|
Size: 481
Comment:
|
← Revision 4 as of 2012-05-12 05:24:57 ⇥
Size: 904
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 13: | Line 13: |
| * It is schemaless (may be seen as a benefit sometimes) | * 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. |
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.