Error with subsynchro - HTTPError

[Supernaz]

Describe the bug
Getting errors when searching for subtitles with subsynchro.

To Reproduce
Steps to reproduce the behavior:
Go to a movie.
Click on search
Go to System > Logs
Throttling subsynchro for 10 minutes, until 26/05/25 22:15, because of: HTTPError. Exception info: "'404 Client Error: Not Found for url: https://www.subsynchro.com/include/ajax/su' ~ subsynchro.py@91"

Expected behavior
The provider shouldn't raise any errors.

Software (please complete the following information):
Bazarr: 1.5.6
Radarr version: 6.1.1.10360
OS: Win11 x64
Python Version: 3.13.11.1

[morpheus65535]

They've completely reworked their website 6 months ago and the API we were using doesn't exist anymore. We would need to refactor this provider to do web scrapping instead. I'm not quite inclined to go down that path when we try to avoid this. Maybe you could contact the website admin to see if they can provide a new API?

[luuvasara]

@morpheus65535 you're spot on — confirming with specifics, and following up on your suggestion to contact the site admin:

• The endpoint the provider uses, …/include/ajax/subMarin.php, now returns HTTP 404 (UA-independent, with and without params), which is what trips the 12h HTTPError throttle. (The …/include/ajax/su… in the original log is just a truncated line; the full path is subMarin.php.)
• Per the site's own changelog, the rework was forced: on 2026-02-04 they lost their entire subtitle-file database (FTP emptied, no working backup) and are rebuilding from scratch. It's a one-person, actively-maintained PHP/MySQL hobby project.
• The new search is a POST https://www.subsynchro.com/tous-les-films.html with a multipart field q=<title> (PHP-backed .html). No JSON endpoint exists anymore — so, as you said, a fix would mean scraping the /{year}/{id}-slug.html pages for download links.

Following your suggestion, we've contacted the developer by email, asking whether they can expose a stable (ideally JSON) search endpoint and on what timeframe, and offering to handle the Bazarr-side work once there is one.

Given the Feb data-wipe (the provider would return almost nothing right now) plus the niche FR-only scope, a scraping rewrite looks low-value until either the site stabilises or the dev provides an endpoint — otherwise this may be a deprecation candidate. Will report back here when we hear from the admin.

Sources (compact)

• Endpoint gone: subsynchro.py:20 → SERVER_URL = ".../include/ajax/subMarin.php" → now returns 404 (verified with ?title=…&year=…, any UA).
• Data wipe / rework — site changelog (https://www.subsynchro.com/changelog.html):
  • Version 5 (05/02/2026): « le 04/02/2026 … tous les fichiers *.srt ont disparus du FTP … on va repartir une base de données vierge » → "on 2026-02-04 all .srt files had vanished from the FTP … restarting from a blank file database."
  • Surrounding entries show frequent PHP/MySQL breakage + repairs (one-maintainer site).
• New search endpoint — homepage search form: with → POST /tous-les-films.html (q=<title>).
• No official API historically — the site's todo page records a 2015 user request "Un API … couple avec du python" that was never implemented; subMarin.php was an internal AJAX endpoint.

[morpheus65535]

Thanks @luuvasara!