* Previously we would convert from python to C++ arrays immediately by
copying, and vice-versa convert TO python immediately by creating a
new python list by copying.
* This however behaves rather poorly in common situations, e.g.:
> foo.bar.append(5)
Would not append 5 to foo.bar, but copy foo.bar to a temporary, append
5 to it, then destroy it leaving foo.bar untouched.
* Instead we leave the C++ array type as a pointer for as long as we can
and instead implement the python sequence API as extensions/slots that
work in-place on the original array.
* The former is only needed inside tp_init of a new object. Instead when
we want to pass in and own a pointer, we use SWIG_POINTER_OWN.
* This also removes the need to pass 'self' all the way down in
ConvertToPy which tidies up a lot of code.
* The bug seems to happen if two raw strings concatenated together are
large enough, so instead we pass them as separate parameters to a
different macro then concatenate them inside the macro.
* In future we could handle async exceptions by storing the exception
information in a std::function derived object (instead of the separate
ExceptionHandling that lives on the stack) and query it out in a new
WaitForInvoke function maybe. Right now we just print the exception
to the output log and abort the callback.
* Even if we don't have pyside2 to treat qwidgets as full objects and
access their methods and data, we still need to be able to pass around
the QWidget* as an opaque pointer to be able to use the API properly.
* This change falls back to just using SWIG's default opaque pointer
wrapping and unwrapping when pyside2 isn't available.
* This means that if we BlockInvoke a python callback, we don't end up
in a deadlock where we're holding the GIL during script execution but
also waiting on the renderer running (which is trying to acquire the
GIL before calling the callback).
baldurk