[
https://issues.apache.org/jira/browse/LUCENE-5154?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13724851#comment-13724851
]
Robert Muir commented on LUCENE-5154:
-------------------------------------
The issue with DIH is that one test writes a dataimport.properties (i havent
looked further).
The issue with velocity is that its logging is using log4j (not slf4j).
Does this module really work if you don't have log4j setup, or do the tests
only pass because of the test environment...?
Cleanest seems if we could use
http://velocity.apache.org/engine/devel/apidocs/org/apache/velocity/slf4j/Slf4jLogChute.html
but I dont see that in a release. I'm sure there are other solutions (not a
fan of logging, sorry)
> ban tests from writing to CWD
> -----------------------------
>
> Key: LUCENE-5154
> URL: https://issues.apache.org/jira/browse/LUCENE-5154
> Project: Lucene - Core
> Issue Type: Test
> Reporter: Robert Muir
> Attachments: LUCENE-5154.patch
>
>
> Currently each forked jvm has cwd = tempDir = .
> This provides some minimal protection against tests in different jvms from
> interfering with each other, but we can do much better by splitting these
> concerns: and setting cwd = . and tempDir = ./temp
> Tests that write files to CWD can confuse IDE users because they can create
> dirty checkouts or other issues between different runs, and of course can
> interfere with other tests in the *same* jvm (there are other possible ways
> to do this to).
> So a test like this should fail with SecurityException, but currently does
> not.
> {code}
> public void testBogus() throws Exception {
> File file = new File("foo.txt");
> FileOutputStream os = new FileOutputStream(file);
> os.write(1);
> os.close();
> }
> {code}
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
