Note that while your switch() will only call the function needed, ALL the functions will be loaded with the file, whether they’re called or not, if they exist in the file. This is unneeded overhead.
If you have a file of say 200k, consider a redesign of the application. You could still use the switch() method…but instead, separate out your functions into separate, logical files. This way, you can then simply include the files (er, functions) that are needed at the given moment without extra overhead.