The `test.support` has been split into sub-modules. Adapt `test_exceptions` (originally coped from CPython) for these changes.

Eliminate C compiler warnings by making the constness of the
char* pointers in the declarations in pycapsule.pxd match
the actual types of the PyCapsule_* functions.

Disable time_pxd test in Python 3.4

It uses features that are unavailable thus always fails

Test was introduced in https://github.com/cython/cython/pull/3767

Closes https://github.com/cython/cython/issues/4000

Closes https://github.com/cython/cython/issues/4000

It was broken in https://github.com/cython/cython/pull/3976 which added an additional error message.

Fixed e_decorators test

Warning message on cpdef variables (0.29.x)

Expose "long and overflow" cpython functions

These functions are available since Python 2.7.

Close #3980

Changed "declared after use" warning to error

* ENH: implement cpython/time.pxd
Co-authored-by: scoder <stefan_ml@behnel.de>

In 0.11.1 is was marked as an warning that would become an error.
3.0 is probably either the time to make it an error (or to remove
the note that it'll be an error in future)

Allowing these gives people the false impression that they
do something meaningful. "Fixes" https://github.com/cython/cython/issues/3959

Always use "new" string->float conversions

This ensures that 1_2_3 type numbers are handled correctly
whether or not the argument type is known.

Fixes https://github.com/cython/cython/issues/3958 (and thus the
currently broken builtin_float test)

Fix missing utility code for complex annotations

Generating utility-code for complex types is special-cased, but
it wasn't when they were generated in annotations.

Fixes https://github.com/cython/cython/issues/3949

Though pretty standard, I can see how the use of these terms could
reasonably be viewed as offensive.

MAINT: Add const qualifiers to char* types in pycapsule.pxd.

Eliminate C compiler warnings by making the constness of the
char* pointers in the declarations in pycapsule.pxd match
the actual types of the PyCapsule_* functions.

Use Py_TYPE() and __Pyx_SET_SIZE() functions