[
https://issues.apache.org/jira/browse/LUCENE-929?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12986844#action_12986844
]
Shai Erera commented on LUCENE-929:
-----------------------------------
I didn't find a way to check if an Ant directory is empty, and even if there is
a way, the extract process can still fail in the middle. So maybe an
alternative solution is to extract the files in ExtractReuters to a tmp
directory first, and rename it if everything succeeds. Also, always delete the
target dir at startup. That way, if extraction fails, the directory doesn't
exist, and if it does exist, we know it's ok (unless manually tempered, but
that's not for us to solve).
What do you think?
> contrib/benchmark build doesn't handle checking if content is properly
> extracted
> --------------------------------------------------------------------------------
>
> Key: LUCENE-929
> URL: https://issues.apache.org/jira/browse/LUCENE-929
> Project: Lucene - Java
> Issue Type: Bug
> Components: contrib/benchmark
> Reporter: Grant Ingersoll
> Priority: Minor
>
> The contrib/benchmark build does not properly handle checking to see if the
> content (such as Reuters coll.) is properly extracted. It only checks to see
> if the directory exists. Thus, it is possible that the directory gets
> created and the extraction fails. Then, the next time it is run, it skips
> the extraction part and tries to continue on running the benchmark.
> The workaround is to manually delete the extraction directory.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
