SharedObjectManager
Provides a high-level interface to create a manager that supports shared memory objects through the multiprocessing interface Only supports data that can be marshalled into a NumPy array.
primitive_types: tuple
__init__(self, obj, base_dict=None, parallelizer=None):
mem_manager:Anya memory manager like
multiprocessing.SharedMemoryManagerobj:Anythe object whose attributes should be given by shared memory objects
base_dict:SharedMemoryDictthe dict that stores the shared arrays (can also be shared)
is_primitive(val):
save_attr(self, attr):
del_attr(self, attr):
load_attr(self, attr):
get_saved_keys(self, obj):
save_keys(self, keys=None):
share(self, keys=None):
load_keys(self, keys=None):
unshare(self, keys=None):
__del__(self):
list(self, *l):
dict(self, *d):
array(self, a):