use fallback for magic requires when there's no middlware

[a13]

added a new defcustom cljr-magic-requires-fallback defcustom, so we can generate require expressions (from cljr-magic-require-namespaces) even if middleware is not available for some reason (e.g. when using babashka)

• also, I updated cljr-magic-require-namespaces with more idiomatic aliases from The Clojure Style Guide.

• The commits are consistent with our contribution guidelines

• You've added tests (if possible) to cover your change(s)

• The new code is not generating byte compile warnings (run cask exec emacs -batch -Q -L . -eval "(progn (setq byte-compile-error-on-warn t) (batch-byte-compile))" clj-refactor.el)

• All tests are passing (run ./run-tests.sh)

• You've updated the changelog (if adding/changing user-visible functionality)

• You've updated the readme (if adding/changing user-visible functionality)

[vemv]

Hi @a13 !

Thanks for the PR.

From my side I literally don't have the time in my hands to do a careful review.

My first instinct would be not to like the change because there's a lot of smartness in the middleware, and I wouldn't want users to miss out on that.

It's also the overall philosophy of CIDER - require an active connection to use most of the features.

To make matters more complicated, babasha-nrepl doesn't use the traditional nrepl middleware design, so refactor-nrepl cannot work there.

I don't have a definitive opinion on how to proceed.

Perhaps one can simply observe that most time one has at least one repl open. And that one could write .cljc files that work with bb and JVM alike, with reader conditionals. So that would work and could be documented.

Cheers - V

[vemv]

These almost certainly will conflict will real-world preferences/usages from users out there, so I'd rather leave them untouched.