Elements scripts = doc3.select("script");
Its contents has then to be extracted as text by Element#text() and parsed further by a different library which is capable of parsing JS code, such as Mozilla Rhino. You could of course also perform trivial String parsing using indexOf(), substring(), etc methods or perhaps even using some good regex.
Jsoup is a HTML parser, not a JS parser. Best what you could get with Jsoup is getting the HTML <script> element(s).