Andi,
I tested what happens when I fail to call initVM, and unfortunately I
hit a SEGV (different from previous Bus Error):
Thread 0 Crashed:
0 libjcc.dylib 0x000ab4a6 JCCEnv::fromUTF(char
const*) + 50
1 _lucene.so 0x01303449 p2j(_object*) + 47
2 _lucene.so 0x01305096 _parseArgs(_object**,
unsigned int, char*, ...) + 5366
3 _lucene.so 0x010a0139
org::apache::lucene::store::t_FSDirectory_getDirectory(_typeobject*,
_object*) + 131
4 org.python.python 0x0018f51e PyEval_EvalFrameEx + 17116
5 org.python.python 0x00191173 PyEval_EvalCodeEx + 1638
6 org.python.python 0x00191260 PyEval_EvalCode + 87
7 org.python.python 0x001a883c PyErr_Display + 1896
8 org.python.python 0x001a8e66 PyRun_FileExFlags + 135
9 org.python.python 0x001aa7d2 PyRun_SimpleFileExFlags
+ 421
10 org.python.python 0x001b5a57 Py_Main + 3095
11 org.python.pythonapp 0x00001fca 0x1000 + 4042
This is the code:
import lucene
#lucene.initVM(lucene.CLASSPATH)
store = lucene.FSDirectory.getDirectory('/lucene/test.idx')
Playing with it a bit, if I call:
lucene.Document()
I get the nice RuntimeError back telling me to initVM. But if I do
this instead:
lucene.Document
(don't apply it), then I hit a SEGV.
Sorry I didn't test out your fix until the RC...
Mike
Andi Vajda wrote:
Last week's vote failed to get the 3 PMC votes to make an official
release. Only one PMC vote was cast, a +1 from Mike McCandless.
In the meantime, some bugs were fixed and PyLucene 2.4.1 RC5 was
uploaded to the staging area:
http://people.apache.org/~vajda/staging_area/
Please vote to release these artifacts as PyLucene 2.4.1.
Thanks !
Andi..
