This removes files and code supporting scons builds. After this change
numpy will only support builds using distutils or bento. The removal of
scons has been discussed on the list several times and a decision has been
made that scons support is no longer needed. This was originally discussed
for numpy 1.7 and because the distutils and bento methods are still
available we are skipping the usual deprecation period.

The different sort kinds are now located in

    quicksort.c.src
    mergesort.c.src
    heapsort.c.src

Both direct and indirect sorts are included in each file. This cleanup
was done so that additional sorts or quickselect could be added without
undue clutter.

There are two reasons to want to keep PyArray_ReduceWrapper out of the
public multiarray API:
 - Its signature is likely to change if/when masked arrays are added
 - It is essentially a wrapper for array->scalar transformations
   (*not* just reductions as its name implies -- the whole reason it
   is in multiarray.so in the first place is to support count_nonzero,
   which is not actually a reduction!). It provides some nice
   conveniences (like making it easy to apply such functions to
   multiple axes simultaneously), but, we already have a general
   mechanism for writing array->scalar transformations -- generalized
   ufuncs. We do not want to have two independent, redundant
   implementations of this functionality, one in multiarray and one in
   umath! So in the long run we should add these nice features to the
   generalized ufunc machinery. And in the short run, we shouldn't add
   it to the public API and commit ourselves to supporting it.

However, simply removing it from numpy_api.py is not easy, because
this code was used in both multiarray and umath. This commit:
 - Moves ReduceWrapper and supporting code to umath/, and makes
   appropriate changes (e.g. renaming it to PyUFunc_ReduceWrapper and
   cleaning up the header files).
 - Reverts numpy.count_nonzero to its previous implementation, so that
   it loses the new axis= and keepdims= arguments. This is
   unfortunate, but this change isn't so urgent that it's worth tying
   our APIs in knots forever. (Perhaps in the future it can become a
   generalized ufunc.)

This removes the include dependency on PY_UNICODE_WIDE in preparation
for defining more general unicode functions in the ucsnarray.c file.

Doing this so that it can be used both in multiarray and umath, to make
writing new reduction operations generally easier.

Also made PyArray_Squeeze work with NA-masked arrays.