ported properly)
- a fix to the IComparator to ensure that fields are ordered by name and value
(otherwise, fields with the same values were in an undefined order)
The recent fix
- removing the field sort when copying from original doc to new doc (when
multiple fields of the same name exist in a doc, changing their order of
addition changes the term position and offsets, causing the tests to fail)
> NUnit test for Index/TestStressIndexing2
> ----------------------------------------
>
> Key: LUCENENET-143
> URL: https://issues.apache.org/jira/browse/LUCENENET-143
> Project: Lucene.Net
> Issue Type: Bug
> Environment: Lucene.Net 2.3.1
> Reporter: Digy
> Priority: Minor
> Attachments: FieldsReader.patch, TestStressIndexing2.patch,
> TestStressIndexing2.patch, TestStressIndexing2.patch
>
>
> Following patch resolves a bug, but this test continues to fail.
> DIGY
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
