All 10 Enums & Flags questions
More in Enums & Flags
What is the performance consideration when using Enum.HasFlag() versus a direct bitwise AND check? How would you write an extension method to safely add or remove a specific flag from a [Flags] enum value without manual bitwise operators scattered throughout your code? How do you safely validate that an enum value received from external input (like deserialized JSON or a database) is actually a DEFINED member, given that enums don't enforce this by default? How would you use pattern matching (switch expressions) with an enum, including exhaustiveness warnings from the compiler? How would you design an enum-based state machine using pattern matching to enforce valid state transitions at compile time or runtime?
Advertisement
Back to results for “Enums”
C# › Enums & Flags

How do enums interact with generic constraints, and why has it historically been awkward to write generic, reusable flag-manipulation code?

Advanced

Answer

Prior to more recent C# versions, you couldn't constrain a generic type parameter to 'must be an enum' cleanly enough to perform bitwise operations generically (the 'where T : struct, Enum' constraint exists, but bitwise operators aren't directly usable on a generic Enum-constrained parameter without casting through the underlying integral type) — this has led to workarounds using Convert.ToInt64() or reflection for fully generic flag utilities.
public static bool HasFlagGeneric<T>(T value, T flag) where T : struct, Enum {
  long valueAsLong = Convert.ToInt64(value);
  long flagAsLong = Convert.ToInt64(flag);
  return (valueAsLong & flagAsLong) == flagAsLong; // workaround: convert through long
}
Real-world example Writing a single, reusable flag-checking utility that works generically across MULTIPLE different [Flags] enum types in a codebase.

Common follow-ups: How do newer .NET generic math features (INumber<T> and related interfaces) begin to address this kind of generic numeric limitation?

Generics
All 10 Enums & Flags questions