I did a sweep through of our MIME parser code. I regenerated the test cassettes and discovered that several of the parsers are no longer working properly. Some of the broken parsers were easy to diagnose and fix. As for the other broken ones, they were all for niche sites and I decided to remove them.
Added better error handling for the VideoTagMIMEParser
Added better error handling for the RedditVideoMIMEParser
Removed the fallback Imgur Album Scraper. The scraper is no longer working because Imgur doesn't return the album image URLs in HTML any more (I'm guessing they're dynamically inserted through JS)
Removed Twitch TV, twitch no longer adds the twitter:image meta tag.
Removed Oddshot, the website got shut down
Removed Imgtc, the image used for testing was removed and I couldn't find another one. Googling the site leads me to believe that it's primarily used for hosting illegal media.
Removed Vidme, this one hasn't been working for a while now
Fixed LiveLeak, their id tag changed names
Updated a couple of the test URLs that had expired
I did a sweep through of our MIME parser code. I regenerated the test cassettes and discovered that several of the parsers are no longer working properly. Some of the broken parsers were easy to diagnose and fix. As for the other broken ones, they were all for niche sites and I decided to remove them.
twitter:imagemeta tag.