fix: Updated shimmer to handle instrumenting named and default exports of CommonJS modules in ESM

[jsumners-nr]

This PR resolves #1821.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (31dc717) 96.87% compared to head (b47a52c) 96.88%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1894   +/-   ##
=======================================
  Coverage   96.87%   96.88%           
=======================================
  Files         209      209           
  Lines       39774    39863   +89     
=======================================
+ Hits        38533    38622   +89     
  Misses       1241     1241           

Flag	Coverage Δ
integration-tests-16.x	78.69% <46.77%> (-0.06%)	⬇️
integration-tests-18.x	78.97% <46.77%> (-0.06%)	⬇️
integration-tests-20.x	78.98% <46.77%> (-0.06%)	⬇️
unit-tests-16.x	90.88% <60.00%> (-0.10%)	⬇️
unit-tests-18.x	90.86% <60.00%> (-0.10%)	⬇️
unit-tests-20.x	90.86% <60.00%> (-0.10%)	⬇️
versioned-tests-16.x	73.78% <100.00%> (+0.03%)	⬆️
versioned-tests-18.x	73.78% <100.00%> (+0.03%)	⬆️
versioned-tests-20.x	73.79% <100.00%> (+0.03%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[mrickard]

LGTM, and much more readable. @bizob2828 should weigh in as well.

[bizob2828]

overall looks good. I'm not sure of the impact of proxying proxies but appears to work with our versioned tests. I do have a few comments and small things that need addressed

[jsumners-nr]

I'm not sure of the impact of proxying proxies but appears to work with our versioned tests.

Frankly, it's a terrible idea. But we need to double up the decoration for CJS imported as ESM, otherwise we have to educate users that the imports do not work as they expect. I do think the IITM team is open to moving this work into the IITM module. So I'm going to work up a patch over there that will do the same thing and see if we can then simplify here. But I am convinced we need the symbol added in this implementation and I do not think they are open to that bit.