Background information
Wikis have content in different language variants, that will get rendered depending on the language sent by the user agent.
Right now, the server side PDF renderer doesn't do anything particularly smart with the language, resulting in PDFs that are always rendered in the default variant for the wiki.
This makes it impossible to get PDFs in different variants in wikis like the chinese and serbian.
What
It should be possible to get rendered PDFs with a different language variant.
How
Proton should take into account the language header (accept-language) in the requests, and forward it to the renderer process in the best way possible.
The simplest solution could be just setting the language in chrome at instantiation, passing args: ['--lang=bn-BD,bn'] when doing a .launch, and that should automatically add the appropriate headers to all requests. Chrome args, Puppeteer docs.
If this approach had any problems, we can probably do with [page.setExtraHTTPHeaders(headers)](https://pptr.dev/#?product=Puppeteer&version=v1.12.2&show=api-pagesetextrahttpheadersheaders) instead, setting Accept-Language there manually before making the page go to the wiki URL.
Acceptance criteria
- Asking for a PDF sending a language variant header in the request, should result in the PDF content rendered in the requested language variant