Introduction
------------

pyperf is a simple performance measuring library for python which uses python 
decorators to mark the functions you'd want to get performance data. After
marking your functions witht the @pyperf.measure decorator you can then turn the 
library on and off by simply setting the global variable pyperf.PYPERF to True. 

Using PyPerf
------------

Using the pyperf library is very easy and straightforward to use and here is a
quick example:

import pyperf

@pyperf.measure
def do_something_function():
    return "some data"

Now by default pyperf is turned off and will have minimal application overhead
when turned off. To turn it on you can just set the pyperf.PYPERF to True and 
when you exit your application automatically print out the stats of the 
various methods you decorated earlier, like so:

PyPerf Report:
==============

function | tot calls | tot dur(ms) | avg dur(ms) | max dur(ms) | min dur(ms)
----------------------------------------------------------------------------
 func3   |     30    |      5      |      0      |      0      |      0     
 func2   |     27    |      4      |      0      |      0      |      0     
 func1   |     18    |      2      |      0      |      0      |      0     

There are a few other options you can enable when using the pyperf module that
actually can give you more advanced information about the methods that you want
to track performance on. The available options are:

pyperf.PYPERF_TRACKARGUMENTS - when set to True pyperf will track the stats per
                               call to each of the functions but making sure to
                               separate the stats by the actual arguments used
                               during calling. The stats output will contain
                               function names with the arguments used during
                               each of the calls.

pyperf.PYPERF_TRACKCALLER - when set to True pyperf will track the stats by 
                            separting the calls to any decorated function by the
                            caller to that function. The output will contain 
                            function names that look like so: 
                            caller_function->called_function
                            
Installing
----------

python setup.py install

or install directly from github with:

pip install -e git+git://github.com/rlgomes/pyperf.git#egg=pyperf

Running Built-In Tests
-----------------------

Go to the tests directory and you can run all the tests by executing:

cd tests
python alltests.py

License
-------

Apache 2.0 License (http://www.apache.org/licenses/LICENSE-2.0.html)