Brain F interpreter and Compiler with TCC

Experimenting with using TCC to compile to binaries. I didn't love it, and I think I almost prefer compiling to assembly with nasm. Started out in pure C, but then constructing strings got really annoying so I switched to C++. I've included a single additional langauge command, ~ which dumps 100 cell values for debugging

bf compiler

Run the interpreter

bf hello_world.bf

Compile to binary

bf -c hello_world.bf

TCC as a backend

As previously mentioned it wasn't great. First of all TCC has very scare documentation so it took a while to figure out how to use libtcc as a library. Also one thing I've found very strange is that you must link to libtcc1.a at runtime. This is very strange to me, I'm not sure why it isn't just a dynamic library instead of an archive file, and more than this why isn't it a real static library that I can just link with at compile time. Another annoyance is that TCC seems to be really particular about syntax, for example, it wouldn't allow me to use #define after the includes, in fact I couldn't get it to allow the #defines I wanted at all. It seems that TCC can't link to normal static libraries built with clang or another compiler. I tried a project using Raylib and could only successfully link with a dynamic library. I thought maybe I could get around this by compiling to an object file and then linking with the system linker, however, TCC uses some custom format for its archive and object files and ld doesn't actually understand how to link tcc .o files. I think maybe TCC is sort of like emscripten and wasm. If you want to use libraries they must be compiled from source using the compiler in order to be compatible.

Building

requires C++ 17. makefile provided for unix (developed on Mac OS).

make tcc
make
./bf
Usage: bf [options] file_path

Options: 
    -h, --help      Display this message
    -c              Compile to executable