Skip to content

Serialization/deserialization using Python Pickle format

License

Notifications You must be signed in to change notification settings

ftomassetti/python-pickle

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

python-pickle

python-pickle is a Haskell library to serialize and deserialize Python objects using the Python Pickle format. It supports protocols 0, 1, and 2 for deserializing and uses protocol 2 for serialization.

It offers a Value data type to represent Python objects, and two functions, pickle and unpickle, to serialize and deserialize such values.

Implementation detail

In practice the Python Pickle format is a list of serialized opcodes. By deserializing those opcodes then interpreting them, one can reconstruct the original object. Interpreting the opcodes is done using a simple stack-based Pickle machine.

If you want to learn about the Pickle format, the standard Python pickletools library source is a good place to start.

Install

Installing from Hackage with Cabal is straightforward:

> cabal update ; cabal install python-pickle

The development version can be installed by cloning the Git repository and using Cabal:

> git clone git://github.com/noteed/python-pickle.git
> cd python-pickle && cabal install

Command-line tool

A pickle executable is provided. It can be used to inspect a pickle file (i.e. a file containing a pickled Python object).

Limitation

  • Not all opcodes are implemented.
  • Strings are not correctly escaped.
  • Can pickle (i.e. serialize) only with protocol 2.
  • Unpickling a protocol 0 string (which uses literal representation of e.g. float values) is not accurate (possibly the Python implementation is not accurate either).

About

Serialization/deserialization using Python Pickle format

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Haskell 96.2%
  • Python 3.8%