Netty's "4.1.60.Final" format not supported

[Olical]

Hi there! Thank you for providing such a great tool, I rely on it over in https://github.com/Olical/depot and have only recently updated to the latest version (from 0.x.something!) which I think fixed some of my issues but may have introduced others.

Specifically Olical/depot#52 and Olical/depot#18 (may be something else).

It seems to stem from this 4.1.60.Final format, which doesn't look standard or okay to me. So I was wondering if this format used in the various netty projects (specifically io.netty/netty-transport-native-kqueue) is supposed to be supported? Or if it's just plain broken and you don't think it should work in the first place?

I don't think I have enough domain knowledge on these version string formats to make a judgement call, so I thought I'd see if you had any thoughts before digging deeper?

[xsc]

@Olical Thanks for reporting! Generally, those versions seem to be handled perfectly fine:

(require '[version-clj.core :as v])
(v/newer? "4.1.63.Final" "4.1.60.Final") ;; => true

My hypothesis is that "4.1.63.Final" is not even considered as a valid version-to-upgrade-to, i.e. filtered out of the list of available versions at some point.

Looking at depot, I see that version-type returns :qualified here. Since the default for consider-types is #{:release} I could see it removing these versions from the list.

Interestingly, this must mean that previously the .Final suffix was not considered a qualifier (and in my opinion, it should be). I think, to solve your problem, you'll need to explicitly treat "Final" as a release, and version-clj.core/parse might help:

(v/parse "4.1.60.Final")
;; => {:version [(4 1 60) ["final"]], :qualifiers #{"final"}, :snapshot? false, :qualified? true}

Update: It could make sense to have a release? predicate directly in version-clj, and additionally have parse return a :release? key. If you want, you could open a PR in version-clj to add those two things; otherwise, I'll probably do it at some point during the upcoming week. :)

[Olical]

Oh interesting! Thank you so so much for the insight, I'll take this info and see what I can do with it 😄 just wanted your opinion before I went down a rabbit hole that could've ended with "they just have a completely invalid version syntax".

[Olical]

If you don't think this is an issue with version-clj in any way (which I don't think it is, mostly just my misunderstanding) then please feel free to close the issue! The advice should be enough for me to fix the issues in my project.

[xsc]

@Olical Yeah, I think this is not an issue of version-clj, but with deciding if a given version is a release, based on what version-clj provides. So, I'll close it for now. :)