flutter-base-classes

Flutter Base Classes

Full reference: template.md

Key rules

Cubit

• All cubits extend BaseCubit<State> or use SafeEmitMixin. Never call raw emit() — always safeEmit() (guards against closed-cubit errors).
• BaseHydratedCubit<S> — extends HydratedCubit<S> with SafeEmitMixin<S>. Use when state must survive process death. Only base_hydrated_cubit.dart imports hydrated_bloc.
• EnumHydratedCubit<E extends Enum> — extends BaseHydratedCubit<E>. Eliminates duplicated fromJson/toJson for enum-based settings (theme, language). Subclass declares values, jsonKey, defaultValue.
• States are sealed classes extending Equatable. No boolean flags in state — use separate state subclasses.
• Cubits depend on use cases, never on repositories directly.

Use cases

• Extend UseCase<Return, Params> or StreamUseCase<Return, Params>.
• Use NoParams for no-argument use cases. Use a record typedef (typedef FooParams = ({String x, int y})) for multi-param.
• File: *_usecase.dart. Class: *UseCase.

Repository

• Use BaseRepository mixin. It provides safeCall<T>() (catches FailureException → rethrows; catches others → wraps as UnknownFailure) and mapToEntities<D,E>().
• AppLogger injected via constructor, exposed as AppLogger get logger.

Entity / DTO

• Entities: extend BaseEntity (or just Equatable), immutable, no fromJson. Live in domain/entities/.
• DTOs: extend BaseDto, annotated @JsonSerializable(), expose toEntity(). Live in data/dtos/.

Paginated cubit

• Use BasePaginatedCubit<Item> for lists with load-more. States: PaginatedInitial, PaginatedLoading, PaginatedLoaded, PaginatedEmpty, PaginatedError.

Co-load with

• flutter-architecture — how base classes fit into clean-arch layers
• flutter-di — cubits as registerFactory
• flutter-network — BaseRepository.safeCall catches FailureException