It will be consistent with the logic used by the codebase that actually fetches the URL (curl), so it won't leave gaps between the expected behavior and actual

It will be consistent with the logic used by the codebase that actually fetches the URL (curl), so it won't leave gaps between the expected behavior and actual behavior the way two disjoint implementations would.