• Pickle

What is Pickle?

http://docs.python.org/library/pickle.html

Should I use Pickle?

Of course not.

• It is insecure. Untrusted pickles can do arbitrary things. For example, this pickle executes arbitrary Python expressions: pickle.loads("c__builtin__\neval\n(c__builtin__\nraw_input\n(S'py> '\ntRtR.")

• It is Python-only: pickles cannot be loaded in any other programming language / environment.

• It is schemaless and opaque. See: http://bpaste.net/show/d0UZUSsAHtekA0pDiQzu/

Alternatives

• The json module (http://docs.python.org/library/json.html) can handle dicts, lists, ints, floats, strings, booleans, and None. It cannot handle reference cycles, however. (pickle can.)

• PyYAML (http://pyyaml.org/) can handle everything json can, as well as having comments inside source data, and reference cycles. However, it defaults to an unsafe pickle-like loader.