A module for serving up python data in a stand-alone process.


PythonDataServe is a simple library that allows you to serve data from a different python process. This eliminates the need to load large data files every time you want to run your program.

The library includes a very simple HTTP Server for basic data types and a Proxy Server for python objects. Both servers implement a similar methodology, allowing the user to define functions on the server side which the client can call with a data payload to and get a return response. This is a remote-procedure-call type of implementation which also makes the library an excellent tool for doing multiprocessing across computers.

The library also contains a DataContainer utility class to encapsulate pickling, gzipping, loading and saving functionality. This simplifies some common data operations.

The only external requirement is the commonly used "requests" library.


For the latest documentation, see ReadTheDocs. Alternatively, you can view the markup files directly in the docs folder of this project.

You can see example implementations in the tests folder.