node: add shebang rewriting

[samford]

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same change?
• Have you added an explanation of what your changes do and why you'd like us to include them?
• Have you written new tests for your changes? Here's an example.
• Have you successfully run brew style with your changes locally?
• Have you successfully run brew typecheck with your changes locally?
• Have you successfully run brew tests with your changes locally?

---

Formulae that depend on node sometimes contain files that use a shebang like #!/usr/bin/env node and this can lead to issues when the node in a user's environment isn't brewed node.

For example, some node modules are compiled when the formula is built but if the user's node is a different major version than brew's node, the differing NODE_MODULE_VERSION can produce an error when certain parts of the application are used. The formula may build and test fine and the issue may only become apparent when more of the application is exercised (e.g., Homebrew/homebrew-core#139389).

This adds a Language::Node::Shebang module, which allows us to use rewrite_shebang detected_node_shebang, ... in formulae to address this type of issue. I borrowed from the existing Perl and Python implementations but omitted logic that doesn't seem to apply to node (though do let me know if I've overlooked anything).

[apainintheneck]

Makes sense to me. Might be good to mention it here in the docs as well.

• https://docs.brew.sh/Formula-Cookbook#rewriting-a-script-shebang

[MikeMcQuaid]

Looks great, nice work @samford! One nit.

[samford]

Might be good to mention it here in the docs as well.

• https://docs.brew.sh/Formula-Cookbook#rewriting-a-script-shebang

Thanks for the heads up. The docs use non-exhaustive wording (i.e., "such as Python or Perl"), so that's technically still correct. I can update the text if we think it will be beneficial (e.g., "such as Python, Perl or Node") but I'm not sure if it would be helpful or just more verbose.

---

I've pushed a commit that incorporates the above feedback and extracts the length of the longest matching shebang string into a constant. I also created a constant for the regex, so it's not recreated every time #node_shebang_rewrite_info is called.

While I was at it, I reworked the existing Shebang modules for Perl and Python to use the previously-mentioned setup (adding constants and also type signatures) and to better align with each other (e.g., adding a perl_shebang_rewrite_info method). Lastly, I expanded the tests for the Shebang modules to hit 100% coverage. That said, if we would prefer to handle the Perl/Python changes in a separate PR, I can pare this down to the Node changes.

[apainintheneck]

How worried should we be about name clashes from these mixins? They now all have constants with the same names so if you included multiple of them in your formula (something that is technically possible but not present at the moment in core) at least one of them would get overwritten, right?

I also don't think we should really be that worried about the regexes being redefined each time since it doesn't seem to be a hot codepath and should really only be needed once for a handful of formula (67 by my count) during bottling. I guess what I'm saying is that I'm partial to how it was before.

[apainintheneck]

This needs to be changed so that the mixins can't overwrite each others constants. Probably the easiest things to do would be either prefix the constant with the language like PYTHON_SHEBANG_REGEX or use inline regexes like before.

[apainintheneck]

Thanks for updating it!

[MikeMcQuaid]

This PR is AWESOME 😍. Really love how you've not only added new functionality here @samford but also cleaned up a bunch of related additional code.