WIP: inject closure-calls for loop externs in closures

[jaekwon]

This is an alternative implementation to #1768 and #1585 for #1135.
This should be much more performant overall, as loop externs are rare.

• as pointed out by Maxwell, this would also need to work with GOTO implicit loops. Should be doable, but another approach would be to disallow this edge case (probably very rare).

UPDATE: noticed that "TestSearchEfficiency" is affected but the runtime is not because the closure is called synchronously.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 45.75%. Comparing base (ac78e0b) to head (fba2746).
Report is 114 commits behind head on master.

Additional details and impacted files

@@             Coverage Diff             @@
##           master    #1780       +/-   ##
===========================================
- Coverage   55.94%   45.75%   -10.20%     
===========================================
  Files         422      460       +38     
  Lines       64269    69518     +5249     
===========================================
- Hits        35956    31808     -4148     
- Misses      25483    35049     +9566     
+ Partials     2830     2661      -169     

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

[jaekwon]

NOTE: I haven't actually tested the running of this logic, yet.

[jaekwon]

"TestSearchEfficiency" is triggered by this but the tests have been passing because the closure is called synchronously. good sign that it is picked up by this panic, and probably a good one to focus on first.

[petar-dambovaliev]

ROFL! Epic panic message.

[jaekwon]

working on #2429 instead.