Interesting... Assuming the article is correct in that Google's tool says that no copyrighted information is left, it seems that Google is saying that header files stripped of all of english/whatever-language text and left with only pure code are not copyright-able because it's just code?