Fortitude 0.9 happiness

[Matt Shin (matthewrmshin)]

Should now passes fortitude check fully at fortitude 0.9.0.

Fix syntax errors manually.

Add config.

• target-std="f2008" (fortitude default is "f2018")
• line-length=140 (fortitude default is 102)
• Ignore various rules in deps/ directory. (Many of which were legacy source files.)

Then ran with fortitude check --fix --unsafe-fixes.

Fix remaining failures manually.

Add fortitude check to CI.

See also:

• JCSDA-internal/lfric-jedi#1276
• JCSDA-internal/ops-um-jedi#269

[Matt Shin (matthewrmshin)]

Test failing. I'm going to reverse some of the less safe changes.

[Matt Shin (matthewrmshin)]

Created a new branch and blasted away the old one with a better set of changes. Tested the changes in my local environment. Now waiting for CI. All good in CI environment.

[David Davies (DJDavies2)]

The tests pass. I am not very happy with the save and public stuff, as these are redundant and simply re-state the defaults. I know people say they like this for "clarity", but it doesn't seem useful to me: it confuses me when I see this sort of thing because i wonder why they have been stated, and if people see both

integer :: foo = 1

and

integer, save :: foo = 1

around they might start to think there is a difference between the two when there isn't.

[Matt Shin (matthewrmshin)]

Note: agree that some of the declarations may be redundant, but they allow us to satisfy these rules:

• C081 https://fortitude.readthedocs.io/en/stable/rules/initialisation-in-declaration/
• C131 https://fortitude.readthedocs.io/en/stable/rules/missing-accessibility-statement/

[David Davies (DJDavies2)]

Note: agree that some of the declarations may be redundant, but they allow us to satisfy these rules:

* C081 https://fortitude.readthedocs.io/en/stable/rules/initialisation-in-declaration/

* C131 https://fortitude.readthedocs.io/en/stable/rules/missing-accessibility-statement/

As noted, these pages make arguments related to clarity, but I disagree with these rules.