I'm on nginx and I just found some sites using embedded content from my site without my consent.
I have sites using my content that way and I would like to permit them but I need to stop some sites of doing it.
Is this possible?
This would be tricky to implement, but you could have a list of IP/mask pairs that when a they got an page request was served from a different directory.
Within that directory have have a process that randomizes blocks of 3 to 10 words. This makes all the text read like bad auto-translation of chinese with an intermediate stop in sanskrit. Process the images to either lower the contrast, or half the intensity of the green channel, or invert one of the channels.
If nginx can't do this, then pf firewall should be able to. Run another instance of nginx on a different port, and the bad-asses get sent to the scrambled pages.
To automate finding these guys, for each page pick a string that is long enough to unique to that page, and search google for that, -your site name. I think you could script this with lynx.
nginx valid-referers may be able to help, assuming they're not linking direct to the URL of the content.
Here is a good write-up with examples.
If you'd prefer to take more of a blacklist approach, you could also try the http_map_module.