fix(ss): tolerate ScreenScraper's invalid JSON escapes

[gantoine]

Description
^{Explain the changes or enhancements you are proposing with this pull request.}

ScreenScraper's API occasionally returns malformed JSON: some text fields (e.g. game synopses) contain raw, unescaped backslashes, producing sequences like \e that aren't valid JSON escapes. Python's strict parser (via aiohttp's res.json()) rejects the entire response with:

[screenscraper] Error decoding JSON response from ScreenScraper: Invalid \escape: line 26 column 11

_request() caught the JSONDecodeError and returned {}, which reads as "no match". So ScreenScraper silently found nothing, even for trivially-matchable titles like Banjo-Kazooie, while IGDB/SGDB/Libretro (clean JSON) matched fine.

Fix: parse leniently. On a decode error, double any backslash that isn't a valid JSON escape and retry, instead of discarding the whole response. Valid escapes (\n, \", \uXXXX, etc.) are preserved; if the retry still fails, the existing {} fallback applies.

Applied at both request sites via aiohttp's res.json(loads=...) hook to keep the change minimal.

Fixes #3764

Note: the PlayMatch error in the same issue is tracked separately and already fixed for the next release. The issue also mentions LaunchBox, but there was no LaunchBox error in the reporter's logs to diagnose.

Checklist
^{Please check all that apply.}

• I've tested the changes locally
• I've updated relevant comments
• I've assigned reviewers for this PR
• I've added unit tests that cover the changes

Screenshots (if applicable)

AI assistance disclosure: This PR was written with AI assistance (PostHog Code / Claude). The root-cause analysis, code change, and tests were AI-generated and reviewed by a human before submission.

---

Created with PostHog Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[greptile-apps]

Greptile Summary

This PR adds tolerant JSON parsing for malformed ScreenScraper responses. The main changes are:

• Repairs invalid backslash escapes before retrying JSON parsing.
• Uses the tolerant loader in both ScreenScraper request paths.
• Adds tests for malformed backslashes and valid JSON escapes.

Confidence Score: 4/5

The malformed Unicode escape path needs a fix before merging.

• A raw backslash followed by u is preserved without checking for four hexadecimal digits.
• The retry still fails for values such as C:\users, causing the response to be discarded as {}.
• Valid JSON and ordinary invalid escapes follow the intended path.

backend/adapters/services/screenscraper.py

Important Files Changed

Filename	Overview
backend/adapters/services/screenscraper.py	Adds the tolerant loader and integrates it into both request attempts, but incomplete Unicode escapes are incorrectly preserved.
backend/tests/adapters/services/test_screenscraper.py	Adds direct coverage for ordinary invalid backslashes and valid escapes, but omits malformed Unicode escape prefixes.

_{Reviews (1): Last reviewed commit: "fix(ss): tolerate ScreenScraper's invali..." | Re-trigger Greptile}