Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 88.86%. Comparing base (6621015) to head (f89980b).
Report is 10 commits behind head on main.

@@            Coverage Diff             @@
##             main   #11137      +/-   ##
==========================================
- Coverage   88.91%   88.86%   -0.06%     
==========================================
  Files         183      186       +3     
  Lines       23937    23971      +34     
==========================================
+ Hits        21284    21302      +18     
- Misses       2653     2669      +16     

wow, I didn't know I fixed an issue! good job me!

The one check CI is uhappy with is code coverage, here: https://github.com/dbt-labs/dbt-core/pull/11137/checks?check_run_id=34330125418

86.17% (-0.12%) compared to 6621015

So probably what's happened here is that we've added 1 line of imports (which is not considered covered even if it is actually exercised by tests), and then removed ~10 lines of bugs & duplication. These removed lines were covered, and thus coverage has gone down (it's also possible that we're accidentally counting changelog lines as uncovered code, but this is less likely).

Anyway, the moral of the story is that code coverage is not an ultra-precise measure of code quality. We could add a few lines of no-ops to an otherwise-covered function to "fix" this, but I don't recommend this.