Yes, if you write the code to do it. This would involve writing a web client which parses the page and finds the relevant link.
Provided you have permission to do this from Jupitermedia, it should be straightforward- just load the page (e.g. with file_get_contents), parse the HTML (DOM::loadHTML) and find the element you want by interrogating the DOM, read its contents etc.
Be mindful of encodings when using the DOM to parse HTML.
Mark
