Should we implement hook methods from `cmd.Cmd` so we can document them?

[kotfu]

We are not consistent in defining and documenting methods from cmd.Cmd which we use/rely-on, but which perform the same function as they do in our superclass. For example, cmd2.Cmd provides an implementation of the precmd() method from our superclass, which does exactly the same thing as the superclass method. We do not implement postcmd() or preloop() or postloop(), which are no-ops in the superclass, i.e. they are stub methods waiting to be over-ridden.

Providing an implementation of these methods allows us to reference them in our documentation, which is nice.

I don't care whether we do or don't provide an implementation for these methods, but whichever we choose, we ought to do it the same for all four of these hook-type methods.

Any preference?

[tleonhardt]

I’m in favor of providing an override of all cmd.Cmd hook methods within cmd2.Cmd for the following reasons:

• It makes them more discoverable to anyone browsing the cmd2 code
• We can provide a better doctring which can help developers properly understand correct usage
• As you mentioned, it makes it easier for us to document them in our Sphinx docs