command for macroexpanding the expression

[alexander-yakushev]

I mostly stole this from SLIME because I feel a critical lack of it in nREPL.

Having little experience with elisp there could be mistakes, but I'd like to give this function a spin.

[kingtim]

Looks good, thank you!

[kingtim]

Hi Alexander,
I just pushed some changes to the macroexpansion handling. I cloned most of the popup handling from slime, so that pressing 'q' in a popup buffer will dismiss it. I also refactored the handlers so that the buffer can be read-only after the expansion. Also, I changed the pprint macroexpansion so that only the stdout is emitted, since pprint sends it output to stdout, so you don't get that extra "nil" in the output. I tested all of this locally and it seems to be working for me.

When you have some time, please give it a spin and let me know what you think?
I am also thinking of adding a C-c M-m which will do macroexpand, and change C-c C-m to do macroexpand-1, What do you think of that?

Again, thanks for the pull request!

[alexander-yakushev]

Hello Tim,
It is a great idea of yours to make buffers read-only and closeable by pressing "q". Changing C-c C-m to macroexpand-1 and introducing C-c M-m is also good (AFAIK SLIME works exactly this way).

Though I can't get the changed macroexpansion working. First of all, in the nrepl-macroexpand-last-sexp there is a form variable which should be substituted by (format command expr). But even with this change I still get the following message:

error in process filter: Wrong number of arguments: ((nrepl-input-start-mark nrepl-prompt-start-mark t) (buffer value) (with-current-buffer buffer (let ((inhibit-read-only t) (buffer-undo-list t)) (erase-buffer) (insert (format "%s" value)) (goto-char (point-min)) (indent-sexp) (font-lock-fontify-buffer)))), 1

I can't find what's wrong. Can you take a look?

[kingtim]

Sorry about that, I hadn't pushed my latest changes. Should be working now.

[alexander-yakushev]

Thanks, it works great!

Though I found an issue with macroexpanding (it's not unique to the introduced functions, but also happens in the interactive REPL). When I macroexpand some macros here's what I receive in the Messages buffer:

error in process filter: let: Args out of range: 102, 5427
error in process filter: Args out of range: 102, 5427
error in process filter: cond: Cannot decode object: 1
error in process filter: Cannot decode object: 1

Trying to macroexpand the same expression from REPLy or CCW works fine. I thought maybe you've run into this kind of error?

[kingtim]

That sounds like a bug in the bencode decoder. If you can give me a reproducible test case I can track it down.
Thanks!

[alexander-yakushev]

It's hard to provide my exact reproducible case but here's another one for the time being. I don't know if they are quite connected, but...

(str \u2190)

Should just return a string with an Unicode character. It does for the REPLy but returns this in nrepl.el:

error in process filter: Args out of range: 100, 105

Typing this one yields the same:

(str "←")

The case I'm struggling with doesn't have any Unicode characters bit I currently can't minimize it down to the single problem.

[kingtim]

I think I have tracked down the multibyte decoding issue and pushed a fix. Let me know if that fixes the above issue.