New rule: DisallowedDeclarationName

[a-frantz]

Credit to @adthrasher for the idea proposed in a comment on #133

We have DisallowedInputName and DisallowedOutputName. Those should remain unchanged, but there should be a new rule that checks all Decls (IO and private) for "type prefixes/suffixes". e.g. File gtfffile would be bad and so would Int my_int

[adthrasher]

One consideration (via Mike Rusch), what about something like bcftools annotate (https://samtools.github.io/bcftools/bcftools.html#annotate). It has parameters that can be specified as a string or as a file. So the file suffix might make sense in that case to distinguish.

-r, --regions chr|chr:pos|chr:from-to|chr:from-[,…​]
see [Common Options](https://samtools.github.io/bcftools/bcftools.html#common_options)

-R, --regions-file file
see [Common Options](https://samtools.github.io/bcftools/bcftools.html#common_options)

[Gyan-max]

Hi @a-frantz and @adthrasher
I'd love to work on this issue.
Here is my plan:

1. Review the existing rules (DisallowedInputName, DisallowedOutputName) to understand how they work.
2. Add a new DisallowedDeclarationName rule to catch bad names in declarations.
3. Write tests for valid/invalid cases.
4. update the docs to explain the new rule.

I would like to confirm if this plan is good. Please let me know if there is any additional guidelines i should follow.

[claymcleod]

This looks good. Let us know if you have any questions!

[Gyan-max]

Hey @claymcleod , I'm trying to test my new DisallowedDeclarationName rule but running into issues with the shellcheck tests failing. These failures seem unrelated to my code - they're showing different shellcheck diagnostics than what's expected in the test files. Is there a way to run just my specific test without the shellcheck tests interfering? Or should I ignore these errors for now?

[Gyan-max]

Although I'm submitting a PR for this, please review it also and suggest to me if any changes are needed.