ShokoServer Ongoing APIv3 Request Thread

[x] Return count of missing episodes (collecting), missing episodes (total). Count as in count of episodes and not count of groups.
[x] Return count of unreconized files, multiple files and duplicate files.
[x] Return breakdown of series by type (Series, OVA, Movie, Others)
Reference image
[x] Return epname and series name for recognized files in recent files. Not very important as we could query /api/ep/getbyfilename but that takes a lot of queries. It would be very faster if everything's returned in a single call.
[x] Return size and number of series in /api/v3/ImportFolder
[x] Fix /api/modules to return proper counts of watched_files, watched_series, hours_watched. Currently, it returns 0 for all three of then. EDIT: Fixed in v3 with a new endpoint
[x] Fix /api/file/recent to return datetime values for created and updated fields. Currently it only returns date and it should return datetime according to Swagger.
[x] ~~Make /api/file/recent return only the filename instead of the relative path for key filename. Maybe keep the relative path but rename it accordingly.~~ Just use Substring(LastIndexOf(pathSeparator)) in whatever language
[x] Return src info (BD, HDTV, www, etc) in /api/ep. In level=3 I guess, as it's different for each file. EDIT: Fixed in v3 with new endpoint /api/v3/File/{id}/AniDB
[x] v3 equivalent of /api/file/recent
[x] Fix the endpoint for changing password. The current needs the jmmuserid (#818 )

May 18 '20 10:05 harshithmohan

Need clarification on: Return count of missing episodes (collecting), missing episodes (total). Count as in count of episodes and not count of groups. Return count of unreconized files, multiple files and duplicate files. Just want an endpoint with that data? Is there a place you expected it that doesn't have it.

May 19 '20 04:05 da3dsoul

As for the Import Folder stuff, it was copy and pasted from APIv1, so it's a mess that I need to rewrite from square one

May 19 '20 06:05 da3dsoul

Need clarification on: Return count of missing episodes (collecting), missing episodes (total). Count as in count of episodes and not count of groups. Return count of unreconized files, multiple files and duplicate files. Just want an endpoint with that data? Is there a place you expected it that doesn't have it.

Just send that data in the dashboard endpoint, that's the only place it's needed.

May 19 '20 11:05 harshithmohan

The number of multiple files was scrapped, because it's confusing at its very concept. I know the system, and coming up with math that made sense wasn't easy. So, I gave Percentage of Multiples/Duplicates: how much of the collection is a multiple or duplicate.

May 19 '20 22:05 da3dsoul

Essentially, because totalling these gives very skewed data, based on the fact that Kaguya has anywhere from 1-3 files per episode, and all of A Dude and His Cat has exactly 2 files/ep.

Screenie

ShokoDesktop_2020-05-19_18-29-09

May 19 '20 22:05 da3dsoul

Just a quick one, I think

Make /api/file/recent return only the filename instead of the relative path for key filename. Maybe keep the relative path but rename it accordingly.

Isn't too needed, if you have the path, there are many libraries to extract the filename/basename from it.

May 19 '20 22:05 Cazzar

I agree

May 19 '20 22:05 da3dsoul

Trying to figure out how to have certain properties have one DateTime style and the rest something different.

May 19 '20 22:05 da3dsoul

~~Why didn't I think of that?~~ proving I'm stupid I'm marking that as done. EDIT: Well, what do you know? I did do that. I don't know why I put that there then..

let displayName = filename.split('/');
displayName = displayName[displayName.length - 1];

May 20 '20 08:05 harshithmohan

@da3dsoul Did you complete the 4th one (Return epname and series name)? You marked it as complete but I still don't see it in the endpoint response.

May 20 '20 08:05 harshithmohan

I gave you ID, as names are user configurable and language dependent, plus you could use that to link to the series/episode in ClientV2. Though, in the future you'll want to use APIv3 anyway

May 20 '20 08:05 da3dsoul

Well, fair enough. At least saves me one API request of getting the ID.

May 20 '20 09:05 harshithmohan

APIv3 models are made light so you can grab a bunch really quick

May 20 '20 09:05 da3dsoul

For File/Recent, I would have it return the new FileDetailed model, which includes the relevant xrefs. If there are none, then it's unrecognized

Jun 05 '20 21:06 da3dsoul

The reason I asked that is, currently I have to run a for loop to separate out the recognized and unrecognized files. If I'm not wrong, the server can differentiate them without a loop? So having different endpoints would be better. The loop isn't slow in anyway, just trying to avoid unneeded things. But I can work with it, no problem.

Jun 06 '20 14:06 harshithmohan

No the server just runs a loop, too.

Jun 06 '20 15:06 da3dsoul

Ok then. I was wrong.

Jun 06 '20 18:06 harshithmohan

Those are done now.

Jun 17 '20 09:06 da3dsoul

API v3 equivalents needed

[x] /version
[ ] /webui/update and /webui/latest
[x] All init endpoints
[x] /os/drives and /os/folder
[ ] Queue status and commands related endpoints
[ ] /serie/infobyfolder
[x] /folder/import, ~/folder/scan~ - Scan is pointless, use import (Run Import) or folder/id/scan

Actions:

[x] ~/rescanunlinked~ - APIv2 version is replaced by folder/import
[x] /avdumpmismatchedfiles - *
[x] /anidb/updatemissingcache - *
[x] ~/trakt/scan~ - APIv2 version is replaced by folder/import
[x] /tvdb/regenlinks - *
[x] ~/tvdb/checklinks~ - This was a test endpoint for making the algorithm.
[x] /moviedb/update - This may be added to actions anyway as a different kind of endpoint, but the APIv2 version is replaced by folder/import

I'll add if any more come to my mind.

Jun 18 '20 22:06 harshithmohan

* I can do this, but it should be in a separate menu for power users, alongside several other dangerous commands

Jun 21 '20 09:06 da3dsoul

Just for your info, Run Import checks for new files, hashes them etc, Scans Drop Folders, Checks and Scans for community site links (tvdb, trakt, moviedb, etc), and Downloads missing images. Also some cleanup stuff, but that's not important to mention

Jun 21 '20 09:06 da3dsoul

I can do this, but it should be in a separate menu for power users, alongside several other dangerous commands

Those actions will be hidden until and unless the user specifically goes and enables them to show up on the dashboard. So that's okay.

Jun 21 '20 09:06 harshithmohan

For ShokoMetadata and future Jellyfin plugin

[x] ~~/ep/getbyfilename~~ DO ENOUGH RESEARCH!
[x] ~~/serie/fromep~~

Jun 21 '20 20:06 harshithmohan

@Cazzar Can you get the Filesystem endpoints (os/drives and os/folder)? The point of them is to be able to remotely browse for an import folder when adding them, but linux and docker stuff gets in the way there. Feel free to not use CloudFileSystem there. I have tried that and didn't have much luck, but I know you are more familiar with the implications.

Jun 22 '20 01:06 da3dsoul

Init is done. Note that you need to use the settings controller during init for settings things. init user should have access

Jun 22 '20 03:06 da3dsoul

Need v3 of /api/trakt/code or just copy-paste the trakt ones to v3.

Jun 23 '20 16:06 harshithmohan

[x] Need an endpoint for getting the list of unrecognized files. The current way I'm doing it is very stupid.

Jul 06 '20 10:07 harshithmohan

[x] Add a new property property Name on Shoko.Server.API.v3.Models.Shoko.Episode, where Shoko decides what the preferred name/title of the episode is.

Sep 22 '20 15:09 revam

Closing this for now, it's easier to ask/be-asked on discord, and we haven't used this issue in... a while.

Apr 07 '23 04:04 revam