fix(binding-http): fill security scheme when baseURI defined

[relu91]

Found a bug when exposing things with a baseURI defined. I added a failing test to verify the correct behavior and avoid regression.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 77.81%. Comparing base (450ebd9) to head (c283db9).
Report is 5 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1244      +/-   ##
==========================================
+ Coverage   77.79%   77.81%   +0.01%     
==========================================
  Files          84       84              
  Lines       17484    17523      +39     
  Branches     1771     1781      +10     
==========================================
+ Hits        13602    13635      +33     
- Misses       3846     3853       +7     
+ Partials       36       35       -1     

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

[danielpeintner]

LGTM

The only concern I had initially was whether it makes any difference to put
this.fillSecurityScheme(thing);

after the if else ... as it was before but I don't think this matters in this case.

[relu91]

I reviewed the code and I can confirm that addEnpoint does not have a direct dependency on what fillSecurityScheme is doing. You can check the code and see that addEnpoint does not manipulate anything other than affordances objects inside the thing. I would say that it is safe to keep the code as proposed and it will increase readability (e.g. since security schemes are often coming before affordances the coder is less surprised to see them handled first).

[danielpeintner]

Latest changes look good again 👍