What is a partial class, and when is it used?
Quick Answer
The partial keyword lets a single class (or struct/interface) be split across multiple files; the compiler merges them into one type at build time. The main use is separating generated code from hand-written code — a WinForms Designer.cs, an EF-generated entity — so regenerating the tool file never overwrites your customisations.
partial lets you spread one class's definition across more than one file, and the compiler stitches them together into a single type when it builds. The reason it exists is code generation — a designer or a scaffolding tool owns one file it can freely regenerate, and you own a second partial file with your own members and overrides, so regenerating never wipes out your customisations. EF Core's scaffolded entities and WinForms designer files both use this. There's also partial void, which lets generated code declare a hook that's simply skipped by the compiler if you never implement it. I wouldn't split an ordinary class into partials just to make files shorter — that's not what it's for.
- All parts must share the same accessibility and be compiled together; you can put different members and even different attributes/base-interface declarations on different parts.
partial voidmethods let a generator declare an optional extension point that's a no-op if you never implement it (and the call is compiled away entirely if unimplemented).- Modern EF Core scaffolding and source generators lean on partial classes/methods heavily.
- Overuse to split an ordinary hand-written class for “organisation” is usually a smell — it's for tooling boundaries, not general file-size management.