Ohhhhh I'm adding that as a filter for newsboat tonight.
Instead of loading castrated headline + tagline stubs, then I'll be able to read the whole article in terminal without any of the commercials and "also see" BS. <3
I was just setting the BROWSER environment variable to a script that makes me choose if I want to open the article on rdrview (with lynx) or firefox.
Newsboat filters operate on the whole feed. So the script needs to read the input rss (cf feedparse) then retrieve the HTML (cf rdrview) and then generate rss errrr.... Feedparse doesn't do that last bit. And then add a layer of on disk caching because otherwise it would keep re-downloading all entries over and over again.
I'll get back to that tomorrow I guess. Now for shut-eye