Use list of approved domains instead of ignored ones

[frvge]

Instead of ignoring domains, which can be tricky if you have ad networks, I'd like to propose a new feature that only checks a resource if its domain matches a regex.

This way, I can allow cdn.site.com, www.site.de and a few other variants (preferably in 1 regex) and ignore all the other domains.

See #766 (comment) for more details about implementation.

[molant]

@frvge good idea. Any proposal for the name of the property?
Also, do you want to try to implement it? We'll be more than happy to help you in the process.

[alrra]

The blacklist approach makes users actually acknowledge the issues, and only after, decide if they want to ignore certain domains or try to do something about them.

With the whitelist approach the biggest downside is that over time the site may include new things, users will most probably forget to update the list, and while sonarwhal may suggest that everything is fine, the site as a whole may be getting worst.

---

Note: I'm fine with having this feature, but we should make the advantages and disadvantages of the approaches perfectly clear.

[frvge]

@molant , I don't have experience with TypeScript, or anything after ES3, but I can probably get some assistance from my colleagues. I'll give it a try.

@alrra , correct. There is a risk of missing out on issues. This should be worded clearly in the documentation.

[alrra]

I don't have experience with TypeScript, or anything after ES3, but I can probably get some assistance from my colleagues. I'll give it a try.

If there is something we can help with, let us know!

[frvge]

I started working on it today, but it's too early to really say if I'll be able to get it done.

[molant]

Awesome!

[alrra]

@frvge Let us know if we can help with anything!

[frvge]

Do you want to have an additional parameter, just like the ignoredUrls one? Or maybe a setting within the ignoredUrls like "reverse"? Or maybe it's better to abstract it away in a new structure / interface?

Currently I went with the additional parameter, but I'm doubting if it's the cleanest way.

The idea is that if the whitelist has at least 1 element and a resource is not in the whitelist, it gets ignored. Afterwards, the ignore list gets applied.

I'm not sure if I want to have a "rules" part on the whitelist because it doesn't make sense to whitelist a domain/resource for only a few rules. That functionality is better done by ignoreUrls.