On Tue, 2006-10-03 at 18:19, Samuel wrote: > Thanks, that's what I was looking for. > > > >>> m = __import__( "StringIO" ) > > >>> x = getattr( m, "StringIO" )() > > >>> x > > <StringIO.StringIO instance at 0x00A47710> > > >>> > > For the records: If the module is already loaded, this also works: > > if my_type_is_not_yet_loaded: > module = __import__(type) > obj = getattr(module, type) > else: > obj = globals().get(type) > resource = obj(my_arg1, my_arg2)
You seem to be under the impression that importing an already imported module is a hideously expensive operation that must be avoided at all costs. It's not. If you import an already imported module, python simply returns the module object from sys.modules without executing any of the module's code. Your code will be much easier to read if you eliminate the "my_type_is_not_yet_loaded" check, whatever that may be under the hood, and always perform the __import__. It's also likely to be at least as fast, but I couldn't run a timing comparison even if I wanted to because you didn't post working code. -Carsten -- http://mail.python.org/mailman/listinfo/python-list
