Integration tests

[XachaB]

Per #83

[julesbouton]

I fixed the variables issue, but we have other problems.

[XachaB]

I think you added a guard to avoid importing overabundant patterns -- but I do it for the lattices stuff (just not for entropies). I'll remove that and try to fix :)

[julesbouton]

Oh true, good catch. In between I managed to break the patterns. We really needed these integration tests.

[XachaB]

We also do need proper unit tests (remember the "dire testing situation"), but I just don't have the time for now, so this will have to suffice temporarily

[julesbouton]

I think we fixed simultaneously but differently the patterns issue. Everything should be back now.

Yes propert unit tests will be for another time. This is already a lot of cleaning up / improving the code and how Qumin works.

[XachaB]

No it's broken now, I think it was working this afternoon ?

[julesbouton]

No it's broken now, I think it was working this afternoon ?

My last commit fixed it. But the entropies are broken apparently.

[XachaB]

Try this, it fails:

qumin hydra.run.dir=tests/scripts/test_run_pat data=tests/data/TestPackage/test.package.json pats.overabundant.keep=True pats.defective=True action=patterns

[julesbouton]

Try this, it fails:

qumin hydra.run.dir=tests/scripts/test_run_pat data=tests/data/TestPackage/test.package.json pats.overabundant.keep=True pats.defective=True action=patterns

That's my fault then, sorry

[julesbouton]

Sorry, everything is going so quickly that I didn't notice in time that you already fixed the patterns thing. And my fix was probably worse than the problem.

[XachaB]

Lol, no worries, working on the same branch simultaneously is bound to lead to madness. I just fixed something that was my fault (and caused the regression in entropies). I have to go now, so if you're not off, you've got the branch for yourself for a while :)

[julesbouton]

Passed the tests! 🥳