Sanitise incoming bridge_reachability reports for bogus bridge data

[isislovecruft]

It seems that someone¹ ran ooni-probe's bridge_reachability test with an input list of a bunch of pastebin and source code/git hosting sites. This caused the charts on https://beta.chokepointproject.net/measurements/tor-bridge-reachability#Bridge_explorer to have a bunch of URLs in the "bridge select" dropdown.

While there's nothing wrong with a Tor bridge having an URL, and while fte sometimes did have URLs in some actual deployments… perhaps we could try to avoid making it so easy to spam the bridge_reachability test results?

---

¹ The Chokepoint Project devs said it was someone operating a probe from Canada, and that the report this came from was bridge_reachability-2015-04-03T183901Z-AS812-probe.yamloo, but that report now strangely doesn't exist in http://pipeline.infra.ooni.nu/CA/ (!!).

[hellais]

I believe the report is this one: http://pipeline.infra.ooni.nu/CA/bridge_reachability-2015-04-03T183901Z-AS812-probe.yamloo.gz