JSpecify: initial handling of generic enclosing types for inner classes

[msridhar]

This is a step towards fixing #836. We now properly handle the case of a generic enclosing type for an inner class type, both in checking matching of type argument nullability and in pretty-printing for error messages. We still don't handle NewClassTree types correctly which is why #836 is not yet fixed.

Also fixes a bug in where generics checks were performed for return statements. Previously we would only check generic type arguments if the top-level nullability of the return type was @NonNull.

[codecov]

Codecov Report

Attention: 6 lines in your changes are missing coverage. Please review.

Comparison is base (e7623f7) 86.85% compared to head (68cccfc) 86.86%.

Additional details and impacted files

@@            Coverage Diff            @@
##             master     #837   +/-   ##
=========================================
  Coverage     86.85%   86.86%           
  Complexity     1879     1879           
=========================================
  Files            74       74           
  Lines          6194     6196    +2     
  Branches       1201     1202    +1     
=========================================
+ Hits           5380     5382    +2     
  Misses          406      406           
  Partials        408      408           

Files	Coverage Δ
...away/src/main/java/com/uber/nullaway/NullAway.java	89.73% <100.00%> (ø)
...rc/main/java/com/uber/nullaway/GenericsChecks.java	90.28% <85.36%> (+0.05%)	⬆️

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[msridhar]

The missing coverage report from @codecov is due to an incorrect computation of what code is actually changing in this diff

[lazaroclapp]

LGTM. Some nits.