Rewrite tests to use Base.Test

[lbenet]

Currently, some tests (involving some functions) do not pass due
to CRlibm

[lbenet]

Addresses #193

ITF1788 files not yet modified, since this is better to using directly the tools provided by ITF1788...

[lbenet]

Incidentally, these changes will only be useful once we leave 0.4 without support.

[dpsanders]

Great, thanks! On Julia 0.4 you need using BaseTestNext to use @testset.

[lbenet]

What about @compat?

[dpsanders]

Not sure what you are asking about @compat?

[dpsanders]

https://github.com/JuliaCI/BaseTestNext.jl

[coveralls]

Coverage decreased (-1.3%) to 89.329% when pulling c0b4b24 on Base.Test into 4af546d on master.

[lbenet]

I though that @compat using Base.Test could save us replacing that for

if VERSION >= v"0.5.0-dev+7720"
    using Base.Test
else
    using BaseTestNext
    const Test = BaseTestNext
end

in each file. But i couldn't find it.

How do we deal with the REQUIRES in this case?

[dpsanders]

I think you only need it in runtests.jl, since everything is included from there?

…

On 16 Feb 2017 4:59 p.m., "Luis Benet" ***@***.***> wrote: I though that @compat using Base.Test could save us replacing that for if VERSION >= v"0.5.0-dev+7720" using Base.Testelse using BaseTestNext const Test = BaseTestNextend in each file. But i couldn't find it. How do we deal with the REQUIRES in this case? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#205 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AALtTsQhKdf6aMG4nivYAbUs5MAluJ5Hks5rdNTHgaJpZM4MDm_M> .

[lbenet]

No, though that could indeed be a solution; each included file has the using command.

[dpsanders]

If I remember correctly, each file has it so that they can be run independently if needed. But the tests should work on 0.4 just adding the basetestnext in runtests.jl

…

On 16 Feb 2017 5:27 p.m., "Luis Benet" ***@***.***> wrote: No, though that could indeed be a solution; each included file has the using command. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#205 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AALtTn_OnlHDeV6PbnrYt-vHD-n_CYEKks5rdNtJgaJpZM4MDm_M> .

[codecov-io]

Codecov Report

Merging #205 into master will decrease coverage by -0.21%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #205      +/-   ##
==========================================
- Coverage   90.65%   90.45%   -0.21%     
==========================================
  Files          23       23              
  Lines         985      985              
==========================================
- Hits          893      891       -2     
- Misses         92       94       +2

Impacted Files	Coverage Δ
src/ValidatedNumerics.jl	100% <ø> (ø)	✅
src/intervals/arithmetic.jl	98.64% <ø> (-1.36%)	❌

---

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 4af546d...402b9c4. Read the comment docs.

[coveralls]

Coverage decreased (-0.9%) to 89.736% when pulling bc4a242 on Base.Test into 4af546d on master.

[coveralls]

Coverage decreased (-0.2%) to 90.457% when pulling 402b9c4 on Base.Test into 4af546d on master.

[dpsanders]

Looking good! Is this ready to merge?

[lbenet]

Not yet; on the one hand, there is something weird with Julia 0.6; I'm also trying to transform the ITF1788 tests. I'll let you know when I think it is ready.

[dpsanders]

Don't worry about 0.6. I finally got tests to pass on the fix_radius_0.6 branch that requires us to change from FixedSizeArrays.jl to StaticArrays.jl.

I suggest we merge this and the itf1788 tests can be in a separate pr.

[lbenet]

Ok. Then go ahead.

[lbenet]

One of the tests that is not passing in v0.6 yields the following output

Operations on boxes: Error During Test
  Test threw an exception of type ErrorException
  Expression: 2A == IntervalBox(2 .. 4, 6 .. 8)
  type TypeName has no field primary
  Stacktrace:
   [1] promote_type_nested(::Type{T} where T<:Tuple) at /Users/benet/.julia/v0.6/FixedSizeArrays/src/core.jl:181
   [2] construct_similar(...) at /Users/benet/.julia/v0.6/FixedSizeArrays/src/core.jl:230
   [3] macro expansion at /Users/benet/.julia/v0.6/FixedSizeArrays/src/mapreduce.jl:81 [inlined]
   [4] map at /Users/benet/.julia/v0.6/FixedSizeArrays/src/mapreduce.jl:119 [inlined]
   [5] *(::Int64, ::ValidatedNumerics.IntervalBox{2,Float64}) at /Users/benet/.julia/v0.6/FixedSizeArrays/src/ops.jl:95
   [6] macro expansion at /Users/benet/.julia/v0.5/ValidatedNumerics/test/multidim_tests/multidim.jl:13 [inlined]
   [7] macro expansion at ./test.jl:853 [inlined]
   [8] anonymous at ./<missing>:?

I guess this is related to the new type nomenclature...

I haven't checked if there are other problems.

[dpsanders]

I'm not sure what exactly it's related to, but it's fixed by the latest StaticArrays.jl update.

[dpsanders]

Thanks!

[dpsanders]

Do you have the ITF1788 tests switched to @testset for the files that are already included (not the new ones)?