This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2017-11-08
Channels
- # bangalore-clj (4)
- # beginners (88)
- # boot (12)
- # cljs-dev (10)
- # cljsjs (1)
- # clojure (284)
- # clojure-denmark (2)
- # clojure-dev (35)
- # clojure-italy (8)
- # clojure-russia (36)
- # clojure-spec (38)
- # clojure-uk (51)
- # clojurescript (145)
- # cursive (6)
- # data-science (1)
- # datomic (8)
- # duct (43)
- # emacs (9)
- # figwheel (2)
- # fulcro (29)
- # graphql (1)
- # immutant (3)
- # instaparse (1)
- # jobs (1)
- # jobs-discuss (1)
- # lumo (16)
- # off-topic (50)
- # onyx (90)
- # re-frame (6)
- # reagent (20)
- # remote-jobs (3)
- # ring-swagger (18)
- # schema (8)
- # shadow-cljs (141)
- # slack-help (3)
- # spacemacs (36)
- # unrepl (7)
- # vim (1)
- # yada (2)
Here is a proposal for the completion action:
It takes 3 arguments:
• before
a string of what is before the cursor (at least the prefix to complete but may encompass more eg up to the start of the form)
• after
a string too, may be empty or missing
• namespace
a symbol
It returns a list of maps with key :candidate
(string) (more optional keys to be defined later)
@cgrand not that unrepl is intended for this, but: would before
handle "(ns foo) (de"
as a value.
As an aside, I believe the completion action would translate fairly well to compliment, which is considered the best available completion.
@dominicm sure because the MVI that @pesterhazy is working on is going to first run a regex to ignore everything but the last word 🙂
heh. Does it matter that different implementations may or may not handle that? I guess I'm putting a question over the potential ambiguity of "more".