Fix: URL-encode credentials in connectToNzbGet() to handle special characters

[MichielDean]

Description

• Fixed connectToNzbGet() to URL-encode credentials before embedding them in the XML-RPC URL. Previously, special URL characters (@, #, %, $) in ControlPassword would corrupt the URL and crash the script with socket.gaierror: [Errno 11001] getaddrinfo failed.
• The code already had a # TODO: encode username and password in URL-format comment noting this needed to be fixed.
• This is the same bug fixed in Extension-Completion#25, Extension-FakeDetector#11, and Extension-PasswordDetector#6.

The problem

# TODO: encode username and password in URL-format
rpcUrl = 'http://%s:%s@%s:%s/xmlrpc' % (username, password, host, port)

Special characters in the password corrupt the URL structure:

• @ separates credentials from host -> password spills into host/port position
• # starts the URL fragment -> everything after # is discarded, text before # becomes the hostname
• % is interpreted as a percent-encoding prefix

The fix

from urllib.parse import quote

rpcUrl = 'http://%s:%s@%s:%s/xmlrpc' % (quote(username, safe=""), quote(password, safe=""), host, port)

Testing

• test_download_another_release_special_chars_password - integration test running the full script end-to-end with a password containing @, #, %, $
• test_connectToNzbGet_encodes_special_chars - unit test verifying URL-breaking characters are percent-encoded in the XML-RPC URL
• test_connectToNzbGet_normal_password - happy-path unit test verifying alphanumeric credentials produce a valid URL unchanged
• Existing test_downlaod_another_release (with TestPassword) serves as the integration happy-path test

[dnzbk]

Thank you