feat/range requests

[juliushaertl]

This PR implements rudimentary support for Range headers in the WOPI endpoint for downloading the file allowing the Collabora server to download the file in multiple chunks in parallel. During testing I've noticed that the webserver infront of Nextcloud might already cover general support but the full file would still be read, so this should bring a performance improvement by only reading relevant parts.

cc @mmeeks Since you were bringing this up for loading optimizations

[pedropintosilva]

I think this was brought up by @mmeeks sometime ago (that we could try to split the document and downloaded its parts concurrently). @Ashod you are probably the right the person to look at this (to check if it brings a performance benefit when integrating with Collabora Online). What do you think?

Notes:

• It doesn’t parse the Range header fully standard compatible, only one range is supported for now, however should be enough to proof that this brings a performance benefit when integrating with Collabora Online
• There is already ongoing work to load the file earlier (at the request of the iframe, @Ashod was working on this, I think)

[Ashod]

Thanks @juliushaertl. Good stuff!