Since most of those are run commercially and don't make their data easily accessible, that'll be a much different process, I assume. You'll basically have to scrape them like any other web site, except you'll specifically be targeting the edit/source view pages. Then find a wiki implementation that has as close a syntax as possible to the one they use (that could be tricky ...) and upload there. So unless you happen to find some code from someone who wanted to do the exact same thing, I'm afraid this would involve quite some programming/scripting.