Support for reset/image/load activation profiles

[andy-maier]

Ready for review.
Has been tested on A65.
To try this PR yourself, run make install after checking out the branch, in order to install the updated "zhmcclient" package.

[coveralls]

coverage: 36.876% (+2.5%) from 34.352%
when pulling 8222fae on andy/add-activation-profiles
into 4d5d7a5 on master.

[EdGue42]

Note: this is so much code that I just scrolled over some parts.
I think you could significantly reduce line count just by introducing helper methods for a lot of repeated structures, that do always the same, only using different property name strings for example.

[andy-maier]

I have introduced helper functions for the setting of properties from specially handled options. Please re-review.

[EdGue42]

Looks better now.

[andy-maier]

I wonder whether it is a good idea to copy the properties from the default profile ourselves (i.e. on the client side) as opposed to letting that be done by the HMC. I think my motivation to copy it on the client side was that we would be on the safe side given that we are not representing all resource properties 1:1 as command options, but I don't think initializing the properties from the default profile is necessary for that to work.

Also, since the HMC uses the profile named "DEFAULT" when --copy-name is not specified, it is inconsistent to copy the profile when --copy-name is specified, but not to copy the "DEFAULT" profile when it is not specified.

Overall, I suggest that we do not initialize the properties from the default profile, and let the HMC do that.

Applies to all three activation profile types.

[andy-maier]

I removed the client-side copying.

Please review that approach.
@EdGue42

[andy-maier]

I looked again at the find_...() functions and whether they take cpc_name or cpc_or_name as an argument. All of them take cpc_or_name as an argument, except the three new ones for the activation profiles. But none of them uses the ability for Cpc as a resource object, it was just passed through, but never actually called that way.

So we could remove the support for a Cpc resource object in the find_:::() functions and reduce that to just always the CPC name. That would eliminate some other portions of duplicated code.

I'll do that in a separate PR.

Update: That became PR #525