Contributing.md

Adding new components

See the developers guide for a description on how components can be defined and what rules are to be followed.

TOOD: Description on the idea/design-vision of the workbench (accessibility) --Best-practice on that here-- (direct input vs load-like extra component)

(Feel free to talk to Jan von Harten/the entire team to check if the visions are compatible :) )

For each component (added):

• Follow the naming guide for the identifier
• Use pydocs explaining your component and its inputs
• Add a unit test of the execution
  • use a small input to keep it fast
  • See the tests unit tests for an example of this
• Consider invalid inputs/limitations and define pre-checks for them
  • Add unit tests for each pre-check
    • Each pre-check MUST be tested
• Add the new component to an example (python) workflow
  • And ensure the workflow runs (on your machine)

Adding new types

See the developers guide for a description on how types can be defined and what rules are to be followed.

For each type (added):

• Follow the naming guide
• Use pydocs explaining your type
• Add a unit test ensuring it fulfills the following requirements (TODO: example):
  • a valid pydantic model
  • serializable and deserializable
• Use it in an example and run the example on your machine

Note: pydantic does not support some types, such as numpy arrays, without annotated information on how to (de)serialize them. For numpy types, please use pydantic_numpy and see the MeshResults here for additional examples and here for the pydantic docs on that topic.