The following common pattern could possibly be optimized further: ``` Object.assign({}, nonEmptyObject); ``` If we know that {} is an empty final object, could we reuse nonEmptyObject's JSC::Structure? > https://github.com/WebKit/WebKit/pull/13220#issuecomment-1524421218 > I think probably opposite case is getting many benefit from the past optimizations already, but we could get better by detecting empty object of the target in DFG / FTL and change it to object-cloning fast path. Can you file a bugzilla to track this kind of possibility?
<rdar://problem/108875645>
Pull request: https://github.com/WebKit/WebKit/pull/17242
Pull request: https://github.com/WebKit/WebKit/pull/17312