This is to announce the release of PyLucene 0.9.8.
Besides a few minor additions and fixes the major change in this release is that all calls made to Java now yield the Python GIL (Global Interpreter Lock). The few APIs I had overloaded with an optional 'yield' argument, such as IndexWriter.optimize(), have been removed.
This change is risky as calling python without holding the GIL is a fatal error. I believe I fixed up all the code to ensure that the GIL is held when it needs to be but errors are always possible.
The complete change list is below.
Andi..
- added support for RangeFilter, ported RangeFilter unit tests
- added support for Throwable
- java exceptions now reported in a JavaError python exception
- added support for Object.getClass()
- added partial support for Class
- Windows _PyLucene.pyd now stripped
- fixed bug in Indexer.py Lucene in Action sample, it did not recurse dirs
- changed Document.__getitem__() to return field value instead of field object
- added Document.__getattr__() to return field object from field name
- replaced Document.__delitem__() with Document.__delattr__()
- all calls to java now yield GIL, additional yield API overloads removed
- fixed ParallelMultiSearcher with python extension searchers deadlock bug
- incorporated Erik Hatcher's man page indexer samples
_______________________________________________ pylucene-dev mailing list [email protected] http://lists.osafoundation.org/mailman/listinfo/pylucene-dev
