Andi Vajda <va...@apache.org> wrote: > > On Wed, 18 Mar 2009, Andi Vajda wrote: > > > > > A long standing problem with PyLucene and JCC has been the crash one > > gets when forgetting to call initVM() to boot the embedded Java VM. > > > > After thinking about this some more, it became clear that a Java VM > > without classes is not very useful and that hence a good place to > > put a is-there-a-vm-yet check would be in the one place that finds > > java classes, the one place one must at least go through once, > > JCCEnv's findClass method. > > > > In rev 755707, I checked in an implementation of this. Getting this > > to work was not exactly trivial as JCC generated lots of unprotected > > initializeClass() calls. > > > > I hope I found them all. Please try the latest out and let me know > > if you can get it to crash without initVM() being called. > > > > >>> import lucene > > >>> lucene.Document() > > Traceback (most recent call last): > > File "<stdin>", line 1, in <module> > > RuntimeError: initVM() must be called first > > > > If this checks out OK, given that we haven't yet received the > > required three PMC votes to issue an official release, I'm ready to > > upload an RC5 into the staging area. > > On the heels of the same fix, a failure to find a class because of, > say, a CLASSPATH error, also no longer crashes. > > Committed in rev 755738. > > Andi..
Excellent! Bill
