Evaluate Scrapling + StealthyFetcher for Planespotters source

[michaelmorandi]

Background

Research was done evaluating whether adopting Scrapling with StealthyFetcher
would benefit the project.

Finding

The only source where adoption makes sense is Planespotters.
Current pipeline:

1. Playwright → browser visit → harvest Cloudflare cookies (cached 1h)
2. curl_cffi → reuse cookies for HTML page requests
3. BeautifulSoup → 6-method fallback chain to extract aircraft data

StealthyFetcher uses Camoufox (modified Firefox) with engine-level
fingerprint spoofing (stronger than current JS injection) and a built-in
session class that can replace all three steps.

Expected Benefits

• Remove beautifulsoup4 dependency (only used in planespotters.py)
• Eliminate ~120 lines of Playwright boilerplate + 6-method HTML parsing fallback
• Stronger Cloudflare resilience (C++ engine spoofing vs JS injection)
• Scrapling's selector API replaces manual BeautifulSoup traversal

Non-benefits

• No Docker image size reduction — StealthyFetcher still needs a browser binary, comparable size to current Chromium
• ADSBexchange: not worth migrating — JSON API + custom TokenBucketRateLimiter; cookie-harvest-once + curl_cffi pattern is more efficient
• Flightradar24: no change needed

Scope (if implemented)

• planespotters.py — replace PlanespottersCookieManager + curl_cffi session + BeautifulSoup parsing
• pyproject.toml — add scrapling[fetchers], remove beautifulsoup4
• Dockerfile — add scrapling install step; verify Firefox system deps

Notes

• Scrapling recently migrated StealthyFetcher from Camoufox to Patchright — verify current engine before implementing
• Scrapling API is still evolving; check stability before adopting