Separate messages for Copy and Push options regarding lock output flag

[vicmarbev]

This PR is a response to issue #354. Different functions to create the Cobra flag for push and copy commands, each one with its own text.

[vicmarbev]

It seems that the linter does not like the exported methods (even though the Set function was with capital 'S'). Should I do something about this?

Also, the performance test is .8 seconds slower, but I haven't changed anything about that.

[joaopapereira]

When you create new public functions you need to add a comment.
This linter was added after a big chunk of the code was already in place, so when changing older functions we need to add the comments.
You can see an example in https://github.com/vmware-tanzu/carvel-imgpkg/blob/develop/pkg/imgpkg/registry/registry.go#L102

[vicmarbev]

It seems there was a problem reading the images to copy:
Error: perf_test.go:63: Failed to successfully execute 'imgpkg copy -b 127.0.0.1:35721/repo/perf-image --to-repo 127.0.0.1:35721/repo/perf-imagecopy --yes': Execution error: stdout: '' stderr: 'imgpkg: Error: Reading Images from Bundle: Checking if 'index.docker.io/cloudfoundry/minio-cf-for-k8s@sha256:5cb64001e21d6e3adb05717ca49aca34d52f576cf8365af28860d4afad383ea4' is a bundle: GET https://production.cloudflare.docker.com/registry-v2/docker/registry/v2/blobs/sha256/2a/2aa1d6d1e1dccae62ad2c2c3c35a303b9edfb793e57d39004540c54eb7781478/data?verify=REDACTED: unexpected status code 409 Conflict: error code: 1023 ' error: 'exit status 1'

Also, other test shows that we are too slow by 0.2 seconds:
=== RUN TestCopyingLargeImageWithinSameRegistryShouldBeFast Running 'imgpkg copy -i 127.0.0.1:53077/repo/perf-image --to-repo 127.0.0.1:53077/repo/perf-image1 --yes'... initial push took: 21.7880665s imgpkg copy took: 1.6420038s assertion_compare.go:313: Error Trace: perf_test.go:49 Error: "1452537766" is not greater than "16[42](https://github.com/vmware-tanzu/carvel-imgpkg/runs/6161215029?check_suite_focus=true#step:7:42)003800" Test: TestCopyingLargeImageWithinSameRegistryShouldBeFast Messages: [copying a large image took too long. Expected it to take maximum [%v] but it took [%v] 1.[45](https://github.com/vmware-tanzu/carvel-imgpkg/runs/6161215029?check_suite_focus=true#step:7:45)2537766s 1.6420038s]

[vicmarbev]

After a rerun, the only failing test is the performance one. As we haven't changed behavior, should we modify the failing threshold? or does it merit investigation?

[joaopapereira]

LGTM

[joaopapereira]

@vicmarbev I created this issue #380 to track the investigation and eventually take some measures to save some of this information in order to understand in the future if we see degradation with time or not.

Nevertheless, I will merge your PR.