When generating code, we don't want to remove any classes, as we want to support more versions than just the most recent one. For example, if Harbor removes support for the Robots V1 API in some new version, we still want to support it for users who are using older versions of Harbor.
Solutions
We can go one of two ways in my mind: Manual or automatic
Manual
The detection needs to find out which classes were removed, if any, and tell the user to add the class definition to the fragments directory so it is added to the final generated file.
Automatic
The code generator finds classes that were removed and automatically creates a fragment containing that class definition. This could be somewhat difficult to implement. We probably need to do it as some separate step prior to running parser.py.
When generating code, we don't want to remove any classes, as we want to support more versions than just the most recent one. For example, if Harbor removes support for the Robots V1 API in some new version, we still want to support it for users who are using older versions of Harbor.
Solutions
We can go one of two ways in my mind: Manual or automatic
Manual
The detection needs to find out which classes were removed, if any, and tell the user to add the class definition to the fragments directory so it is added to the final generated file.
Automatic
The code generator finds classes that were removed and automatically creates a fragment containing that class definition. This could be somewhat difficult to implement. We probably need to do it as some separate step prior to running
parser.py
.