Verify on CI that no reflection or boxing warnings are emitted

[vemv]

This makes Expound a friendlier library for consumers that observe these warnings.

[vemv]

@bhb : hope this PR sounds reasonable. No idea if the build passes btw, although I did run the lein check locally

Cheers - V

[bhb]

@vemv Thanks for this! I tried your branch and I get the following warning when I run lein with-profile -dev,+check check 2>&1

Boxed math warning, /private/var/folders/gm/p9fh4d3n5m11nnyjrsry1cm80000gq/T/form-init6307258272756268834.clj:1:1000 - call: public static boolean clojure.lang.Numbers.isZero(java.lang.Object).
Compiling namespace expound.util
Compiling namespace expound.specs
Compiling namespace expound.paths
Compiling namespace expound.alpha
Compiling namespace expound.problems
Compiling namespace expound.printer
Compiling namespace expound.ansi

Do you get the same thing? I don't think it's only my machine because I get a similar failure in CI.

I tried this with my branch that make a few minor changes to your code https://github.com/bhb/expound/pull/181

In both cases, I get a similar warning. Do you know how to fix this? I'm afraid I'm not an expert on fixing these boxing warnings.

[vemv]

Hey there,

thanks to you as well!

Do you get the same thing?

Yes I do, even when voiding my profiles.clj. I suspect the dyn var affects Lein itself, so that's the thing that complains.

Maybe you can grep -v "/tmp/\|/private/var/" the output?

[bhb]

Yes I do, even when voiding my profiles.clj. I suspect the dyn var affects Lein itself, so that's the thing that complains.

Ah, yes, that would match what I see here. I think you're correct, I'll just need to filter out that warning.

[bhb]

@vemv Thanks again for this PR! It's merge into my next branch as of d2087150e527c8b601e7a4a7edd90862cba6990e

[vemv]

🙌!