This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2020-10-11
Channels
- # announcements (1)
- # babashka (132)
- # beginners (52)
- # calva (46)
- # clj-kondo (8)
- # cljdoc (17)
- # clojure (13)
- # clojure-australia (1)
- # clojure-dev (3)
- # clojure-europe (4)
- # clojurescript (4)
- # cloverage (1)
- # conjure (22)
- # datomic (9)
- # emacs (2)
- # fulcro (16)
- # leiningen (5)
- # malli (26)
- # off-topic (16)
- # pathom (3)
- # portal (5)
- # reagent (10)
- # reitit (5)
- # rewrite-clj (1)
- # ring (1)
- # shadow-cljs (14)
- # spacemacs (6)
- # tools-deps (10)
- # vim (11)
- # vscode (1)
- # xtdb (10)
I have some time this afternoon to have a stab at fixing that a little bit so pom.xml is less required.
So if I understand correctly, the .pom file in the maven repo is a copy of pom.xml. But cljdoc is relying on the pom.xml in the jar, which some jars don’t include?
the POM is accessible via the repo url and most packaging tools include it inside the jars „META-INF“ directory. In some places we rely on the file being included inside the jar but I’m not 100% sure if that’s a cljdoc thing or a tools.deps thing
Tools deps supports adding a jar as a dependency in which case it has(?) to rely on the pom being inside the jar
It would be duplicate work, but it could just use the project instead of the jarpath/pompath.
The other option is to download it and then just assume the location in $M2, checking that out now too.
For example, this (contains? #{"provided" "system" "test" "compile" "runtime" nil} (:scope %))
will solve the problem where the pom is already used.
I've opened a PR to fix this :) https://github.com/cljdoc/cljdoc-analyzer/pull/30