Implementated global validation of domain strings

[CyberAustin]

Should close #626 and provide a basic sanity check for all providers, agnostic of provider specific requirements.

[qdm12]

Thanks for the PR 😉
A few changes to be made though 🤔

[CyberAustin]

Ugh. Can't run linters or devcontainer in Replit. Have to fix this later at home.

JK. Forgot GitHub codespaces was a thing.

[CyberAustin]

I don't know why the linter is failing in GitHub Actions because it passes in the dev container. Anyway, ready for review.

[qdm12]

Nice 👍 💯

Now, we need to add a check the domain isn't "" for the providers actually using the domain parameter. Most of them do, but there are 2-3 exceptions (like duckdns, goip). So that would be edits in each of the provider subpackage within internal/provider/providers

[CyberAustin]

Now, we need to add a check the domain isn't "" for the providers actually using the domain parameter. Most of them do, but there are 2-3 exceptions (like duckdns, goip). So that would be edits in each of the provider subpackage within internal/provider/providers

I'm not sure what you're asking here. Could you explain it a different way.

[qdm12]

We need to add a check in most providers that the domain is not empty "". Because in the current "above layer" check, we check the domain is empty (🆗) or valid (🆗), but some providers require a non-empty domain (most except a few).

For example in

ddns-updater/internal/provider/providers/cloudflare/provider.go

Line 97 in 6a6b1a8

switch {

add a case to check the domain is not empty:

case p.domain == "":
    return fmt.Errorf("%w", errors.ErrDomainNotSet)

[CyberAustin]

Should be good now. Please take a look @qdm12

[qdm12]

Thanks!
Positive is it does it in 3 lines for all providers except the two weird ones
Negative side is it it adds coupling in the code, but oh well that's fine this time 😉

Just rebase/fix the go.mod dependency diff and I'll merge, thanks! 🎖️