Html extraction about trafilatura

[adbar]

Hi @slavaGanzin, thanks for your feedback!

• You could definitely write a PR to raise an exception for unknown output formats.
• HTML as output won't be possible as is because the content is converted to XML internally. However, you could use XML as output format and work on the nodes as you'd do with HTML.
• The output has to be a string sooner or later because there are people using the package on the command-line. I didn't think of actually using something else than a string for consistency reasons, maybe that's something we could tackle later.

[slavaGanzin]

What was I expecting is to get this result using extract(downloaded, output_format='html'). But it just return text, and as I now know understand, it should raise an exception 😄

BTW, it's very confusing that extract(downloaded, output_format='json')
return a string. Yes I know that I can use bare_extraction, but still unexpected.

Overall thanks for your work, I could make a PR. If this will be approved.

[slavaGanzin]

Thanks Adrien