False positives in the `snake_case` rule

[a-frantz]

In workflows we have multiple variables that end in the suffix _md5. These get flagged by the snake_case rule as needing to be changed to *_md_5. I'd call that a bug in the code which needs to be addressed.

From the convert_case docs:

By default, to_case will split along a set of default word boundaries, that is

space characters , underscores _, hyphens -, changes in capitalization from lowercase to uppercase aA, adjacent digits and letters a1, 1a, A1, 1A, and acroynms AAa (as in HTTPRequest).

adjacent digits and letters a1, 1a, A1, 1A, is the specific issue here.

---

We also have variables that contain the substring dbSNP that get flagged by snake_case as needing to be db_SNP. Any sane to_snake_case algorithm would make this "correction" so it's not a "false positive". However the camelCase string dbSNP is meaningful as it refers to the NIH's dbSNP database, and changing the variable substring to db_SNP would be obfuscating the connection to the NIH's dbSNP. This is a special case where the wdl crates need to be configurable so that they ignore specific instances of a lint. We don't want to completely disable the snake_case rule, just silence it in this instance.

We should add a feature that allows lints to be ignored for a specific line (File dbSNP_vcf. # ignore: snake_case or something)