python-tinycc

Module to use the Tiny C Compiler for inlining C Code in Python.

Tested with:

• CPython 2.7.6 and PyPy 2.2.1 Linux (both 64bit on x86_64, Ubuntu 14)
• CPython 2.7.5 Windows (32bit on x86, Windows XP)
• CPython 2.7.8 Windows (32bit on x86_64, Windows 10)
• CPython 2.7.9 and PyPy 4.0.1 Linux (both 32bit on ARMv7 emulated, Raspberry Pi 3, Raspbian Jessie armhf)
• CPython 2.7.3 Linux (32bit ARMv7, Wandboard, Ubuntu 12 armel)
• CPython 3.4.3 Linux (64bit on x86_64, Ubuntu 14)

Get the Tiny C Compiler

Linux

Run the following commands:

cd <modulepath>
git clone git://repo.or.cz/tinycc.git
mkdir build
cd build
../tinycc/configure --prefix=../linux --with-libgcc --disable-static
make all
make install

Windows

For Windows the repo contains a binary version (x86 only). It was build from source with MinGW:

cd <modulepath>
git clone git://repo.or.cz/tinycc.git
mkdir build
cd build
../tinycc/configure --prefix=../win32 --disable-static --extra-ldflags=-static-libgcc
make all
make install

Examples

A simple example with the 'run' state:

>>> from tinycc import TinyCC
>>> c_code = '#include <stdio.h>'
>>> c_code += 'void main(void) {'
>>> c_code += '    printf("Hello World!\n");'
>>> c_code += '}'
>>> state = TinyCC().create_state('run')
>>> state.compile(c_code)
>>> state.run([])

Example with inline code:

>>> from tinycc import TinyCC, InlineGenerator
>>> from ctypes import c_int
>>> 
>>> gen = InlineGenerator()
>>> 
>>> # C function to be used from Python
... @gen.c_function(c_int, c_int, c_int, c_int)
... def add_mul(a, b, c):
...     "return mul(a + b, c);" #  calls the Python function mul
... 
>>> # Python function to be used from C
... @gen.callable_function(c_int, c_int, c_int)
... def mul(a, b):
...     return a * b
... 
>>> # compile the code
... state = TinyCC().create_state()
>>> state.compile(gen.code)
>>> state.relocate()
>>> 
>>> # bind to state for symbol resolution
... gen.bind_state(state)
>>> 
>>> # use it
... add_mul(23, 42, 7)
455

See the example files for more usage ideas.

TODO

• rework error handling
• Testing
• PyPI package