Codecov Report

Base: 92.38% // Head: 92.59% // Increases project coverage by +0.20% 🎉

Coverage data is based on head (186325f) compared to base (3c4d3b3).
Patch coverage: 100.00% of modified lines in pull request are covered.

@@            Coverage Diff             @@
##           master    #2344      +/-   ##
==========================================
+ Coverage   92.38%   92.59%   +0.20%     
==========================================
  Files          37       37              
  Lines        4454     4441      -13     
==========================================
- Hits         4115     4112       -3     
+ Misses        247      239       -8     
+ Partials       92       90       -2     

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

it is not very useful or proper way to test close. I would not mind or even notice if this would be with some other changes but alone this stands out - as artificially improving coverage. This is not what tests suppose to do or are meant for.

If you feel that close needs to be checked that it does what it is supposed to - please write proper test for it.

This is not useful at all.

I understand that you have checked the coverage of that file and decided that out 4 places (full coverage of that file) that are not covered yet you want to cover that method. Even though reader that is being called is io.NopCloser which has implementation func (nopCloser) Close() error { return nil }

but you choose not to test skipper parts which would be actually useful and provide more coverage

I am going to accept this PR if you add at least one of these options:
a) proper testcase for close that check that calling close actually calls Close on wrapped reader
b) separate testcase for skipper
c) separate testcase for panic