Bugfix 2596 main v11.1 rpath compilation

[jprestop]

Expected Differences

• Do these changes introduce new tools, command line arguments, or configuration file options? [No]
  
  If yes, please describe:
  

• Do these changes modify the structure of existing or add new output data types (e.g. statistic line types or NetCDF variables)? [No]
  
  If yes, please describe:
  

Pull Request Testing

• Describe testing already performed for these changes:
  
  Tested compilation of METv11.1.0-rc1 and its dependent libraries on a MacOS (narya) using the gcc version 12.3.0 compiler and the 11.3.0 compiler. I got a successful compilation using the 11.3.0 compiler, but not the 12.3.0 compiler (new issue to be created for this problem that seems unrelated to the script).
  After reproducing the problem described in the issue, I was able to resolve the problem with these changes to the script and got a successful compilation and a successful run of "make test". I also tested the changes in the script with METv11.1.0-rc1 and its dependent libraries the gcc version 12.1.0 compiler on cheyenne and got a successful compilation and a successful run of "make test".

• Recommend testing for the reviewer(s) to perform, including the location of input datasets, and any additional instructions:
  
  Review the files that changed and ensure that all tests pass. You could compile on a Mac using gcc if desired, noting that MET won't compile, but that the errors seems unrelated to this script. (new issue to be created for this problem that seems unrelated to the script).
  Note that @HathewayWill also tested the script, where the compilation of MET also failed, however the errors he received also seem unrelated to the script. The errors he received were different from mine and were noted in the issue description for Bugfix: Problem compiling MET on MacOS using GCC 12.3.0 #2615.

• Do these changes include sufficient documentation updates, ensuring that no errors or warnings exist in the build of the documentation? [Yes]

• Do these changes include sufficient testing updates? [Yes]

• Will this PR result in changes to the test suite? [No]
  
  If yes, describe the new output and/or changes to the existing output:
  

• Please complete this pull request review by [20230726].
  

Pull Request Checklist

See the METplus Workflow for details.

• Review the source issue metadata (required labels, projects, and milestone).
• Complete the PR definition above.
• Ensure the PR title matches the feature or bugfix branch name.
• Define the PR metadata, as permissions allow.
  Select: Reviewer(s)
  Select: Organization level software support Project or Repository level development cycle Project
  Select: Milestone as the version that will include these changes
• After submitting the PR, select Development issue with the original issue number.
• After the PR is approved, merge your changes. If permissions do not allow this, request that the reviewer do the merge.
• Close the linked issue and delete your feature or bugfix branch from GitHub.

[georgemccabe]

There is a lot of duplication here. I think this could be simplified by creating a variable to store the the sed command to use based on the OS.

After the case statement (around line 356):

if [[ $machine == "Mac" ]]; then
    sed_inline="sed -i ''"
else
    sed_inline="sed -i''"
fi

then wherever you need to run a sed command, use the variable instead of "sed" e.g.

Before:
sed -i '' "s|INC=.*|INC=-I${LIB_DIR}/include -I${LIB_DIR}/include/jasper|g" makefile
After:
$sed_inline "s|INC=.*|INC=-I${LIB_DIR}/include -I${LIB_DIR}/include/jasper|g" makefile

[jprestop]

Thanks @georgemccabe for this simplification! I'll change the code and retest to ensure the functionality still works. I'll follow up.

[georgemccabe]

This could be simplified with a variable:

  configure_args=""
  if [[ $machine == "Mac" ]]; then 
    configure_args="LIBS=\"${LIBS} -lhdf5_hl -lhdf5 -lz\""
  fi
  run_cmd "./configure --prefix=${LIB_DIR} LDFLAGS=-L${LIB_DIR}/lib CPPFLAGS=-I${LIB_DIR}/include LIBS=\"${LIBS} ${configure_args}\"> netcdf-cxx.configure.log 2>&1"

[jprestop]

Thanks @georgemccabe for this simplification! I'll change the code and retest to ensure the functionality still works. I'll follow up.

[georgemccabe]

Why was this added? I don't see it used anywhere in the script.

[jprestop]

Thanks @georgemccabe. I'll take a closer look at the other comments tomorrow, but I added the COMPILER_MAJOR_VERSION and COMPILER_MINOR_VERSION because it was originally thought that the problem lied with gcc 12.3.0. It didn't, however, I could see potentially using these in the future so I left them in.

[georgemccabe]

I made a few suggestions in-line to clean up the new logic to reduce duplicate code.

[jprestop]

Thanks @georgemccabe. I made the suggested changes with a tweak and retested which results in the same outcome. I'm going to remove @JohnHalleyGotway as a reviewer, since you have already been looking at this and will request a re-review from you.

[georgemccabe]

Why was LIB_Z removed from the LDFLAGS here? Does it need to be set there if LIB_Z is set before running the script?

[jprestop]

@georgemccabe Thank you for you attention to detail. That was removed during testing and never put back. I'll add that back in and then will re-request a review.

[georgemccabe]

Looks good. Thanks for making the changes.

Line 595 doesn't need to be indented but that is not important.

Line 693: I'm not sure if exporting LDFLAGS to an empty string for MacOS is needed or not, but if it works it works.

I approve.

[jprestop]

@georgemccabe

Line 595 doesn't need to be indented but that is not important.

While not important, I want to be consistent, so I removed the indentation. Thanks for noting that.