Fork me on GitHub
#calva
<
2020-02-18
>
pez08:02:05

Dear Calva-friends: Can I get some help testing this PR? https://github.com/BetterThanTomorrow/calva/pull/562 It is about a more Paredit-aware, and much more efficient, highlighting of matching brackets and rainbows. The efficiency part is visible when you test it, pay attention to the forms at the very edges of the window as you scroll it slowly. This makes a big difference mostly for large files (like Clojure core.clj, but also for slightly more common cases). The Paredit aware bit is the most important. If something is indicated to be a bracket pair by the highlighting, then Paredit should navigate and edit it as such, awlays. It is a rather big change, even if not touching all that much of the code, so testing it a bit before release is extra important. Here's the VSIX, for your convenience. (See the PR for known issues.) https://3916-125431277-gh.circle-artifacts.com/0/tmp/artifacts/calva-2.0.77-wip-token-cursor-rainbow-d1b1f2e9.vsix

✔️ 4
pez07:02:34

I need help testing some more. Here's something like a rewrite of that highlighting PR. https://3956-125431277-gh.circle-artifacts.com/0/tmp/artifacts/calva-2.0.77-wip-token-cursor-rainbow-b9b020ba.vsix It is about 5X faster, and more accurate. Also ignores (`#_`) work again, more accurate than the current Calva. In theory it should in all just be better than the current Calva release, so please run with it and help catch any unlikely quirks.

👍 4
pez09:02:47

Alright so there was a problem with regexes there. Fix is building. Grab it from the PR or wait for me to post a link. The fix also addresses an old issue with bracket literals like \(.

bringe15:02:01

I installed it and will use it while working today

pez15:02:19

Awesome. I will probably back away from the performance part of the fix, b/c really tricky corner cases.

pez15:02:34

The latest build fixes some of the issues with #_ highlighting, but not all. So run with that, preferably.

hoppy17:02:21

so far so good on my end

❤️ 4
benny22:02:15

asked this in the general channel, but was rightfully redirected here. any suggestions?

benny22:02:34

seems like this issue is related but i’m on a 2020.* version: https://github.com/borkdude/clj-kondo/issues/342

borkdude23:02:31

I'm off to bed in a minute, but if you can paste a repro at the clj-kondo channel or repo I'm willing to take a look tomorrow.

borkdude23:02:20

clj-kondo now has built-in support for compojure, but when you use refer all it might not be able to resolve the symbol. just not use that 🙂

benny05:02:54

well that sure is easy, i didn’t realize the fix is just to be more accurate with the refs , thanks @U04V15CAJ!