Native targets for Functional and Logging #26
Conversation
I'd love to have this setup, just haven't gotten around to it, since it requires using a macOS machine on CI. Created #27 to track. |
|
Honestly I started with Stately out of fear, but AtomicFU turned out to be much easier to work with. |
Codecov Report
@@ Coverage Diff @@
## main #26 +/- ##
============================================
+ Coverage 92.85% 92.98% +0.12%
- Complexity 57 58 +1
============================================
Files 12 12
Lines 112 114 +2
Branches 12 12
============================================
+ Hits 104 106 +2
Misses 8 8
Continue to review full report at Codecov.
|
| public synthetic class com/juul/tuulbox/logging/LogAtomicTagGeneratorRefVolatile { | ||
| public fun <init> (Ljava/lang/Object;)V | ||
| } |
There was a problem hiding this comment.
Can't quite pin down where this is coming from (and if it needs to be on the public API)?
There was a problem hiding this comment.
It's auto-generated from atomicfu's plugin. Not totally sure why
There was a problem hiding this comment.
I think the synthetic marker prevents library consumers from accessing it directly?
There was a problem hiding this comment.
It's auto-generated from atomicfu's plugin. Not totally sure why
Oh gotchya. If it's generated from atomicfu, not much we can do (seems fine) — I was just a little confused where it was coming from.
I think the
syntheticmarker prevents library consumers from accessing it directly?
Not sure actually.
Functional
No changes, everything just worked.
Logging
No
StackTraceTagGeneratorfor now, not ruling it out for later. Hit two snags in my first stab:Throwablewas setting off the memory-leak crash. I think it was a false positive, but didn't want to spend too much time debugging, because...Releasebuilds are super useless without the mapping file, which wouldn't be available at runtime.No console logger tests for now. In theory it should be possible to
freopenthe streams forstdoutandstderr, but I had trouble getting that working.