MerC

A Python tool for detecting and translating easy to translate C macros

Dependencies:

• Python 3.10
• Maki

Instructions for using MerC

1. Download and extract source for target program.

For example, if running on bc:

wget https://mirrors.kernel.org/gnu/bc/bc-1.07.tar.gz
tar -xzf bc-1.07.tar.gz

Ensure you have the bc dependencies:

apt-get install ed
apt-get install texinfo

2. Use Bear or cmake to intercept the build process to create a compile_commands.json

For example, if running on bc, from within the bc directory:

bear -- ./configure
bear -- make

You should now have a compile_commands.json in the directory of your target program.

3. Run run_maki_on_compile_commands.py from MerC

The format for running the script is as follows:

python3 run_maki_on_compile_commands.py \
    -p <path to maki plugin> \
    -i <target program source directory> \
    -c <path to compile_commands.json> \
    -o <path to output maki analysis file, default is analysis.maki> \
    -j <number of threads, default is number of CPUs on system> \
    -v <verbose> \
    --cache-dir <optional, use directory to store intermediate analysis results> 

For example, running on bc may look like this:

python3 run_maki_on_compile_commands.py \
    -p ../maki/build/lib/libmaki.so \
    -i ../bc-1.07/ \
    -c ../bc-1.07/compile_commands.json

You should now have a generated analysis.maki file in the MerC directory (unless you specified another output directory)

4. Run emit_translations.py on the analysis.maki file

The format for running the script is as follows:

python3 emit_translations.py \
    -i <path to target program source directory> \
    -m <path to maki analysis file> \
    -o <(required) path to output directory for translation> \
    -v <verbose>

For example, running on bc may look like this:

mkdir output
python3 emit_translations.py -i ../bc-1.07/ -m analysis.maki -o ./output

You should now have a translated source for your target program.