ClojureScript support relies on the piggieback nREPL middleware being present in your REPL session. There’s one exception to this, though: shadow-cljs. It has its own nREPL middleware and doesn’t rely on piggieback at all.
If cider-inject-dependencies-at-jack-in
is enabled, which it is by
default, then piggieback will be automatically added and configured
for your project when doing cider-jack-in-cljs
.
If cider-inject-dependencies-at-jack-in
is disabled or you’re going
to connect to an already running nREPL server using
cider-connect-cljs
, use the configuration in the following section.
To setup piggieback, add the following dependencies to your project
(project.clj
in a Leiningen based project or build.boot
in a Boot
project or deps.edn
):
;; use whatever are the most recent versions here
[cider/piggieback "0.5.3"]
[org.clojure/clojure "1.11.1"]
as well as piggieback
nREPL middleware:
in project.clj
:
:repl-options {:nrepl-middleware [cider.piggieback/wrap-cljs-repl]}
or in build.boot
:
(task-options!
repl {:middleware '[cider.piggieback/wrap-cljs-repl]})
or in deps.edn
:
{:aliases { :cider-cljs { :main-opts
["-m" "nrepl.cmdline" "--middleware"
"[cider.nrepl/cider-middleware,cider.piggieback/wrap-cljs-repl]"]}}}
or in build.gradle
:
dependencies {
devImplementation 'nrepl:nrepl:0.9.0'
devImplementation 'cider:cider-nrepl:0.45.0'
devImplementation 'cider:cider-piggieback:0.5.3'
}
tasks.named('clojureRepl') {
middleware = ['cider.nrepl/cider-middleware', 'cider.piggieback/wrap-cljs-repl']
}
Open a ClojureScript file in your project and type M-x
cider-jack-in-cljs
RET. With the correct configuration
and after answering a few prompts this will start up the
nREPL server and create a ClojureScript REPL buffer.
Note
|
Prior to CIDER 0.18, cider-jack-in-cljs would create both a Clojure and
a ClojureScript REPL. In CIDER 0.18+ if you want to create both REPLs
you’ll have to use cider-jack-in-clj&cljs instead.
|
When you have a combination of Clojure and ClojureScript REPLs, CIDER
will automatically direct all the usual CIDER commands to the
appropriate REPL based on whether you’re currently visiting a .clj
or
.cljs
file.
cider-jack-in-cljs
will prompt you for the type of ClojureScript
REPL you want to start. Keep in mind that some of the REPLs will
require you to configure additional setup. For example, you’ll need to
have Node.js installed to be able to start a Node REPL.
Tip
|
The only ClojureScript REPL type that doesn’t require any additional setup is the browser REPL. To use the Node.js REPL you’ll need to install Node.js. |
CIDER will automatically try to check if the required dependencies (e.g. Clojure libraries and/or additional tools like Node.js) to start a certain ClojureScript REPL are present. If you happen to run into a faulty requirements check you can disable them like this:
(setq cider-check-cljs-repl-requirements nil)
Ordinarily, CIDER dispatches code from clj
files to Clojure REPLs
and cljs
files to ClojureScript REPLs. Butcljc
files have two
possible connection targets, both of which are valid. So, by default,
CIDER tries to evaluate cljc
files in all matching connection
buffers, both clj
and cljs
, if present.
Thus, if you’re evaluating the code (+ 2 2)
in a cljc
file and you
have both an active Clojure and ClojureScript REPL then the code is
going to be evaluated twice, once in each of the REPLs. In fact, you
can create multiple clj and cljs sibling connections (C-c C-x C-s C-s/j) within a CIDER session and evaluation will be directed
into all REPLs simultaneously. See Managing
Connections for more details.
If you would prefer to evaluate cljc
files in only one of the matching REPLs, you can customize the variable cider-clojurec-eval-destination
to either clj
or cljs
. For example, to evaluate cljc
files in the ClojureScript REPL only, add the following to your Emacs configuration:
(setq cider-clojurec-eval-destination 'cljs)
Note: This variable can also be set on a per-buffer basis using setq-local
, or project-wide in dir-locals.el
.