I am currently studying the Invariant part of a course and have a question regarding the execution order of functions when using invariant testing. Specifically, I am looking at the statefulFuzz_testInvariantBreakHandler function which includes the following code snippet:

function statefulFuzz_testInvariantBreakHandler() public {
 vm.startPrank(owner);
 handlerStatefulFuzzCatches.withdrawToken(mockUSDC);
 handlerStatefulFuzzCatches.withdrawToken(yeildERC20);
 vm.stopPrank();

 assert(mockUSDC.balanceOf(address(handlerStatefulFuzzCatches)) == 0);
 assert(yeildERC20.balanceOf(address(handlerStatefulFuzzCatches)) == 0);
 assert(mockUSDC.balanceOf(owner) == startingAmount);
 assert(yeildERC20.balanceOf(owner) == startingAmount);
}

My question pertains to the order in which handlerStatefulFuzzCatches.withdrawToken(mockUSDC) and handlerStatefulFuzzCatches.withdrawToken(yeildERC20) functions are executed in relation to the fuzzing process:

1. Are these withdrawToken functions called at the beginning of the test, and then the fuzzer attempts to break the invariants by calling other functions?
2. Or, are these specific withdrawToken calls supposed to be executed after the fuzzer has run through other functions?

I am seeking clarification on whether the specified withdrawToken operations are meant to set up the state for invariant testing. Thank you for your assistance.