Fix UndefVarError in Julia v1.0-1.3

[giordano]

Ref: https://github.com/JuliaTesting/Aqua.jl/pull/71/files#r834842406. If you have better ideas about how to more proper fix this or how to actually test this code path which is clearly untested, please take over this PR or open a new one, but at least this doesn't break the package in older versions of Julia. Not very useful to have a QUality Assurance package that breaks your tests.

CC: @Roger-luo

[codecov]

Codecov Report

Merging #73 (18e3ac5) into master (b1f583a) will decrease coverage by 0.11%.
The diff coverage is 66.66%.

@@            Coverage Diff             @@
##           master      #73      +/-   ##
==========================================
- Coverage   73.66%   73.54%   -0.12%     
==========================================
  Files          10       10              
  Lines         619      620       +1     
==========================================
  Hits          456      456              
- Misses        163      164       +1     

Flag	Coverage Δ
unittests	73.54% <66.66%> (-0.12%)	⬇️

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

Impacted Files	Coverage Δ
src/ambiguities.jl	82.85% <66.66%> (-0.80%)	⬇️

📣 Codecov can now indicate which changes are the most critical in Pull Requests. Learn more

[tkf]

Thanks! It'd be nice to actually hit this case in the test but let's register this version first. JuliaRegistries/General#57279

If you have some simple-ish examples for exercising #71, it'd be great if you can add some in https://github.com/JuliaTesting/Aqua.jl/blob/master/test/pkgs/PkgWithAmbiguities.jl (cc @Roger-luo)

[Roger-luo]

Yeah I'm not sure how to test the checks tho, but I can have a look at that morespecific function

[tkf]

That'd be great. But I'm also OK with just doing a "smoke test" to hit the code using morespecific with some MWE functions. It won't be a comprehensive test but it's still useful since morespecific is an internal function. (And many of Aqua's tests are smoke tests anyway.)

[giordano]

For what is worth, I found this error in JuliaPhysics/Measurements.jl#82, but I have little clue of where the ambiguity comes from: tests are passing on master, and that PR is only changing the signature of an internal macro, in addition to loading a new package (FiniteDifferences instead of Calculus). Can the ambiguity come from the new package?

Thanks for the quick merge! And sorry for the salty comment above, but a failure from a testing package triggered me yesterday night 🙂

[tkf]

Oh, no, that's OK :) I totally missed that my test didn't hit the code.