Interface from IIS.

[dourouc05]

Exact copy from the common bits in the existing solvers, except that compute_conflict! now has a !.

https://github.com/JuliaOpt/CPLEX.jl/blob/2756458f5f6c02b784e5c2fb4594c930d6febf60/src/MOI/MOI_wrapper.jl#L2745

https://github.com/JuliaOpt/Xpress.jl/blob/master/src/MOI/MOI_wrapper.jl#L3291

https://github.com/JuliaOpt/Gurobi.jl/blob/master/src/MOI_wrapper.jl#L2498

This follows jump-dev/JuMP.jl#1035 (comment).

[codecov-io]

Codecov Report

Merging #1056 into master will decrease coverage by 0.05%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1056      +/-   ##
==========================================
- Coverage   95.53%   95.48%   -0.06%     
==========================================
  Files         100      100              
  Lines       11447    11441       -6     
==========================================
- Hits        10936    10924      -12     
- Misses        511      517       +6

Impacted Files	Coverage Δ
src/attributes.jl	85.58% <ø> (ø)	⬆️
src/MathOptInterface.jl	100% <ø> (ø)	⬆️
src/Utilities/constraints.jl	85.71% <0%> (-14.29%)	⬇️
src/instantiate.jl	88.46% <0%> (-3.85%)	⬇️
src/Utilities/functions.jl	92.56% <0%> (-0.65%)	⬇️
src/Utilities/mutable_arithmetics.jl	90% <0%> (-0.1%)	⬇️
src/FileFormats/utils.jl	95.45% <0%> (-0.07%)	⬇️
src/FileFormats/LP/LP.jl	97.8% <0%> (-0.03%)	⬇️
src/FileFormats/MPS/MPS.jl	99.23% <0%> (-0.01%)	⬇️
src/FileFormats/MOF/MOF.jl	100% <0%> (ø)	⬆️

---

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 cdf87e9...de1eb13. Read the comment docs.

[dourouc05]

Thanks for the pointer.

Is it required to add tests for this? I'm not sure exactly what to test… And it looks like not all attributes are tested either.

[dourouc05]

@blegat @joaquimg Is anything else required to make this PR go on?

[blegat]

Is anything else required to make this PR go on?

Oops, I did comments on my phone but forgot to click on "Submit", you should see them now, thanks for the reminder.

[dourouc05]

GitHub should really warn you more when you do not post all your messages, but at least they were kept!

[dourouc05]

I just did this in the last commit. I also added an enum for the status, so that all the information the solver gives is accessible through MOI.

[dourouc05]

@blegat I just changed the value for the enum. However, for the indenting, you had a look at an old version :).

[odow]

I vote we: (i) keep this API as simple as possible; and (ii) rename the enum's to have CONFLICT in their name to avoid ambiguity with other enums (e.g., FEASIBLE could be a termination status, and INCLUDED and EXCLUDED are pretty broad terms).

[dourouc05]

I modified the PR accordingly. Indeed, having just one conflict simplifies things quite a bit.

[odow]

Yes, simplicity is good.

Still needs some tests and additions to the documentation.

[dourouc05]

What tests would be needed, here? I'm not sure what to test for MOI… Or you mean a few things for the solver to check their compatibility in MOIT?

For the documentation, is there anything to add in docs/src/apireference.md? Maybe a section in https://www.juliaopt.org/MathOptInterface.jl/dev/apimanual/#Advanced-1 could be useful?

[blegat]

Add the new types in the docs

[dourouc05]

Is this enough?

[blegat]

Thanks, looks good to me once CI passes.

[dourouc05]

By the way, should I include an entry for NEWS in this PR?

[blegat]

We usually collect NEWS entry separately as otherwise all PRs will conflict with each other.