simpler parm/makefile

[SamuelTrahanNOAA]

Replaces parm/makefile with a simpler one that will be easier to maintain.

• fixes The parm/makefile is overcomplicated and error-prone #787

Regression tests will not test these changes. Instead, you must regenerate the text files. Ensure that both make clean and make work. The resulting txt files should be identical to the ones in the repository.

This script will do that for you. If it prints "SUCCESS" at the end, then the test passed. Otherwise, the test failed.

Expand for test script

#! /bin/bash

# Ensure the test fails if any step failed.
set -xue

# This is the list of all txt files generated by the old makefile.
all_txt_files='postxconfig-NT-GFS.txt postxconfig-NT-GFS-GOES.txt postxconfig-NT-GFS-ANL.txt
postxconfig-NT-GFS-F00.txt postxconfig-NT-GFS-FLUX.txt postxconfig-NT-GFS-FLUX-F00.txt
postxconfig-NT-NMM.txt postxconfig-NT-NGAC.txt postxconfig-NT-GEFS.txt postxconfig-NT-GEFS-F00.txt
postxconfig-NT-hafs_sat.txt postxconfig-NT-hafs_nosat.txt postxconfig-NT-hafs.txt
postxconfig-NT-GFS-TWO.txt postxconfig-NT-GFS-F00-TWO.txt postxconfig-NT-hrrr.txt
postxconfig-NT-rap.txt postxconfig-NT-fv3lam_rrfs.txt postxconfig-NT-UFS-aerosol.txt
postxconfig-NT-UFS-aerosol-F00.txt postxconfig-NT-GEFS-aerosol.txt postxconfig-NT-GEFS-F00-aerosol.txt
postxconfig-NT-AQM.txt'

# Delete all txt files that were generated by the old makefile
rm -f $all_txt_files

# Ensure that all those files were deleted.
for x in $all_txt_files ; do
    ! test -s "$x"
done

# Regenerate all txt files using 8 make threads.
make -j 8

# Confirm there are no changes.
! ( git status --porcelain -uno | grep . )

# Delete all txt files that were generated by the old makefile.
make clean

# Ensure that all those files were deleted.
for x in $all_txt_files ; do
    ! test -s "$x"
done

# Regenerate all txt files using 8 make threads.
make -j 8

# Confirm there are no changes.
! ( git status --porcelain -uno | grep . )

echo SUCCESS

[SamuelTrahanNOAA]

There is no use in running the regression tests on this PR, and they wouldn't test the changes anyway. The only file that should change is the makefile, which is never used when running UPP. As long as the makefile regenerates the txt files correctly, the changes are safe.

[WenMeng-NOAA]

@SamuelTrahanNOAA Thanks for your help on optimizing UPP scripts. The testing instructions you provided is working for me.

[WenMeng-NOAA]

The testing of recreating UPP control files in .txt is successful.

[WenMeng-NOAA]

This PR is ready for merging.