feed_seeker scans the same links repeatedly

[pushshift]

The current implementation appears to have no memory of previously searched urls. This can cause large circular loops that never end and increases scan time and bandwidth.