Allow building of the ufs-weather-model, WW3 pre/post execs for GFS, GEFS, SFS in the same clone of global-workflow

[aerorahul]

Description

GFS and GEFS (and now SFS) uses different compile time options for the UFS-weather-model. For the purposes of CI testing, a multi-build pipeline under Jenkins was created by @TerrenceMcGuinness-NOAA. This served well, until now. With the inclusion of SFS, a third variety of the model is being built. Under the multi-build pipeline paradigm, a second (or third) clone and build of the global-workflow is required. This adds cloning and compilation time of the global-workflow.

This PR allows compiling the ufs-weather-model in a single clone of the global-workflow. The compiled executables based on the options for GFS, GEFS, and SFS results in a model executable as gfs_model.x, gefs_model.x, and sfs_model.x. The forecast script uses the right executable.

This PR also differentiates the WW3 pre/post executables based on gfs_ or gefs_. SFS variants need to be introduced, when NET=sfs is added.
In the process of updating the WW3 pre/post executable names, it was discovered (via grep) the following are not used:

• ww3_prep
• ww3_outf
• ww3_ounf
• ww3_ounp
  These need to be confirmed by running the workflow to ensure they are indeed not used at runtime.

This PR also updates the Jenkinsfile to use the multi-build from the single location.

Type of change

• New feature (adds functionality)
• Maintenance (code refactor, clean-up, new CI test, etc.)

Change characteristics

This PR does not update any submodules.

How has this been tested?

In progress

Checklist

• Any dependent changes have been merged and published
• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have documented my code, including function, input, and output descriptions
• My changes generate no new warnings
• New and existing tests pass with my changes
• This change is covered by an existing CI test or a new one has been added
• Any new scripts have been added to the .github/CODEOWNERS file with owners
• I have made corresponding changes to the system documentation if necessary

[aerorahul]

I am going to mark this as ready for review to gather feedback on the code changes before I commence extensive testing.

[DavidHuber-NOAA]

Does SFS get its own executable or does it use the GEFS executable?

[aerorahul]

Eventually, it will. For now, SFS_POST=YES will use the gefs prefix.

[DavidHuber-NOAA]

Under what circumstances should ufs_model.x be used? Since both config.fcsts reference gfs_model.x, gefs_model.x, or sfs_model.x, it may cause issues to have ufs_model.x as the default. I'd suggest at least a warning be raised if build_ufs.sh is called without specifying an executable to build.

[aerorahul]

I can completely remove ufs_model.x if that is the source of confusion. ufs_model.x is going to be used if EXEC_NAME is not provided.

[DavidHuber-NOAA]

I'm OK with a default, but I think it would make sense to have a default that is useable. How about defaulting to gfs_model.x?

[WalterKolczynski-NOAA]

Suggested change

	if [[ -d "${HOMEgfs}/sorc/ufs_model.fd/WW3/install/pdlib_OFF" ]]; then # installed with PDLIB=OFF (Used in GEFS)
	for ww3exe in "${ww3_exes[@]}"; do
	[[ -s "gefs_${ww3exe}.x" ]] && rm -f "gefs_${ww3exe}.x"
	${LINK_OR_COPY} "${HOMEgfs}/sorc/ufs_model.fd/WW3/install/pdlib_OFF/bin/${ww3exe}" "${HOMEgfs}/exec/gefs_${ww3exe}.x"
	done
	fi
	if [[ -d "${HOMEgfs}/sorc/ufs_model.fd/WW3/install/pdlib_ON" ]]; then # installed with PDLIB=ON (Used in GFS, may be SFS later)
	for ww3exe in "${ww3_exes[@]}"; do
	[[ -s "gfs_${ww3exe}.x" ]] && rm -f "gfs_${ww3exe}.x"
	${LINK_OR_COPY} "${HOMEgfs}/sorc/ufs_model.fd/WW3/install/pdlib_ON/bin/${ww3exe}" "${HOMEgfs}/exec/gfs_${ww3exe}.x"
	done
	fi
	declare -A wave_systems
	wave_systems["gfs"]="pdlib_ON"
	wave_systems["gefs"]="pdlib_OFF"
	wave_systems["sfs"]="pdlib_OFF"
	for sys in ${!wave_systems[@]}; do
	build_loc="${wave_systems[${sys}]}"
	if [[ -d "${HOMEgfs}/sorc/ufs_model.fd/WW3/install/${build_loc}" ]]; then
	for ww3exe in "${ww3_exes[@]}"; do
	target_ww3_exe="${sys}_${ww3exe}.x"
	if [[ -s "${target_ww3_exe}" ]]; then
	rm -f "${target_ww3_exe}"
	fi
	${LINK_OR_COPY} "${HOMEgfs}/sorc/ufs_model.fd/WW3/install/${build_loc}/bin/${ww3exe}.x" "${HOMEgfs}/exec/${target_ww3_exe}"
	done
	fi
	done

[aerorahul]

There is a typo in the suggestion. Line 310

[WalterKolczynski-NOAA]

Fixed

[JessicaMeixner-NOAA]

What happens when GFS/GEFS/SFS change their wave option? Sorry I didn't catch this before, but seeing this brought it to my attention. This should be tied to the options you pass the build, not assumptions here.. and while yes, if someone did that there would be other changes to make in terms of grid choices, but this seems pretty hidden and potentially arbitrary.