First of all, there are complete solutions, some of which are probably already referenced on SO, but I'm not familiar with any of them, and many are likely to cost money. If I were you, I'd start searching around with the keywords "clean up css lint" and "clean up javascript lint" or something similar. I would not go and write my own code for this purpose.
Now if I weren't you or after a few hours of searching I still hadn't found anything that met my needs, I might decide to try this myself. What you could do if you're willing to get dirty, at least for the CSS, is run a script locally to run through each served HTML file and
- Extract all of the referenced stylesheets from each, making a list
- Extract the names of all CSS classes and ids from each, making a list
- Match the list against the classes and ids actually used in the HTML
- Make a list of all those that aren't, and store it in a file for easy access
You could throw something like this together in Python in maybe a few hours if you're familiar with it. Of course, this depends on what language you use for quick and dirty scripting.
Analyzing javascript files would be a bit more difficult, because of the issues with logic. You could do something similar, but it might not be in your best interests to basically write your own javascript interpreter for the purpose of cleaning out a few unused functions. Yes, you code do it, but you'd be reinventing the wheel. It might actually be a decent idea to look into IDEs that have this functionality built-in. Some are free, and most importantly, you don't have to write them yourself.
