Stabilizer-search

Small changes to setup, including adding a package description and manifest. The dependencies for the Cython extensions are now built locally with setup.py as opposed to distributed with the package.

This release slightly modifies the behaviour of _Result.success, to make it a boolean value for intutivitely checking if a search succeeded, for exampe

python
search = RandomWalkSearch(target, 'A State', n_qubits, chi)
res = search()
if res.success:
print("Yeah!")
else:
print("Oh.")


It also extends the RandomWalkResult and BruteForceResult classes. The resulting states in a successful decomposition are stored in the property `.basis`, and the string representation in `.decomposition`.

---

We are in the process of cythonizing certain components of the code to improve performance, and to potentially allow for parallel processing of searches.

In this version, key functions have been optimised using Cython. This includes Gram-Schmidt orthogonalisation to build a projector, and the random walk.

The default method of the RandomWalkSearch class has been updated to the cython form. If you want to use the old python backed, this can be achieved with:

python
from stabilizer_search.search.random_walk import do_random_walk
search = RandomWalkSearch(.... , func=do_random_walk)

This is the initial release of Stabilizer Search, also available from [PyPI](https://pypi.python.org/pypi?name=stabilizer-search&version=1.0.0&:action=display), and installable using `pip install stabilizer_search`.

This version contains pure python code for generating stabilizer states. Future releases plan to optimise this method. It also contains routine for Brute Force searches for stabilizer rank, and a random walk method adapted from [arXiv:1506.01396](https://arxiv.org/abs/1506.01396).