Type Guards#

weitsicht provides a few small helper functions that act as type guards. They are mainly useful for static type checking (pyright, mypy), where they can help narrow base classes (like ImageBase) to concrete subclasses (like ImagePerspective) based on the runtime type discriminator.

Available guards#

  • is_opencv_camera: narrow CameraBasePerspective → CameraOpenCVPerspective

  • is_camera_type: check a camera discriminator (narrowing works when called with a literal, e.g. CameraType.OpenCV)

  • is_perspective_image: narrow ImageBase → ImagePerspective

  • is_ortho_image: narrow ImageBase → ImageOrtho

They are available from weitsicht (top-level) and from weitsicht.type_guards.

Examples#

from weitsicht import ImageBase, is_perspective_image

def process(img: ImageBase) -> None:
    if is_perspective_image(img):
        # here img is treated as ImagePerspective by type checkers
        _ = img.camera
    else:
        # here img is treated as ImageOrtho | ImageBase (depending on your checker)
        pass

from weitsicht import CameraBasePerspective, CameraType, is_camera_type

def process_camera(cam: CameraBasePerspective) -> None:
    if is_camera_type(cam, CameraType.OpenCV):
        # here cam is treated as CameraOpenCVPerspective by type checkers
        _ = cam.camera_matrix