DXCDT-403: Separate rendering for logs tail and logs list

[willvedd]

🔧 Changes

Separating the LogList rendering function into LogTail and LogList. The original function served two commands – auth0 logs tail and auth0 logs list. Despite the similarities between these commands, their uses differed enough to warrant the separation.

Specific differences include:

• Streaming output vs static rendering
• Accepting channel as requirement vs not
• Handling of zero logs returned – log streaming should not terminate early

🔬 Testing

Primarily manual. Log tail streaming output cannot be tested with current framework though auth0 logs list will be covered.

📝 Checklist

• All new/changed/fixed functionality is covered by tests (or N/A)
• I have added documentation for all new/changed functionality (or N/A)

[sergiught]

Now that we're separating them, if we run auth0 logs list with a filter and we get back no log messages of that type we'll post a message that "No logs are available.", although that's slightly incorrect.

There probably are log messages just not any that satistfy the filter. So maybe we can improve the output with something like:

auth0 logs list -n 10 --filter "type:f"        

=== tenant.example.auth0.com logs

No logs available matching filter criteria.

 ▸    To generate logs, run a test command like 'auth0 test login' or 'auth0 test token'

[willvedd]

The behavior you're describing is not caused by this change, it existed prior. Though I do agree with the general sentiment.

Implementing isn't exactly trivial because the render functions are abstracted away from knowing the filters and other arguments. The best we could do here is to provide a generalized message that would apply to alls situations regardless of provided criteria.

[sergiught]

The behavior you're describing is not caused by this change, it existed prior. Though I do agree with the general sentiment.

Well we can still leave the codebase a bit better than it was, and also improve DX considering we're now touching this part, but I'm also okay if we wanna do this in a separate PR.

To implement this we can check inside listLogsCmd() if we have any filters selected and pass it to the LogList() func.

hasFilter := inputs.Filter != ""
cli.renderer.LogList(list, !cli.debug, hasFilter)

then inside the LogList() we can tweak the if len(logs) == 0 check as follows:

if len(logs) == 0 {
		if hasFilter {
			r.Output("No logs available matching filter criteria.\n\n")
		} else {
			r.EmptyState(resource)
		}
		r.Infof("To generate logs, run a test command like 'auth0 test login' or 'auth0 test token'")
		return
	}

This will have the desired behavior as described above.

go run cmd/auth0/main.go logs ls -f "type:f" 

=== example.auth0.com logs

No logs available matching filter criteria.

 ▸    To generate logs, run a test command like 'auth0 test login' or 'auth0 test token'

[willvedd]

Let's address in a follow-up PR that's more focused.

[sergiught]

The LogTail doesn't seem to attach at the top the table headers like the auth0 logs list does:

  TYPE                     DESCRIPTION                                             DATE                  CONNECTION            CLIENT

[codecov-commenter]

Codecov Report

Patch coverage: 16.66% and project coverage change: -0.05 ⚠️

Comparison is base (054145b) 54.50% compared to head (040c74b) 54.46%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #672      +/-   ##
==========================================
- Coverage   54.50%   54.46%   -0.05%     
==========================================
  Files          89       89              
  Lines       11509    11512       +3     
==========================================
- Hits         6273     6270       -3     
- Misses       4789     4796       +7     
+ Partials      447      446       -1     

Impacted Files	Coverage Δ
internal/display/logs.go	52.58% <12.50%> (-3.67%)	⬇️
internal/cli/logs.go	52.41% <50.00%> (-0.33%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[ewanharris]

Tested locally and looks good, approving as there are no new changes in behaviour and existing items have been captured for future improvements

[willvedd]

@sergiught I've made note of the valid points that you raised. Given the current emphasis on testing, I'm going to opt to address these in a separate, focused PR.