Remove round trim cleanup formatting

[lefticus]

Pull request overview

Refactoring to remove RoundSigDigits and TrimSigDigits while adding format() calls where reasonable.

expected failure:

• 1 regression test is now failing due to a copy-paste-error correction

No other changes expected.

[Myoldmopar]

This is going to be great. I'll keep an eye on it for when CI is happy with it and you feel like you are done with it.

[lefticus]

@Myoldmopar I'm clearly quite confused on this as of course I've been building it all along on my systems and it's not building at all on any CI. I'll see what's going on.

[Myoldmopar]

Definitely closer, but Windows is still having trouble with an ambiguous fmt symbol. Once that is resolved I'll take a final look here.

[Myoldmopar]

I'll resolve the conflicts here and get this ready to go. Thanks for fixing the fmt issue @lefticus.

[lefticus]

Sorry, I missed the conflicts. I headed out of town for a couple of days

[mitchute]

I'm not sure why CI complained on the unit test coverage build, but the AFN error message diff is definitely fixing an erroneous error message here: https://github.com/NREL/EnergyPlus/blob/develop/src/EnergyPlus/AirflowNetwork/src/Properties.cpp#L85

We'll keep an eye on CI for this next round of results and go from there.

[lefticus]

I'm not sure why CI complained on the unit test coverage build, but the AFN error message diff is definitely fixing an erroneous error message here: https://github.com/NREL/EnergyPlus/blob/develop/src/EnergyPlus/AirflowNetwork/src/Properties.cpp#L85

Yes, I saw many minor mistakes during the conversion that I left because I didn't want to cause too many diffs, but when I saw this mistake which was 100% wrong, I felt I needed to fix it (it's mentioned in the PR also)

[mitchute]

Thanks, @lefticus. This is looking really good. I'll try to get this one in right away so we don't keep having to deal with merge conflicts.

[Myoldmopar]

Very nice to see these cleaned up!

[Myoldmopar]

I see the coverage dipped very slightly because of lines changed here. I don't anticipate artificially pumping that back up or adjusting the limit, we'll leave the warning and try to bump it back up through other PRs. This looks otherwise good on CI with the one expected diff and the GPU spurious regressions. I see this is ready and will merge, thanks!