Fluent API Generators should now create a "Data" layer that states how the method should be implemented (see the new DataGeneratorsFactory and DataMethodBodyBuilders).
The non-fluent part, aka. the generators using attributes, should also rely on this abstraction layer.
Generators should build the method body based on the provided "Data" alone, regardless of how it was built. Extend the "Data" layer as needed.
Create tests for the data building layer if they build matching data.
Refactor and tests the code generators themselves, make sure they use the "data" layer as input.
- Focus on Body Generators, entire method generators (including name) are WIP and should be commented out if needed in this PR.
- Explicit SwitchCase statements will be replaced in the future, so comment out those as well (see DataMethodBodyBuilders.cs for details)
Fluent API Generators should now create a "Data" layer that states how the method should be implemented (see the new DataGeneratorsFactory and DataMethodBodyBuilders).
The non-fluent part, aka. the generators using attributes, should also rely on this abstraction layer.
Generators should build the method body based on the provided "Data" alone, regardless of how it was built. Extend the "Data" layer as needed.
Create tests for the data building layer if they build matching data.
Refactor and tests the code generators themselves, make sure they use the "data" layer as input.