Simple source code syntax converter from objc2java based on ANTLR lexer parser.

Originally based on https://code.google.com/p/objc2j/ but basically rewritten from scratch
so changed package and author names since i'm not sure the originally author (Andremoniy) would subscribe to it
or even recognize it:) nevertheless i use this place to thank Andremoniy for giving me the idea.

The antlr-master-3.4-completejar.jar this works with is a part of this repo.
To make ANTLR 3.4 work with Eclipse 4.2.1 I had to build it on Mac from source.
Once that'd been done it worked quite nicely.

Limitations:
- It does only what i need to ease my porting work from iOS/Mac to java
- it doesn't create buildable code
- it doesn't fully cope with C++ syntax (e.g. the gramma doesn't recognize some C++ declarations)
- it doesn't bother with categories declarations as it expects the functions to be implemented somewhere
- it doesn't merge categories either
- it doesn't create separate files in case it finds more interfaces in one .mm/.m, .h file pairs
- it only copes with 1 line #define extept when macro has a parameters then it turns it in sort of typeless function
- it preferrably wants all macros to be defined outside @interface,@iplementation,etc problems:
	- i haven't found a simple way to terminate macro by \n which is a whitespace on hidde channel
	- the syntax to alter whitespaces per rule (e.g. options = {ignore=...}; ) has disappeared in ANTLR 3.x apparently
	- i figured i would need to start writing gramma functions to handle that which i was too lazy to do:)
- it doesn't cope with macro expansion and treats it as a C function call if it sees that left bracket follows
	- leads to recognition exception if macro "parameter" contains anything else but expression
- the gramma is just to keep me happy and
	- doesn't include any operator precedence or any other stuff of that sort useless for this type of conversion
	- due to the above it produces 5 warnings when compiled with ANTLR which i couldn't be bothered to fix:)
- there seem to be a bug in ANTLR in a way it treats '\' char as something special and hnece it wouldn't parse string const split over multiple lines. so i ended up allowing '\' in string efectively forgeting about any 'proper' escaping which produced another 6th warning but worked.
- i'm sure you'll find plenty of others

Usage:
Load the project to Eclipse (with ANTLR installed) build and run it with 2 parameters stating root dir to files to convert
and files to produce. The paths must end with '/' 
It's producing the best results (given the limitations) when it finds matching 
<name>.mm/.m and <name>.h files (same <name> in the same directory).Then it merges interfaces into a single
<name>.java file.
And now the hard bit:
After it converts you load the stuff to some java IDE your project and make it build:)

Disclaimer:
Anyone who tries to use it and gets frustrated please don't panic it's just a piece of code that can be amended:)

License:
You can do whatever you like with it. In case you want to help to improve it you'r welcome to contact me.
In case anyone wants to play with gramma i'd recommend ANTLRWorks 1.4.3 < to debug and visualize it.

Have fun