Detect tracking parameters in URLs

[Famlam]

Implements #1950

This PR detects tracking parameters in URLs, such as fbclid or utm_campaign, and strips them from the URL.
It uses the list of https://github.com/duckduckgo/privacy-configuration/blob/main/features/tracking-parameters.json, which is CC BY-NC-SA 4.0 licensed. Based upon my understanding, we can use this for the detection as long as we refer to its origin (via resource=*) as we're not commercial (but mind, I'm not a legally educated person).

As the easiest way to parse the URL was to use urlsplit, I also added a warning (class 30933) for broken URLs that give a ValueError in that function. These are however pretty rare, so I suspect it'll barely detect something. So if you prefer, I can also skip over those.

[frodrigo]

which is CC BY-NC-SA 4.0 licensed. Based upon my understanding, we can use this for the detection as long as we refer to its origin (via resource=*) as we're not commercial (but mind, I'm not a legally educated person).

I'm unsure about that. As OSM is not NC. Maybe is better to ask ?

cc @jocelynj

[jocelynj]

Are you sure about the license?

I see on https://github.com/duckduckgo/privacy-configuration/blob/main/LICENSE that it is using Apache license v2, which is an open-source license. CC BY-NC-SA 4.0 seems to be only about DuckDuckGo logos and marks.

[frodrigo]

Are you sure about the license?

I see on https://github.com/duckduckgo/privacy-configuration/blob/main/LICENSE that it is using Apache license v2, which is an open-source license. CC BY-NC-SA 4.0 seems to be only about DuckDuckGo logos and marks.

It is not the same as here https://github.com/duckduckgo/privacy-configuration/tree/main#licensing

[jocelynj]

So, we have two different information about license :(

It would be better to ask which license is applied, and if CC BY-NC-SA is confirmed, to ask for explicit authorization.

In the meantime, we could limit to utm_* variables, which can be found in a free license here: https://en.wikipedia.org/wiki/UTM_parameters

[Famlam]

It would be better to ask which license is applied, and if CC BY-NC-SA is confirmed, to ask for explicit authorization.

Drafted an email on Matrix. If that's fine, we could leave it for a few weeks on draft until we get a confirmation?
If we don't hear anything, I'll go for @jocelynj s suggestion to use the UTM parameters.

[Famlam]

New implementation using the tracking parameter list from https://github.com/mpchadwick/tracking-query-params-registry is ready :)

[frodrigo]

Look ok for me. Thank you.