========
Overview
========
A fast and thorough lazy object proxy.
* Free software: BSD license
Note that this is based on `wrapt`_'s ObjectProxy with one big change: it calls a function the first time the proxy object is
used, while `wrapt.ObjectProxy` just forwards the method calls to the target object.
In other words, you use `lazy-object-proxy` when you only have the object way later and you use `wrapt.ObjectProxy` when you
want to override few methods (by subclassing) and forward everything else to the target object.
Example::
def expensive_func():
# create expensive object
return stuff
obj = lazy_object_proxy.Proxy(expensive_func)
# function is called only when object is actually used
print(obj.foobar) # now expensive_func is called
Installation
============
::
pip install lazy-object-proxy
Documentation
=============
https://python-lazy-object-proxy.readthedocs.io/
Development
===========
To run the all tests run::
tox
Acknowledgements
================
This project is based on some code from `wrapt`_ as you can see in the git history.
.. _wrapt: https://github.com/GrahamDumpleton/wrapt
Changelog
=========
1.3.1 (2017-05-05)
------------------
* Fix broken release (``sdist`` had a broken ``MANIFEST.in``).
1.3.0 (2017-05-02)
------------------
* Speed up arithmetic operations involving ``cext.Proxy`` subclasses.
1.2.2 (2016-04-14)
------------------
* Added `manylinux `_ wheels.
* Minor cleanup in readme.
1.2.1 (2015-08-18)
------------------
* Fix a memory leak (the wrapped object would get bogus references). Contributed by Astrum Kuo in
`#10 `_.
1.2.0 (2015-07-06)
------------------
* Don't instantiate the object when __repr__ is called. This aids with debugging (allows one to see exactly in
what state the proxy is).
1.1.0 (2015-07-05)
------------------
* Added support for pickling. The pickled value is going to be the wrapped object *without* any Proxy container.
* Fixed a memory management issue in the C extension (reference cycles weren't garbage collected due to improper
handling in the C extension). Contributed by Alvin Chow in
`#8 `_.
1.0.2 (2015-04-11)
-----------------------------------------
* First release on PyPI.