Relax print tests

[oxinabox]

Wether or not a space is printed changes depending on the julia version.
So we use a regex.

I also hard coded to use Int32 since that simpifies the regex

[KristofferC]

Why is this tested at all? Doesn't it just directly call a Base function?

If we really want to test it, I find it best to just interpolate the whole type, so something like:

@test ... == "3-element $(ForwardDiff.Partials{3,Int}):\n 1\n 2\n 3"

[oxinabox]

Why is this tested at all? Doesn't it just directly call a Base function?

That is a good question.
We do have
https://github.com/JuliaDiff/ForwardDiff.jl/blob/master/src/partials.jl#L228
which itself looks kinda iffy.
Are we testing that that one isn't hit?

That code was added in
1279080
along with a ton of other things.

but that test was added in 799c444
Maybe @timholy can illuminate what we were testing for.

[andreasnoack]

The show method for Partials used to fail which was why the test was added in #437. I've applied the proposal from #476 (comment).

[codecov-io]

Codecov Report

Merging #476 (317859d) into master (3fb16ac) will increase coverage by 3.94%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #476      +/-   ##
==========================================
+ Coverage   84.92%   88.86%   +3.94%     
==========================================
  Files          10       10              
  Lines         703      826     +123     
==========================================
+ Hits          597      734     +137     
+ Misses        106       92      -14     

Impacted Files	Coverage Δ
src/prelude.jl	87.50% <0.00%> (-5.10%)	⬇️
src/jacobian.jl	99.26% <0.00%> (-0.74%)	⬇️
src/hessian.jl	97.43% <0.00%> (+0.46%)	⬆️
src/gradient.jl	100.00% <0.00%> (+1.38%)	⬆️
src/apiutils.jl	100.00% <0.00%> (+2.63%)	⬆️
src/derivative.jl	100.00% <0.00%> (+3.70%)	⬆️
src/dual.jl	78.74% <0.00%> (+4.03%)	⬆️
src/partials.jl	84.21% <0.00%> (+7.19%)	⬆️
src/config.jl	87.03% <0.00%> (+19.64%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3fb16ac...317859d. Read the comment docs.