Validate component IDs when creating component pools

[shsms]

No description provided.

[llucax]

Assigning to rc2300 because this is a breaking change. Also I think we need to mention it in the upgrading section of the release notes.

[llucax]

You could save a line of code and a couple cycles by reversing this:

Suggested change

	if not self._batteries.issubset(all_batteries):
	unknown_ids = self._batteries - all_batteries
	if unknown_ids := self._batteries - all_batteries:

(same for the rest)

[shsms]

that doesn't look like a subset check.

[llucax]

I would also add a note here, as this is effectively a breaking change. Users will need to update their code to deal with the exception, or validate the IDs before passing them to new_xxx_pool().

[shsms]

It is of course not a breaking change, because it was broken much worse earlier when they sent non-battery to battery pool. This is a more predictable meaningful error. Just a bug fix.

[llucax]

It is a breaking change in the sense that raising a new exception is an interface change, but I see your point that passing the wrong ID would also make the user code explode, only in a different and worse way, so OK. Rolling-back the breaking change label and milestone.

[llucax]

Added a commit to add the missing Raises section to the docstrings of methods now raising a ValueError.

Needs approval from someone else now. Enabling auto-merge too, as Sahas will be away for a few weeks.

[llucax]

Tests still failing for the same reasons as #1343, see comments in that PR for details.

[llucax]

Rebased to get the CI fix, and needed to merge some conflicts for that, so it needs a new approval.