Parse article body

[wilsonpage]

We need to show some content in magnet-client when the user taps on a POI, right now the meta-description is pretty crappy content. Ideally we would show an extract/summary of the real article, or even have the ability to show the entire article.

https://github.com/mozilla/readability might come in handy.

[wilsonpage]

@samgiles @arcturus thoughts?

[wilsonpage]

[samgiles]

I think we should use try and use fathom if we try this.

[wilsonpage]

That will depend on us moving everything to JSDoc instead of Cheerio, unless we parse twice :/ Happy to make that change if it gives us better metadata. Number one priority is to not lose any control if we move to a framework.