Simple implementation of the Dog API for finding and displaying images.
This implementation contains a single external dependency: KingFisher. Caching and cache busting is a complicated enough topic that farming this out to a vendor is highly reasonable. Beside this dependency, every other component is built within the Swift language and the UIKit/Foundation frameworks. No Combine, no SwiftUI - just a sample of using core language features together.
Things that are here:
- An initial list of dog breeds is displayed on app launch
- A search box awaits user input; on entry, matching dog breeds are displayed along with sample images
- Tap a cell to auto-search that dog breed
Things that are a bit crusty:
- The DataManager and associated Operations are a bit kludgy, and could use a cleaner abstraction layer. (The DispatchGroup is kinda icky, for example)
- There isn't much of an architecture here; the abstractions are somewhat standalone, as implementing an entirely DI'd, MVVM'd, Tweakable solution is deemed far out of scope
- API Error handling is covered up with default, empty models. Makes the code easier to implement, but of course hides actual errors.
- There is no debouncing against the search API. Each character can potentially result in an iteration through known dogs and respective API calls for each of their image sets.
Things that are missing:
- UI / Data tests
- UI Error states
- Data mocking, UITests