Refactoring and Cleanup web sources

The situatuon: I have a large web site which is used since 8 year for internal purposes running on linux. The site uses php, javascript and tons of css and many jquery libraries etc. As the Linux version and php version is quite old I am refactoring the code to a newly setup Ubunntu linux. Here come the question: The current web dirrctory contain 1145 directories and 9120 files with possibly a lot of files and libs which are never used over the time. I just want to copy these source files from th eold webserver to the new webserver which are really necessary to get rid of unsuded libary files.

There are many tools which do part of the job but still need a lot of work. My first attempt was to analyse the apache log to find out loaded files, the second idea was to use any kind of firefox developer tools. Both attemps have the problem that somebody (me) needs to ckick on all links and functions on the web site. But how do I know that all functions and links where clicked ? Should I just wait some month and then analyse apache logs ? Are there any tools out which can do this semi automated ?

As a result I just want to copy files from the old server to the new server which are really used, and then start the refactoring process based on these files.