feat: add limited support for mixed operators

[rhamzeh]

This PR adds initial support for mixed operators

Description

Cases to be supported (both from and to DSL)
(with no direct assignment):

• define relation: (rel1 and rel2) but not rel3
• define relation: ((rel1 and rel2) but not rel3) # transformation back to DSL will not keep the extraneous surrounding brackets
• define relation: rel1 and (rel2 but not rel3)
• define relation: rel1 and ((rel2 but not rel3) or (rel4 and (rel5 but not rel6 from rel7)))

(with direct assignment, only if there is a single reference to the assignment and only if it is the first entry):

• define relation: [user, user:*, user#follower with condition] or ((rel1 and rel2) but not rel3)
• define relation: ([user] or ((rel1 and rel2)) but not rel3 # Note we can de-prioritize this as pending discussion

Cases NOT planned to supported yet, pending discussion:
Anything involving multiple references to direct assignment, or the presence of direct assignment not as the first entry, or references to self/this:

• define relation: [user] as self | rel1 but not self
• define relation: [user] as self | this but not rel1
• define relation: rel1 but not [user]
• define relation: [user] as self | (self or ((rel1 and rel2)) but not rel3
• define relation: (rel3 or ((rel1 and rel2)) but not [user]
• define relation: (rel3 or (([user] and rel2)) but not self
• define relation: ([user] and allowed) or ([user] but not blocked]

References

• Closes feat: support nested userset rewrite operands in the DSL #16

Review Checklist

• I have clicked on "allow edits by maintainers".
• I have added documentation for new/changed functionality in this PR or in a PR to openfga.dev [Provide a link to any relevant PRs in the references section above]
• The correct base branch is being used, if not main
• I have added tests to validate that the change in functionality is working as expected

[d-jeffery]

Sadly this grammar, while pretty robust doesnt catch define relation: rel1 and ([user] or thing) as expected. Might have to validate it during transformation, unless we can spot a way to catch this with the grammar rules.

Nevermind, think i got it.