Open jpluimers opened 7 years ago
They don't have any of the markup needed to be interpreted. You can, in theory, teach a parser to identify a mediawiki site and deal with it's way of doing things which might be worth it.
I remember the old bookmarklet parsing many of these correctly from a UX perspective.
If the v8 bookmarklet didn't work, then the v4 bookmarklet usually worked fine, but I cannot test with those any more as that functionality got broken.
Related to https://github.com/WordPress/press-this/issues/11
Parsing either these URLs gives an empty post:
I've seen this happen with other Wikipedia links as well, but not enough of them to assume all of them fail.
Will update this issue if I recognise a pattern in the failing links.