Track AI Bots through our HTTP Tracking API

[sgiehl]

Description:

Review

• Functional review done
• Potential edge cases thought about (behavior of the code with strange input, with strange internal state or possible interactions with other Matomo subsystems)
• Usability review done (is anything maybe unclear or think about anything that would cause people to reach out to support)
• Security review done
• Wording review done
• Code review done
• Tests were added if useful/possible
• Reviewed for breaking changes
• Developer changelog updated if needed
• Documentation added if needed
• Existing documentation updated if needed

[sgiehl]

Note: We need to update the tracking documentation afterwards. I'll create a PR for that once this PR has an approval.

[sgiehl]

Note: This will only enable the plugin for new installs by default. Once the plugin is feature complete we can consider to add an update script to also enabled it for existing installs automatically.

[mneudert]

Suggested change

	* This is the first method called when processing a tracker request.
	*
	* Derived classes can use this method to manipulate a tracker request before the request
	* is handled. Plugins could change the URL, add custom variables, etc.
	* This method is called last.
	*
	* Derived classes should use this method to insert log data.

Or something like that to document this is for writing, and not for manipulation/filtering.

[mneudert]

Do we also (right now that is) need a way to delete data for a site? Like a new PrivacyManager.deleteDataForDeletedSites event we can hook onto because a LogTable won't work?

Or can we delay that a for a little while?

[sgiehl]

Actually I think log data removal for deleted sites might work by implementing a LogTable. Didn't hink of this case. For removal by date that didn't work as it first looks up all visit and then iterates over the logtables that have a visitid column, which isn't the case here. I'll see if I can add a LogTable class and an according test.

[sgiehl]

Ok. It doesn't work aswell, as it still tries to check how the table can be join with log_visit. So a LogTable remains useless for now. Will add a new event for it

[mneudert]

We should replace that with a DI injected LoggerInterface.