Fork me on GitHub

@andrzej046 There were some breaking changes to the peg.el library that clj-refactor is using internally. I recall that a workaround was to use an older version. I was also under the impression work started to replace edn.el (which uses peg.el) with @plexus’s parseedn.el. I’ll have to check how far we got there.


I’m assuming the problem you encountered is

Andrzej P19:02:21

Yes! That was exactly it!

Andrzej P19:02:04

Thanks 🙂 . I’ve pinned peg to the melpa-stable archive for now and installed peg version 0.6. It works as expected now.


I also see we have yet to switch to the new library.


As the changes are trivial I’ll try to handle them myself later this week. There’s no time like the present. 🙂


When using cider w/ both cljs + clj, cider-apropos inside a cljs buffer seems to be using the clojure apropos instead of cljs.repl/apropos - is there a different function / nrepl op that I should be using to get cljs results? If not, where would be the best place to extend this functionality (and perhaps submit a PR)


Note that it's using the cljs host process clojure apropos, not the sibling clojure apropos (ie. it seems to be dispatching to the correct nrepl, just using clojure's namespace, instead of cljs')


I don't think we have an implementation in orchard for clojurescript var searching.


Looking through that code now... Looks like perhaps a function could be added to cljs/analysis.cljc that dispatches to the cljs.repl and returns maps in the same shape as the stock apropos? Once that's added, nRepl would also have to be extended to support a new op that calls that code too right? and then lastly cider w/ a function that calls the dispatch op?


Ideally we’ll just have in Orchard a version of apropos that works for the ClojureScript and cider-nrepl will pick the right function in Orchard based on the type of the current nREPL session (Clojure or ClojureScript). You can see how macroexpansion, info or completion works in cider-nrepl to get an idea of what I mean.