New Testsuite

[rs028]

Starting work on the new test suite. This PR contains some reorganization of the directories to allow the new and the old tests to run in parallel until work is completed.

To make things clearer, the command to run the build and behavioural tests has been changed from make tests to make modeltests.

[codecov]

Codecov Report

Merging #464 (eabb9cf) into master (4d72237) will decrease coverage by 10.13%.
The diff coverage is n/a.

@@             Coverage Diff             @@
##           master     #464       +/-   ##
===========================================
- Coverage   65.51%   55.37%   -10.14%     
===========================================
  Files          17       17               
  Lines        2047     2046        -1     
===========================================
- Hits         1341     1133      -208     
- Misses        706      913      +207     

Flag	Coverage Δ
build	52.23% <ø> (ø)
tests	?
unittests	31.61% <ø> (ø)

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

Impacted Files	Coverage Δ
src/interpolationFunctions.f90	0.00% <0.00%> (-36.37%)	⬇️
src/inputFunctions.f90	36.07% <0.00%> (-18.87%)	⬇️
src/constraintFunctions.f90	21.70% <0.00%> (-18.61%)	⬇️
src/solverFunctions.f90	35.80% <0.00%> (-13.59%)	⬇️
src/dataStructures.f90	50.00% <0.00%> (-8.07%)	⬇️
src/outputFunctions.f90	50.00% <0.00%> (-5.85%)	⬇️
src/argparse.f90	24.21% <0.00%> (-3.91%)	⬇️
src/atchem2.f90	85.14% <0.00%> (-2.54%)	⬇️
src/parameterModules.f90	91.30% <0.00%> (-1.45%)	⬇️
test/unit_tests/date_test.f90
... and 9 more

---

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 4d72237...eabb9cf. Read the comment docs.

[rs028]

@spco Am I reading this wrong or CodeCov is working again? I haven't touched it.

[spco]

No idea - as far as I am concerned, CodeCov's input here should just be ignored regardless! :D

[rs028]

You mean ignore it for this type of PR?

[spco]

Frankly, I don't trust it to be meaningful and/or correct on any PR!

[rs028]

Fair enough. Should we consider removing it altogether or mark it as something to be (eventually) fixed?

[rs028]

Any idea why test fails on MacOS with gfortran 8? It gives an error about sundials, but it appears that cvode is installed correctly.

[spco]

I'm not sure, but I think the error in the install cvode (MacOS) step may be something to do with it - it fail to link a simple fortran program.

-- Trying to compile and link a simple Fortran program... FAILED
WARNING: Fortran compiler not functional.
   FCMIX support will not be provided.

It looks like the image we use - macos-latest - has upgraded from 10.15 to 11 recently - see actions/runner-images#4060 .

I suspect this may have something to do with it - perhaps v11 has some different permissions, or something else.

To diagnose whether this is the cause, perhaps you could change macos-latest at

AtChem2/.github/workflows/ci.yml

Line 25 in 4d72237

os: [ubuntu-latest, macos-latest]

to macOS-10.15 (note the capitalisation) and see whether that runs successfully?

[rs028]

Looks like that. I will leave it to version 10.15 for the time being and make a note of it.
If there are no further comments I am ready to merge this and start developing the tests.