Add initialize_constituents scheme

[peverwhee]

Originator(s): peverwhee

Summary (include the keyword ['closes', 'fixes', 'resolves'] and issue number): Adds new (non-portable) initialize_constituents scheme (.F90 and .meta). This scheme will be used to instantiate all constituents in the file (used to validate schemes that need the full constituent array but do not modify all of the individual constituents). Added to new "test_schemes" directory.

Describe any changes made to the namelist: N/A

List all files eliminated and why: N/A

List all files added and what they do:
A test/test_schemes/initialize_constituents.F90
A test/test_schemes/initialize_constituents.F90

• add initialize_constituents scheme

List all existing files that have been modified, and describe the changes:
(Helpful git command: git diff --name-status development...<your_branch_name>)

List any test failures: none

Is this a science-changing update? New physics package, algorithm change, tuning changes, etc? No

[nusbaume]

Thanks @peverwhee for getting this test helper scheme developed! I just had a few change requests.

[nusbaume]

Everything looks good to me now, thanks!

[cacraigucar]

I also have been testing a variation of this routine in the ZM conversion

[cacraigucar]

I think I prefer the old errmsg statements which said what field had the allocate error. Looking through the internet for an example of what errmsg might contain, I saw these two error messages:
errmsg = Attempt to allocate an allocated object
errmsg = insufficient virtual memory
The errmsg is probably compiler dependent, but the important thing is that neither indicate what field had the error being allocated. If you'd like to include the Fortran errmsg, perhaps you can concatentate the module/variable in front of the returned errmsg.