This should make parsing even easier as the internal commands are tokens instead of text. Additionally it is now possible to properly include comments as they are no longer considered separate parts and instead are a single string lasting until the end of the line.
The Lexer should now be able to parse all existing Blitz code into easily identified Tokens. With this information, a Parser can now verify that the syntax was kept correctly, warn about possible dangerous or wrong use and convert the Tokens into actual LLVM code for compiling.
Boost is used to parse program arguments into a usable format, print help and so on. This makes it far easier to quickly get off the ground and is also reasonably fast with relatively small impact. Since Boost provides prebuilt versions with all recent compilers, it shouldn't impact newer people.
LLVM will be used for compiling the actual code, as of right now there is no Tokenizer, Parser or Compiler implemented. This will be the biggest task as it requires careful planning in order to not end up like Mark Sibly did with BlitzBasic and all the programming languages that ended up being created by it.
Additionally removed the VERSION_BUILD from the template so that version updates work better.
Michael Fabian 'Xaymar' Dirks