add integration with opensubtitles mirrors

Nobody has claimed this yet.

many subtitles from opensubtitles have been scraped
and have been made available for free via bittorrent

> many subtitles

about
5_719_123 + (9_521_948 - 9_180_519) + 10*100_000
= 7_060_552 of currently 10_772_237 subtitles
7_060_552 / 10_772_237 = 65%

see also my project [opensubtitles-scraper](https://github.com/milahu/opensubtitles-scraper)
which is currently not active, because [they stole my VIP accounts for "abuse"](https://github.com/milahu/opensubtitles-scraper/blob/main/docs/dont-copy-our-website.md)
which will have to be adapted to [the new opensubtitles.com API](https://opensubtitles.stoplight.io/docs/opensubtitles-api/f65bc8dd4aef7-subtitle-exports)

ideally, all opensubtitles clients should ...

1\. try to get subtitles from a local subtitles database
such a database costs around 160 GB of disk space

2\. try to download subtitles from a remote subtitles database
remote subtitles database = opensubtitles mirror
example implementation: [get-subs.py](https://github.com/milahu/opensubtitles-scraper/blob/main/get-subs.py)
demo server: [milahu.duckdns.org/bin/get-subtitles](https://milahu.duckdns.org/bin/get-subtitles)

3\. cache all opensubtitles API requests to disk
so later, the cached API responses can be released via bittorrent
so future users can get free access to subtitles
the new opensubtitles.com API does no longer return zip files
but now it returns the subtitle files directly
so for releasing, the subtitles have to be compressed
the simplest solution is to compress each subtitle individually into a zip file
and store these zip files in a sqlite database
using the same database layout as previous releases
(feel free to experiment and do your own benchmarks)
ideally such releases should be usable directly
without extraction, to reduce disk usage

4\. send random opensubtitles API requests
to use more of the daily quota of a VIP account
note: if a VIP account uses 100% of his daily quota every day
then [opensubtitles admins will close the VIP account (without a refund!)](https://github.com/milahu/opensubtitles-scraper/blob/main/docs/dont-copy-our-website.md)
claiming that such behavior is "abuse of their service"
so these random requests should
be splitted into multiple smaller batches, distributed over the day
and it should use a random fraction of the daily quota, between 80 and 100%

duplicate issues

- https://github.com/jellyfin/jellyfin-plugin-opensubtitles/issues/199
- https://github.com/josdion/subbuzz/issues/40
- https://github.com/exebetche/vlsub/issues/260
- https://github.com/emericg/OpenSubtitlesDownload/issues/124
- https://github.com/QNapi/qnapi/issues/204
- https://github.com/matcornic/subify/issues/23
- todo: more opensubtitles clients?

The issue names no Subify files, tests, or entry points. Start with the OpenSubtitles API documentation and the linked get-subs.py example, then narrow the proposal to a specific integration target. Done should be defined by an agreed scope and corresponding tests.

Written by the indexing model from the issue text.