011: Splitting the form parsing into objects

Initially a single Form class handled the parsing of the form definition JSON as well as a lot of the logic around what different sections meant. This works fine but led to a lot of places in code where we’re passing around arguments to determine whether a page or section should or shouldn’t do something rather than being able to ask it directly.

Refactoring this into smaller form domain object classes has several benefits:

  • It’s easier to compare the form definition JSON to the code classes and reason about what fields can be passed and what effect they’ll have
  • It moves business logic out of the helpers and keeps them to just dealing with display logic
  • It makes it easier to unit test form functionality, and group that into smaller chunks
  • It allows for less passing of arguments. e.g. page.routed_to?(lettings_log) vs form.was_page_routed_to?(page, lettings_log)

This abstraction is likely still not the best (the form vs lettings log split) but this seems like an improvement that can be iterated on.