I'm trying to learn how raising and handling errors works in packages and saw that in Pandas they define most of the Exception Classes in the __init__.py with only two messages getting put into a separate .py file in the errors folder. I thought that the __init__.py had two primary purposes. A) To tell python that this folder has code in it and should be treated as a package. (which is no longer needed in more recent versions of python) B) To control the imports by defining __all__ = [...] But for the most part, the __init.py can be left empty. So can someone explain why Pandas uses their __init__.py instead of creating exceptions.py (or something) within the errors folder? Continue reading...
