Fourthed, cheers! z

@bbatsov I would like to try submit a PR to fix this, maybe in two/three weeks.

After I start emacs, my Messages buffer contains the following smart-mode-lineerror message:

Error during redisplay: (eval (sml/generate-minor-modes)) signaled (wrong-type-argument number-or-marker-p \10) [24 times]

and I see the same message again when I move the cursor in a new buffer etc.

This is with the latest smart-mode-line on melpa, GNU Emacs 27.2 (build 1, x86_64-w64-mingw32) of 2021-03-26, and Windows 10.

My smart mode line configurations are:

(use-package smart-mode-line 
  :config
  (setq sml/theme nil) ; don't change existing modeline faces
  (sml/setup))

and I have the following emacs customizations:

'(sml/mode-width '\10)
 '(sml/modified-char "•")
 '(sml/name-width 30)
 '(sml/position-percentage-format nil)
 '(sml/prefix-face-list
   '((":SU:" sml/sudo)
     (":G" sml/git)
     (sml/projectile-replacement-format sml/projectile)
     ("" sml/prefix)))
 '(sml/prefix-regexp nil)
 '(sml/replacer-regexp-list nil)
 '(sml/shorten-directory t)
 '(sml/show-file-name nil)
 '(sml/vc-mode-show-backend nil)
'(sml/modified ((t (:inherit sml/not-modified :foreground "firebrick" :weight bold))))

No progress. I've been pretty busy lately.

@bbatsov Any news for this?

Same issue here.

@bbatsov Do you think we should include this workaround in faq?

In gnu guix with Java OpenJDK installed via the packages "openjdk" "openjdk:doc" "openjdk:jdk" info and stacktrace causes this exceptions:

Caused by: java.lang.ClassNotFoundException: jdk.javadoc.doclet.Doclet
	at java.base/java.net.URLClassLoader.findClass(Unknown Source)

My cloure code stays the same. On fedora it is working, on guix it throws exceptions.

Is there a way to find out if javadoc is installed? Any pointer to what could be missing is appreciated!

Expected behavior

:info and :stacktrace ops work

Environment & Version information

guix with

Clojure version

latest stable

Java version

from guix package

Operating system

linux

I solved the issue. In gnu guix the "openjdk" package removes javadoc libs, one has to use ONLY the package "openjdk:jdk".

In gnu guix with Java OpenJDK installed via the packages "openjdk" "openjdk:doc" "openjdk:jdk" info and stacktrace causes this exceptions:

Caused by: java.lang.ClassNotFoundException: jdk.javadoc.doclet.Doclet
	at java.base/java.net.URLClassLoader.findClass(Unknown Source)

My cloure code stays the same. On fedora it is working, on guix it throws exceptions.

Is there a way to find out if javadoc is installed? Any pointer to what could be missing is appreciated!

Expected behavior

:info and :stacktrace ops work

Environment & Version information

guix with

Clojure version

latest stable

Java version

from guix package

Operating system

linux

Thanks for validate.el. I use it throughout in my config.

Unsure if I'm misuing validate-setq or if a definition needs to be updated.

Definition:

(defcustom company-show-numbers nil
  "If enabled, show quick-access numbers for the first ten candidates."
  :type '(choice (const :tag "off" nil)
                 (const :tag "left" 'left)
                 (const :tag "on" 't)))

Tried:

(validate-setq company-show-numbers 't)

Error:

validate-value: Looking for `(choice (const :tag "off" nil) (const :tag "left" ...) (const :tag "on" ...))' in `t' failed because:
  all of the options failed
    Looking for `(const :tag "off" nil)' in `t' failed because:
      not the expected value
    Looking for `(const :tag "left" 'left)' in `t' failed because:
      not the expected value
    Looking for `(const :tag "on" 't)' in `t' failed because:
      not the expected value

Also tried:

(validate-setq company-show-numbers t)

Error:

validate-value: Looking for `(choice (const :tag "off" nil) (const :tag "left" ...) (const :tag "on" ...))' in `t' failed because:
  all of the options failed
    Looking for `(const :tag "off" nil)' in `t' failed because:
      not the expected value
    Looking for `(const :tag "left" 'left)' in `t' failed because:
      not the expected value
    Looking for `(const :tag "on" 't)' in `t' failed because:
      not the expected value

I was just looking for this issue yesterday. It would be very useful to jump to test source lines from exceptions thrown in tests.

