Setting of _main_module about dill HOT 4 CLOSED

Hmm… that looks like it could be an oversight. I don't usually extend dill with the Pickler and Unpickler, so that code's not as strongly tested there. Tell you what… why don't make a pull request and I'll review it? It'd be especially nice if you show how it fixes a simple issue for you. If you don't want to do that, then I'll patch it and write the tests -- but I'd like you to have first crack at it, since it's your patch.

from dill.

The following is a test program that should work:

import dill
import StringIO

def my_fn(x):
    return x * 17

obj = lambda : my_fn(34)
print 'obj() returns %d' % obj()

obj_io = StringIO.StringIO()
pickler = dill.Pickler(obj_io)
pickler.dump(obj)

obj_str = obj_io.getvalue()
print 'obj() pickled to %d bytes' % len(obj_str)

obj2_io = StringIO.StringIO(obj_str)
unpickler = dill.Unpickler(obj2_io)
obj2 = unpickler.load()

print 'obj2() returns %d' % obj2()

An associate of mine is preparing a minimal pull request. If that looks like it's heading in the right direction, I think that a couple of things could be cleaned up to simplify things. I suggest we get something working first. Then we submit a second pull request with additional suggestions that I think (but am not 100% sure) would be good for the code.

from dill.

@citibob: sounds good. Thanks for the follow-up.

The procedure is as you have posed… a minimal pull request for pull request for a patch… and additional pull requests for additional features. The more atomic, the easier to merge, the less likely that a merge affects others negatively and is off scope, and thus is more likely to get in.

from dill.

closed by merging #24.

from dill.