Added function for switching back from an inf-clojure buffer

[JasonKDarby]

This addresses #168 by creating inf-clojure--recent-buffer to keep track of the buffer the repl was switched to from. It's set on both inf-clojure and inf-clojure-switch-to-repl which hopefully means it can never be unintentionally nil. If inf-clojure-switch-to-recent-buffer is called and inf-clojure--recent-buffer IS nil the message "inf-clojure: No recent buffer known." is sent.

This only tracks the most recent buffer among all repls, it will not manage multiple references in the case of multiple repls.

I'm by no means an elisp expert, if there are improvements to be made I'm happy to hear them. I'm gonna try to take on more changes, I love having an opportunity to contribute!

---

Before submitting the PR make sure the following things have been done (and denote this
by checking the relevant checkboxes):

• The commits are consistent with our contribution guidelines
• The new code is not generating bytecode or M-x checkdoc warnings
• You've updated the changelog (if adding/changing user-visible functionality)
• You've updated the readme (if adding/changing user-visible functionality)

Thanks!

[bbatsov]

Looks good to me overall. Just mention this in the changelog and I we can merge this.

[JasonKDarby]

Thanks! I did already, if I should change it let me know.

I did notice I forgot to update inf-clojure--recent-buffer on inf-clojure-connect, I can add that later.

[JasonKDarby]

As far as I know it's good to go, let me know if there's anything else.

[dpsutton]

This attempts to track a buffer you came from. This seems inherently weak as there are only certain times we mark the buffer we came from. Seems like most times it will be nil or stale. Can we do something like what CIDER does and look through the (buffer-list) for the most recent buffer that is in a mode that descends from clojure-mode?

For reference, CIDER's version:

(defun cider-switch-to-last-clojure-buffer ()
  "Switch to the last Clojure buffer.
The default keybinding for this command is
the same as variable `cider-switch-to-repl-buffer',
so that it is very convenient to jump between a
Clojure buffer and the REPL buffer."
  (interactive)
  (if (derived-mode-p 'cider-repl-mode)
      (let* ((a-buf)
             (the-buf (let ((repl-type (cider-repl-type-for-buffer)))
                        (seq-find (lambda (b)
                                    (unless (with-current-buffer b (derived-mode-p 'cider-repl-mode))
                                      (when-let* ((type (cider-repl-type-for-buffer b)))
                                        (unless a-buf
                                          (setq a-buf b))
                                        (or (eq type 'multi)
                                            (eq type repl-type)))))
                                  (buffer-list)))))
        (if-let* ((buf (or the-buf a-buf)))
            (if cider-repl-display-in-current-window
                (pop-to-buffer-same-window buf)
              (pop-to-buffer buf))
          (user-error "No Clojure buffer found")))
    (user-error "Not in a CIDER REPL buffer")))

But the gist is (seq-find clojure-mode-predicate (buffer-list))

[JasonKDarby]

@dpsutton I was only intending to update it when it was swapped with C-c C-z / inf-clojure-switch-to-repl, is this to say the last buffer when swapped to by any means (ex C-xo)?

[dpsutton]

@JasonKDarby exactly. The current implementation would take you back to the last buffer that you used C-c C-z in, which might be a while ago, landing you in a strange buffer. It seems to me that the ideal workflow is a quick way to bounce between the repl and a source buffer, regardless of how you got there.

[JasonKDarby]

@dpsutton I think this works, I've tried it out with two windows.
Before passing '(display-buffer-reuse-window . ()) to pop-to-buffer I kept having this issue where it would change back to the correct buffer but would always do it in the 'first' window. I'm not sure why that wasn't happening before but it looks like it's fixed now.

Let me know if there's anything else to do on this. I love inf-clojure.

[dpsutton]

thanks @JasonKDarby . this is a great addition.