I'm having trouble getting re-jump.el to work in context of a re-frame app. It might have to do with cider-symbol-at-point.

re-jump.el provides emacs navigation for re-frame projects. It helps jump to re-frame effect registrations, which are basically namespace-qualified keywords. Ref. its README for an example https://github.com/oliyh/re-jump.el

re-jump relies on cider-symbol-at-point to give it a fully-qualified keyword, which feeds into a regex-based search, to jump to definition.

And I think cider-find-keyword is broken for me, for the same reason.

Expected behavior

Using re-jump.el as of SHA 443ddfa33dd2ae593cc0a013d16fff21f2afd925

cider-symbol-at-point should return a fully namespace qualified keyword to re-jump.el when we call M-x re-frame-jump-to-reg or M->.

Actual behavior

cider-symbol-at-point returns an empty string to re-jump.el.

At one point I thought refactor-nrepl might be interfering, but I am able to reproduce the problem without refactor-nrepl too.

Steps to reproduce the problem

Setting up an exact replica of my project may be too much. I suspect this is a general problem that will reproduce within any clojurescript project, given the software versions documented below.

A bit of debugging inside cider-util.el/cider-symbol-at-point suggests the function is able to extract the keyword with (thing-at-point 'symbol). But, it appears to not macroexpand the keyword in a clojurescript nREPL.

ref: https://github.com/clojure-emacs/cider/blob/master/cider-util.el#L134

(when (string-match-p "^::.+" str)
  (setq str (or (ignore-errors (cider-sync-request:macroexpand "macroexpand-1" str)) "")))

I found macroexpanison returns nil. I observed this with println style debugging and with edebug step through.

The regex string-match appears to work, but then, cider-sync-request:macroexpand appears to return nil.

When stepping through, I noticed str seems to get bound to a fontified structure.

Result: #("::my-target-event-registration" 0 2 (help-echo cider--help-echo fontified t cider-locals #1=("slug" "_" "_") face clojure-keyword-face) 2 28 (help-echo cider--help-echo fontified t cider-locals #1# face clojure-keyword-face))

I tried to remove font lock, and re-tried step through. Now str gets set to "::my-target-event-registration", but still no dice... macroexpansion still returns nil.

        ;; remove font-locking
        (setq str (substring-no-properties str))
        ;; resolve ns-aliased keywords
        (when (string-match-p "^::.+" str)
          (setq str (or (ignore-errors (cider-sync-request:macroexpand "macroexpand-1" str)) "")))

This is the limit of my emacs-fu, I'm afraid... I'll need help to debug this further to find a fix.

Looking through commit history, I noticed the above keyword-expansion behaviour was added here (starting line 132 in cider-util.el): https://github.com/clojure-emacs/cider/pull/2826/files#diff-21ccb709cb3dcff7d2b47a250e64c363ca182b5a3155ab034ce6743fec58fa03R132

Note: I start the repl in my terminal and connect from Emacs with cider-connect-cljs. The exact REPL invocation is:

lein with-profile cljs update-in :dependencies conj '[cider/cider-nrepl "0.25.3"]' -- run -m shadow.cljs.devtools.cli --npm watch app

Environment & Version information

CIDER version information

;; CIDER 1.0.0snapshot (package: 20201018.957), nREPL 0.7.0
;; Clojure 1.10.1, Java 11.0.8

;; ClojureScript REPL type: shadow-select
;; ClojureScript REPL init form: (do (require '[shadow.cljs.devtools.api :as shadow]) (shadow/nrepl-select :app))

Middleware:

