What are the differences between pickle.dump/load and pickle.dumps/loads?

Question

I've started to learn about the pickle module used for object serialization and deserialization.

I know that pickle.dump is used to store the code as a stream of bytes (serialization), and pickle.load is essentially the opposite, turning a stream of bytes back into a python object. (deserialization).

But what are pickle.dumps and pickle.loads, and what are the differences between them and pickle.dump and pickle.load? I've looked at the documentation, but I am having trouble differentiating between the two.

From the documentation:

pickle.dumps(obj, protocol=None, *, fix_imports=True, buffer_callback=None) ; Return the pickled representation of the object obj as a bytes object, instead of writing it to a file.

pickle.loads(data, /, *, fix_imports=True, encoding="ASCII", errors="strict", buffers=None) Return the reconstituted object hierarchy of the pickled representation data of an object. data must be a bytes-like object.

Answer 1

The difference between dump and dumps is that dump writes the pickled object to an open file, and dumps returns the pickled object as bytes. The file must be opened for writing in binary mode. The pickled version of the object is exactly the same with both dump and dumps.

So, if you did the following for object obj:

with open("pickle1", "wb") as f:
    pickle.dump(obj, f)
with open("pickle2", "wb") as f:
    f.write(pickle.dumps(obj))

you'd end up with two files with exactly the same contents.

The same applies to loading - load "unpickles" from an open (readable) file object, and loads uses a bytes object.