On Apr 22, 2013, at 4:06, "Thomas Koch" <[email protected]> wrote:
>> Thanks again. A trick - on Windows - might be to rename the testrepo dir > to >> testrepo.<current datetime> and move to the next test. And clean them all >> up at the end. Or create a testrepos directory tree where each >> testrepo.<currentTime> is going to be created and remove the whole tree at >> the end. > > Sure, that's similar to what I did for the test_PyLucene (in > setup/teardown): > >>> I then ended up in using a 'fresh' test-repository (i.e. new dir name) >>> if rmtree fails. The test_PyLucene test then passes on windows - the >>> downside is that you end up with some testrepo.1, testrepo.2 etc. dirs >>> - and of course an attempt to cleanup these dirs after >>> unittest.main(exit=False) also fails... > > Problem here is that one cannot cleanup the testrepos while inside the test > loop. It must be done outside, i.e. on Makefile level (well on windows > only). Yes, that was my thought too. >> GIven that Lucene 4.3 is about to be released, I think I'm going to drop > the >> 4.2.1 PyLucene release and move to proposing a PyLucene 4.3 release >> candidate next. > > Is there a release date yet? Lucene/Solr 4.3 RC2 is being voted on right now. Andi.. > Saw something on twitter, but not on the > website and JIRA still states a number of open issues: > https://issues.apache.org/jira/browse/LUCENE/fixforversion/12324143#selected > Tab=com.atlassian.jira.plugin.system.project%3Aversion-issues-panel > > > Regards, > Thomas > >
