PDF Trimmer Worker

Usage: GET /?url=<page-url>
Optional: &sources=domainA.com,domainB.com to override the image-source match list.
Note: only pages on pre-approved domains can be scanned (see ALLOWED_PAGE_DOMAINS in the worker).

It finds a matching <img> inside a link on that page, resolves the link to a PDF (direct PDF link or Google Drive file), strips the first 2 pages, and returns the result.