llvmpy is a Python wrapper around the llvm C++ library which allows simple access to compiler tools. It can be used for a lot of things, but here are some ideas:
Get and extract LLVM 3.3 (or LLVM 3.2) source tarball from llvm.org. Then, cd into the extracted directory.
Run ./configure --enable-optimized --prefix=LLVM_INSTALL_PATH.
Note: Without the --enable-optimized flag, debug build will be selected. Unless you are developing LLVM or llvmpy, it is recommended that the flag is used to reduce build time and binary size.
Note: Use prefix to select the installation path. It is recommended to separate your custom build from the default system package. Please replace LLVM_INSTALL_PATH with your own path.
Run REQUIRES_RTTI=1 make to build.
Note: With LLVM 3.2+, the default build configuration has C++ RTTI disabled. However, llvmpy requires RTTI.
Get llvm-py and install it:
$ git clone https://github.com/llvmpy/llvmpy.git $ cd llvmpy $ LLVM_CONFIG_PATH=LLVM_INSTALL_PATH/bin/llvm-config python setup.py install
Run the tests:
$ python -c "import llvm; llvm.test()"
See documentation at ‘http://www.llvmpy.org‘ and examples under ‘test’.
This project is maintained by Continuum Analytics