Fix precedence of safe-accessor tail

[marcelocantos]

Fixes #552

Change safe-accessor tail to expect an expr instead of the next rule on the precedence stack.

The expression a.b?:c.d used to parse as (a.b?:c).d. This is confusing because it means that .d, which looks like it belongs to c, actually applies to a.b if a has a b attribute. This PR changes it to parse as a.b?:(c.d).

This is actually a bit of a hack. My preference would be to loop back to the current rule. However, wbnf doesn't have a way to refer to the current rule, only the next rule (@) or back to the top (expr). It might be better to add this capability to wbnf. Comments welcome.

Checklist:

• Added related tests
• Made corresponding changes to the documentation

[coveralls]

Pull Request Test Coverage Report for Build c293e8ca5c743b40acee1500092d00da4d6541df-PR-525

Warning: This coverage report may be inaccurate.

This pull request's base commit is no longer the HEAD commit of its target branch. This means it includes changes from outside the original pull request, including, potentially, unrelated coverage changes.

• For more information on this, see Tracking coverage changes with pull request builds.
• To avoid this issue with future PRs, see these Recommended CI Configurations.
• For a quick fix, rebase this PR at GitHub. Your next report should be accurate.

Details

• 0 of 0 changed or added relevant lines in 0 files are covered.
• 3 unchanged lines in 2 files lost coverage.
• Overall coverage increased (+0.04%) to 48.034%

---

Files with Coverage Reduction	New Missed Lines	%
syntax/import_cache.go	1	77.27%
rel/value_tuple.go	2	75.2%

Totals
Change from base Build 6ca2c118006fa40a4732f1387f9a84906ef94f73:	0.04%
Covered Lines:	4423
Relevant Lines:	9208

---

💛 - Coveralls

[orlade-anz]

Can you add a description of what this is actually doing for posterity?

It looks vaguely related to #532 but doesn't fix it.

[orlade-anz]

The description is good but hard to grok; can you also add a test case that would have failed before but should pass now and vice versa?

[marcelocantos]

We really should fix the stack operator in wbnf so that this rule can be self-referential. arr-ai/wbnf#85