Return Torrent hash or TorrentDictionary from Client.torrents.add() and Support Additional Actions when Adding a Torrent

[unsocial-bleach]

I want to enable super seeding at the time that I add a torrent via an additional parameter to qbt_client.torrents_add(urls='...'), like qbt_client.torrents_add(urls='...', enable_super_seeding=True).

I don't want to have to lookup the torrent hash, and make a separate request to enable super seeding. Using qbt_client.set_super_seeding(...) is less convenient.

I understand that this may mean straying from implementing the existing API calls exactly, but I think that the UX benefit of this Python library will make it worth it.

[rmartin16]

Hello @unsocial-bleach,
This is a probably a reasonable request based on your use-case.

However, as in #29, it cannot be guaranteed this client can reliably determine which torrent will be added to qBittorrent when qbt_client.torrents_add() is invoked. That is, from an arbitrary URL, file, or binary blob that torrents_add() can accept, it is non-trivial to determine which torrent was actually added to qBittorrent....if one or more was added at all.

Although, given enough information in a specific example, it would normally be trivial to determine the actual torrent; it's just difficult to generalize to any possible case. I'm definitely open to ideas I may be overlooking.

I think ideally the qBittorrent team would implement qbittorrent/qBittorrent#4879 so that their torrents/add method would return more information about what torrents were added during the call; this would trivialize performing more complex operations on the added torrent.

Alternatively, thinking out loud, it'd probably be technically possible for this client to request a corresponding torrent hash from users for each torrent to be added by qbt_client.torrents_add().....however, that starts to feel particularly kludgy and likely even error-prone.

As you've stated, though, it should be fairly straightforward at least to write a small wrapper for this client to add the torrent and set additional options such as super seeding....albeit with at least two more API calls...

[unsocial-bleach]

All reasonable points! I guess for now, we'll wait on qbittorrent/qBittorrent#4879.

[commonelk]

One can create a temporary category, assign it to the torrents during the initial qbt_client.torrents_add(...) call, use said category to retrieve the hashes, and then remove the temporary category, like so:

temp_cat = "..."
qbt_client.torrents_add(urls=magnet_urls, category=temp_cat)
hashes = [ torrent_info.hash for torrent_info in qbt_client.torrents_info(category=temp_cat) ]
qbt_client.torrent_categories.remove_categories(categories=temp_cat)

I'd further suggest randomly generating the category and or ensuring it is unique (e.g. compare with qbt_client.torrents_categories()) in the off-chance that the temp_cat is already an existing category. Of course, the need for this precaution depends on one's use-case.

There's likely a better method of retrieving hashes from added torrents, but I figured I'd leave this here in case it helps anyone.