I understand that in the DbC method, preconditions and postconditions are attached to a function.
What I'm wondering is if that applies to member functions as well.
For instance, assuming I use invariants at the beginning at end of each public function, a member function will look like this:
edit: (cleaned up my example)
void Charcoal::LightOnFire() {
invariant();
in_LightOnFire();
StartBurning();
m_Status = STATUS_BURNING;
m_Color = 0xCCCCCC;
return; // last return in body
out_LightOnFire();
invariant();
}
inline void Charcoal::in_LightOnFire() {
#ifndef _RELEASE_
assert (m_Status == STATUS_UNLIT);
assert (m_OnTheGrill == true);
assert (m_DousedInLighterFluid == true);
#endif
}
inline void Charcoal::out_LightOnFire() {
#ifndef _RELEASE_
assert(m_Status == STATUS_BURNING);
assert(m_Color == 0xCCCCCC);
#endif
}
// class invariant
inline void Charcoal::invariant() {
assert(m_Status == STATUS_UNLIT || m_Status == STATUS_BURNING || m_Status == STATUS_ASHY);
assert(m_Color == 0x000000 || m_Color == 0xCCCCCC || m_Color == 0xEEEEEE);
}
Is it okay to use preconditions and postconditions with global/generic functions only and just use invariants inside classes?
This seems like overkill, but maybe its my example is bad.
edit:
Isn't the postcondition just checking a subset of the invariant?
In the above, I am following the instructions of http://www.digitalmars.com/ctg/contract.html that states, "The invariant is checked when a class constructor completes, at the start of the class destructor, before a public member is run, and after a public function finishes."
Thanks.