Dart Generics: Covariance, Contravariance, and Type Safety

Dart supports generics similar to Java and Kotlin, but its handling of covariance, contravariance, and invariance differs. In Dart, generic classes are covariant by default, allowing a List<Girl> to be used where a List<Person> is expected. This simplifies code reuse but introduces safety risks because the collections remain mutable.

The article provides concrete examples: a generic PrintMsg<T>(T value) function, usage of List<int> versus List<dynamic> , and how type inference works when generic parameters are omitted. It also contrasts Dart's runtime type retention with Java/Kotlin's type erasure, showing that Dart can perform type checks like value is List<String> at runtime.

Covariance and contravariance are explained using Java's ArrayList<? extends Number> and ArrayList<? super Number> syntax, and the PECS principle (Producer‑Extends, Consumer‑Super) is introduced. The article demonstrates how these concepts enable safe polymorphic behavior in collections, such as adding a List<Boy> to a List<Person> via addAll , and using forEach with a Consumer<? super Person> .

In Dart, because generic collections are both covariant and mutable, writing to a covariant collection can cause runtime type errors. The article shows a problematic case where a PrintMsg(List<Person> persons) function receives a List<Girl> and attempts to add a Boy , leading to a type mismatch.

To mitigate such issues, Dart introduces the covariant keyword, allowing overridden methods to narrow parameter types safely. Examples include BoyHouse and GirlHouse classes that override a checkin method with covariant Boy or covariant Girl parameters, preventing inappropriate objects from being passed.

The article also highlights real‑world Flutter usage, such as passing a List<ElevatedButton> to a function expecting List<Widget> , and shows how the covariant keyword is applied in the project's state management classes to enforce type safety.

Overall, the piece emphasizes understanding generic variance, applying the PECS principle, and using Dart's covariant modifier to write safer, more maintainable mobile code.