Handle punctuation

[jhchen]

Following up on #5 with something to take a look at. If the approach is acceptable perhaps it would be better to move to snake_case.ex instead since it seems like the other cases depend on that one (and update the other tests).

I tried running mix credo --strict and mix dialyzer but they errorred on the main sobolevn/recase master branch as well so I don't think this PR introduces any new cod equality issues. If you're not seeing the same I can provide more details on what I'm seeing.

Last thing (and perhaps this should be separate) but using Lodash's implementation as a guide, they also split up strings with digits ex you can try _.snakeCase("ab-cD-e0-Df-GH-I1-2j-3K-45") === "ab_c_d_e_0_df_gh_i_1_2_j_3_k_45" in the developer console in lodash.com.

[coveralls]

Coverage remained the same at 100.0% when pulling a930f6c on jhchen:master into 9185ac6 on sobolevn:master.

[sobolevn]

@jhchen Thanks for your work! I appreciate it.

What about snake_case? Is it any different from kebab?
If it should behave in the same manner - it means, that we should fix it as well (or fix just snake_case since many other cases rely on it).

I think we should discuss lodash way of handling kebab case as a separate issue. Will you, please, open it?

[coveralls]

Coverage remained the same at 100.0% when pulling f9edaac on jhchen:master into 9185ac6 on sobolevn:master.

[coveralls]

Coverage remained the same at 100.0% when pulling f9edaac on jhchen:master into 9185ac6 on sobolevn:master.

[jhchen]

Okay I updated the code and PR to implement the punctuation fix in snake case instead, updated the affected tests and opened a new issue for the digits question.

[sobolevn]

@jhchen you are awesome!

[jhchen]

Thanks! You have any thoughts on when to release a new version?

[sobolevn]

I am actually releasing it in a few hours.
Right now I am trying to fix some minor issues with the CI and stuff.

[sobolevn]

@jhchen done.