reducing some code duplication in MultisegmentWellSegments

[GitPaean]

A lot of code in the function getSurfaceVolume() and computeFluidProperties() are similar (probably started from the same) while became a little different with time.

In this PR, we put the common part of the two functions inside a third function calculatePhaseProperties() and reuse it.

And also, since the surface densities does not change, it is also put in a function while called once in the constructor instead of calling it again and again.

The function calculatePhaseProperties() is still long after the PR, which probably can be split to smaller functions.

[GitPaean]

jenkins build this please

[GitPaean]

jenkins build this please

[GitPaean]

jenkins build this please

[GitPaean]

I am marking the PR as ready for review to discuss.

[bska]

I haven't looked closely enough at the details to verify that this is in fact completely equivalent to the current master, but I'll assume it is for the purpose of this initial review. I really like the reduction in duplicated logic, so this is a marked improvement over the current master version. Moreover, it opens the path to further refactorings that would hopefully bring additional improvements to the readability.

Other than that, I have just one initial suggestion.

[GitPaean]

I haven't looked closely enough at the details to verify that this is in fact completely equivalent to the current master,

There are some small difference between the original functions getSurfaceVolume() and computeFluidProperties(), for example,

in the original getSurfaceVolume, we check the rvmax and rsmax below 0, while we did not do in the original computeFluidProperties(). Now we do it in the new function calculatePhaseProperties().

In the original getSurfaceVolume(), OpmLog::debug was used and the original computeFluidProperties() used DeferredLogger::debug(). Now we use DeferredLogger::debug() in the new function calculatePhaseProperties().

Basically, calculatePhaseProperties() unified some difference between the original functions getSurfaceVolume() and computeFluidProperties(). It changed some code slightly, while it should not cause issue.

Furthermore, when implementing calculatePhaseProperties(), I decided to clamp the rs and rv against 0.0, which should be fine.

And also, thanks for the suggestion for the private free function surfaceDensities(), I have updated the PR accordingly.

[GitPaean]

jenkins build this please