• refactor-nrepl 2.5.0
• cider/cider-nrepl 0.25.3

Lein/Boot version

• Leiningen 2.9.4 on Java 11.0.8 OpenJDK 64-Bit Server VM
• shadow-cljs 2.10.21 (invoked via yarn + npm)

Emacs version

GNU Emacs 26.3 (build 2, x86_64-pc-linux-gnu, GTK+ Version 3.24.13) of 2019-12-24

Operating system

Ubuntu 20.04.1 LTS

Use the template below when reporting bugs. Please, make sure that you're running the latest stable CIDER and that the problem you're reporting hasn't been reported (and potentially fixed) already.

Remove all of the placeholder text in your final report!

Expected behavior

Start off by visiting a file in https://github.com/Gonzih/cljs-electron/ I chose https://github.com/Gonzih/cljs-electron/blob/master/ui_src/ui/core.cljs and executing M-x cider-jack-in-clj&cljs M-. on 'def' in https://github.com/Gonzih/cljs-electron/blob/master/ui_src/ui/core.cljs should take me to the source code of def.

Actual behavior

It doesn't work. Instead cider-find-var (M-.) prompts "Symbol: " in the minibuffer with no prefilled suggestions. When I write in 'def' and hit return it tells me no cljs REPLs are in current session.

Steps to reproduce the problem

Clone the repository: https://github.com/Gonzih/cljs-electron/ and then try to M-. any function/variable/macro in ui_src/ui/core.cljs.

Environment & Version information

I tried updating the dependencies because someone suggested that in a totally different cljs project that would fix it.

Here it is:

:dependencies [[org.clojure/clojure "1.10.1"]
                [org.clojure/clojurescript "1.10.758"]
                [figwheel "0.5.19"]
                [figwheel-sidecar "0.5.19"]
                [cider/piggieback "0.5.1"]
                [reagent "0.8.1"]
                [ring/ring-core "1.7.1"]]
 :plugins [[lein-cljsbuild "1.1.7"]
           [lein-figwheel "0.5.20"]
           [lein-cooper "1.2.2"]]

CIDER version information

;; CIDER 1.0.0snapshot (package: 20201125.1146), nREPL 0.8.3
;; Clojure 1.10.1, Java 11.0.9.1

Lein/Boot version

skip

Emacs version

27.1

Operating system

Ubuntu 20.10

Expected behavior

eldoc does not disappear

Actual behavior

eldoc disappears

Steps to reproduce the problem

Invoke a loaded function and in any argument position, open a map, vector - eldoc disappears.

I suppose this problem has always been present. Just today I noticed that it already became a habit of mine to skip out of the current structure to see eldoc again. Especially when invoking a function that declares a destructured map arg, its an interruption not to see the :keys list anymore.

Environment & Version information

Cider version

0.26.0

Emacs version

26.3

Expected behavior

Cider jumps to the correct source when clicking a stack trace frame.

Actual behavior

Cider jumps to clojure.core/fn as source of stack frames for top level anonymous functions.

Steps to reproduce the problem

(def ^{:test (fn [] (throw (ex-info "err" {})))} my-fn
  (fn []))

(defn call-my-fn []
  ((:test (meta #'my-fn))))

(call-my-fn)

Click on the first displayed stack frame - opens clojure,core/fn.

Note that this is the pattern used by deftest.

Environment & Version information

All things master head. Mac.

CIDER version information

;; CIDER 0.26.0snapshot, nREPL 0.8.0-alpha5
;; Clojure 1.10.1, Java 1.8.0_252

Emacs version

GNU Emacs 26.3 (build 1, x86_64-apple-darwin18.2.0, NS appkit-1671.20 Version 10.14.3 (Build 18D109)) of 2019-09-02

For example I have the code:

(println "foo")

With cider-pprint-eval-last-sexp-to-comment I can eval the code and get

(println "foo") ;; nil

in clojure buffer and also it print foo in cider-repl buffer.

I would like to get something like this in the current buffer:

(println "foo") ;; nil foo

How can I do this?