DLCanvasRecorder uses an SkCanvasVirtualEnforcer to field the SkCanvas API calls and record them in the internal did<Transform>, onClip<Shape> and onDraw<Thing> methods.

The enforcer will maintain the current transform for you by updating its internal transform before calling the associated did<Transform> methods.

The enforcer will only maintain the current clip bounds, though, if you allow its onClip<Shape> methods to be called. Unfortunately, the DLCanvasRecorder intercepts those calls and does not pass them along to the enforcer.