FFI Lab

This is a lab project for an upcoming workshop on FFI

Requirements

1. cargo
   1. prost
   2. pyo3
   3. tokio
   4. tonic
2. rustc
3. python3.11
4. pip: https://pip.pypa.io/en/stable/installation/
   1. grpcio
   2. grpcio-tools
   3. maturin
   4. protobuf
5. protoc: https://github.com/hyperium/tonic#dependencies

TODO: Can we use Docker to handle this for the users?

Usage

Setup

From the project root:

1. Create a virtualenv to work in: python -m venv .venv
2. Install the necessary python packages: pip install --require-virtualenv -r requirements.txt NOTE: vscode doesn't seem to handle virtualenvs very well. You can either skip making a virtualenv entirely (and the --require-virtualenv), add the new python path (located at .venv/Scripts/python(.exe)) to vscode (ctrl-shift-p -> Python: Select Interpreter), or use a different terminal for running the commands below.

Running the code

There are 2 folders, workshop and solution. As the names imply, workshop will contain partially completed code to be completed during the workshop itself while solution holds a working final version of the code at the end of the workshop. There are 3 main binaries to run (in either the workshop or solution folder):

1. cli.py: cli for executing some python code against a json database
   1. Usage: python cli.py pymk 0 (see --help for more)
2. database_server.py: a "remote" database server implemented in python to replace the in-memory database
   1. Usage: python database_server.py
3. database_server: a "remote" database server implemented in rust to replace the in-memory database
   1. Usage: cd database; cargo run --release

There is also an FFI library (the main point of the workshop) that needs to be built with cd database_ffi; maturin develop. Then you should be able to run python -c import database_ffi