Error if graph has a dependency offset exclusively on the RHS

[wxtim]

Closes #5844

Raise an error if a graph has a dependency offset only on the RHS of a graph pair.

Check List

• [x] I have read CONTRIBUTING.md and added my name as a Code Contributor.
• [x] Contains logically grouped changes (else tidy your branch by rebase).
• [x] Does not contain off-topic changes (use other PRs for other changes).
• [x] Applied any dependency changes to both setup.cfg (and conda-environment.yml if present).
• [x] Tests are included (or explain why tests are not needed).
• [x] CHANGES.md entry included if this is a change that can affect users
• [x] Cylc-Doc pull request opened if required at cylc/cylc-doc/pull/XXXX.
• [x] If this is a bug fix, PR should be raised against the relevant ?.?.x branch.

[oliver-sanders]

@wxtim what does "…aph pair." mean?

[MetRonnie]

I think he used the commit message as the PR title and it got truncated, continuing on the body. I updated the title to something a little more readable

[oliver-sanders]

wxtim requested a review from MetRonnie

@wxtim, you have an outstanding review comment.

[wxtim]

wxtim requested a review from MetRonnie

@wxtim, you have an outstanding review comment.

I think that I was waiting for Ronnie and you finish the discussion. Will fix now.

[oliver-sanders]

@hjoliver could you check you're happy with this approach.

This is back-filling a check which was removed in: https://github.com/cylc/cylc-flow/pull/4343/files#diff-be87b3fcb20fc3228166ad53af61efb79f0d0c84ee01fcd0c9e3092c788e0b93L374-L377 (see the issue)

I think the logic has become more complex in order to permit examples like a => b[-P1] => c which presumably failed validation at Cylc 7.

[wxtim]

@MetRonnie - I've used your words. Please merge at your leisure.

[MetRonnie]

Tests need updating