feat(*): fetch from wasmcloud cache on inspect commands

[matthewtgilbride]

Closes #138

[brooksmtownsend]

Great stuff @matthewtgilbride, really only had one nit comment here.

[matthewtgilbride]

Great stuff @matthewtgilbride, really only had one nit comment here.

Thanks @brooksmtownsend - renamed things to "local_artifact" and "cached_artifact - along moving the local/cache/remote helper to reg.rs.

I'm wondering what the best way to test this is? I was thinking of jamming the echo.wasm file into the cache directory similarly to how it's added to the local directory here. Then I was going to add a call to get the remote actor before the push command to see if it would get pulled from the cache. Feels a bit awkward...thoughts?

[brooksmtownsend]

Thanks @brooksmtownsend - renamed things to "local_artifact" and "cached_artifact - along moving the local/cache/remote helper to reg.rs.

I'm wondering what the best way to test this is? I was thinking of jamming the echo.wasm file into the cache directory similarly to how it's added to the local directory here. Then I was going to add a call to get the remote actor before the push command to see if it would get pulled from the cache. Feels a bit awkward...thoughts?

I think that's a good idea, a good baseline would be a unit test just to verify:

should produce a file like wasmcloud_azurecr_io_kvcounter_v1.bin

One great way you could test this is exactly what you suggested, throw the echo.wasm in the temp directory under a name like fakeoci_io_echo_v1.bin and then try to inspect faceoci.io/echo:v1 and even though that doesn't exist, the cached file should be retrieved instead

[brooksmtownsend]

This LGTM @matthewtgilbride, just need the DCO fix 😄 If you check the workflow details it'll tell you the command you can run

[matthewtgilbride]

👍 think I'm ready. Added the --no-cache option so take a look before approving.

[brooksmtownsend]

I thought of one more request... could you also add a test for the provider so that we test that works as well? I figure having a wash claims inspect + wash par inspect would be great

[matthewtgilbride]

I thought of one more request... could you also add a test for the provider so that we test that works as well? I figure having a wash claims inspect + wash par inspect would be great

🦃 done

Interestingly the wash par inspect integration tests are really slow. Way slower than it should take to download/push the httpclient provider a couple of times. 🤔

[brooksmtownsend]

Interestingly the wash par inspect integration tests are really slow. Way slower than it should take to download/push the httpclient provider a couple of times. 🤔

That's a fun one actually that we worried about as we started, turns out the whole compress/decompress part of the gzip library we use is SUPER slow in debug mode, so hopefully if we try with release mode that doesn't reproduce

[brooksmtownsend]

This LGTM, and should be ready to merge as soon as the tests run and pass 🚀