Fix embed URL for password-protected links and locale subdomains

[N0Arxan]

Summary

• Preserve the secret_password query parameter when building the embed URL, so privately-shared/password-protected Scribd documents export correctly instead of failing with "No printable document pages were detected."
• Match any *.scribd.com subdomain (e.g. es.scribd.com), not just www.scribd.com.

Test plan

• Ran the script against a password-protected www.scribd.com/document/.../...?secret_password=... link and confirmed the embed URL retains the password and the PDF exports successfully (80/80 pages).
• Verified es.scribd.com document links now convert to a valid embed URL.

[fullstackusama]

Thanks for the contribution. Supporting locale subdomains and password-protected shared links is useful.

Before merging, could you please address these points?

1. The hostname regex currently accepts unrelated domains such as notscribd.com and evil-scribd.com. Please parse the URL and accept only scribd.com or hostnames ending exactly in .scribd.com.
2. Preserving secret_password causes the complete secret-bearing embed URL to be printed by print(f"Link embed: {converted_url}"). Please redact or omit the query string in terminal output.
3. Please use parse_qs() or parse_qsl() for extracting secret_password instead of parsing the query with a regex.
4. Please add tests covering www.scribd.com, locale subdomains, encoded passwords, and rejection of lookalike domains.

The overall feature looks good, and the PR is otherwise small and mergeable. Once these security and validation concerns are addressed, I’d be happy to reconsider it.