subliminal
subliminal copied to clipboard
Subscene provider implementation
Please discuss about anything related to Subscene as a provider for this project here.
Sadly this is far from merge-ready imho.
In the worst case this provider hits SubScene on a massive scale for just one video query. Not really your fault, though, it's their tedious website implementation and lack of consistent data. If they'd include the series details (IMDB ID etc.) on the subtitle detail page, it would be a lot easier.
This PR's state is unclear to me. I prefer to get it done and finish writing tests and make it work but the issue is I don't even know what should we do and think of some possible solutions:
- just writing tests and merge it as is; ignoring traffic efficiency
- rewriting the provider with probebly a different design
- giving up about SubScene
Since I've forgotten about the API and architecture because it was for a while ago, it would be easier to do that from scratch.
@pannal I just checked this subtitle page in Subscene for Altered Carbon Season 1 and it seems to have an IMDB link to the series.
@euphoria360 @mamins1376 I've made a separate implementation of subscene for Sub-Zero. It relies heavily on Subscene-API, with some slight modifications.
Also parts of your initial PR have been used, but also heavily modified.
Sadly I can't really merge this into subliminal itself, because it has ties to certain data structures the Plex subsystem provides.
What's wrong with python 3.4? seems like a dependency VersionConflict :/
Can you merge develop and try again?
Maybe this is finally here to be merged? :)
Maybe this is finally here to be merged? :)
We all hope so!:smiley:
Well, there is an issue with Cloudflare bot-detector somtimes, and I feel like being able to get over this is a good idea. I recently found Anorov/cloudflare-scrape, though didn't played with it. If you think it's worth adding another dependency, I can take some time to try this.
Update: It looks like the package requires nodejs. this would be a serious overkill.
is it possible to search for Persian subtitles too? In Subscene website they are shown as "Farsi/Persian". and their ISO 639-1 code is "fa" and ISO 639-2 is "per"
Not sure, but I suppose they are actually equivalent.
On Fri, Jul 5, 2019, 1:20 PM euphoria360 [email protected] wrote:
is it possible to search for Persian subtitles too? In Subscene website they are shown as "Farsi/Persian". and their ISO 639-1 code is "fa" and ISO 639-2 is "per"
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Diaoul/subliminal/pull/740?email_source=notifications&email_token=ACOFOYIMENBM6KBY6EAMU7TP54DO3A5CNFSM4DC4ISI2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZI6RUY#issuecomment-508684499, or mute the thread https://github.com/notifications/unsubscribe-auth/ACOFOYN5HYEYFGVKZ5MR65TP54DO3ANCNFSM4DC4ISIQ .
As of late, subscene have implemented reCAPTCHA all over the place and have locked us out using CloudFlare. I've worked around that by implementing a CloudFlare circumvention system as well as captcha solving services.
Also the search mechanism and endpoint has changed multiple times over the last months.
Captchas are currently not thrown at people who are logged in.
All this would be a lot of work to backport from subliminal_patch, though. (For reference: https://github.com/pannal/Sub-Zero.bundle/blob/master/Contents/Libraries/Shared/subliminal_patch/providers/subscene.py, https://github.com/pannal/Sub-Zero.bundle/blob/master/Contents/Libraries/Shared/subliminal_patch/http.py#L70)
Cleaning up old PRs. Subscene.com closed down.