Add try/catch when reading EDN data

[andersenleo]

Fixes #255

Since the artifacts/get-clojars-artifacts! function is written in a way that isn't easily testable I opted out on tests. I would need to refactor the code a bit more to make it possible to write a test for this. Let me know if you'd like me to do that.

The fix is simply to use keep instead of map - and swallow any exception on edn/read-string.

[benedekfazekas]

A test would be nice I think. Please also add a one liner to the changelog. Thanks again, really appreciated.

[andersenleo]

Moved the "reading" to a separate function and wrote a minimal test for verifying that. Let me know if you'd like to restructure things differently.

[benedekfazekas]

good work @andersenleo! will release a new snapshot in the evening.

[andersenleo]

@expez - the edn/read-string function is called when the lazy-seq returned from get-clojars-artifacts! is evaluated further down the line. The try/catch that was in place would only return [] if the connection to Clojars was down (an error triggered before call to line-seq). And either way -- returning an empty seq would still make the feature unusable (if the artifact seq would have been eagerly evaluated).

[expez]

Thanks @andersenleo I eventually got that and deleted my comment hoping nobody would've seen it! :p

Great job on this! 👍

[benedekfazekas]

new snapshot version is on clojars

[andersenleo]

Lovely - it seems to work. I'm no Emacs expert, but adding

  (setq cider-jack-in-lein-plugins '(("refactor-nrepl" "2.4.1-SNAPSHOT" :predicate cljr--inject-middleware-p)
                                     ("cider/cider-nrepl" "0.22.0-beta1")))

works fine. Should be an easier way to override the refactor-nrepl version I guess? Suggestions are welcome :). Any idea when a "proper" release could be made available?

[benedekfazekas]

looks legit to me. no release is planned atm. nothing blocking one either afaik