Skip to content
This repository has been archived by the owner on Jun 7, 2022. It is now read-only.

Latest commit

 

History

History
132 lines (96 loc) · 2.68 KB

README.md

File metadata and controls

132 lines (96 loc) · 2.68 KB
Raw

RustQL: a Tool for Semantic Querying of Rust Code.

RustQL is a prototype of Niko Matsakis' idea to use Datalog for understanding how Rust programmers write code. The initial version was developed by Nicolas Winkler. The project is currently maintained by the Programming Methodology group at ETH Zurich.

The tool consists of three parts:

  1. Extractor – a Rust compiler plugin that extracts the information about the crate during compilation.
  2. Linker – a program that merges the information about different crates into a single database.
  3. Query Engine – a program that takes a Datalog like query and evaluates it on the database. The engine is based on the Datafrog.

Extractor

Build the extractor and prepare the environment:

cd rustql-extractor
cargo build --release
source prepare_env.sh

Extract information about the crate and its dependencies by compiling it:

cd <some-crate>
cargo clean
cargo build

This will emit information about the compiled crates at ~/.rustql/crates/.

Linker

To create the database, compile the linker and run it:

cd rustql-linker
cargo run --release

This will create a file database.db in the current directory.

Query Engine

When you run a query engine, it will enter a loop in which you can provide paths to queries to you want to execute:

cd ../rustql-query/
env RUST_BACKTRACE=1 cargo run --release

Example queries to try out:

samples/same_type.rql
samples/thief.rql
samples/unsafe.rql

Getting Started

This tutorial will show how to download most popular 500 crates and run a simple query over them.

Check Out Submodules

git submodule update --init

Set Up Vagrant

Note: If you are running on a server, you probably want to allocate a larger VM by editing Vagrantfile.

Install Vagrant and start the VM:

vagrant up
vagrant ssh

All remaining steps should be done inside the VM from the directory /vagrant that is mapped to the directory on the host in which the repository was cloned.

Download Top 500 Crate Metadata

cd top-crates/
cargo run
cd ..

Run Extractor

cd rustql-extractor/
cargo build --release
cd ..
mkdir -p data/cache
python3 bin/run_extractor.py
rm -f /vagrant/data/crates/rustql_dummy_*

Link

cd rustql-linker
export EXTRACTOR_TARGET_DIR=/vagrant/data/crates/
cargo run --release
cd ..

Start Query Engine

cd rustql-query/
RUST_BACKTRACE=1 cargo run --release
cd ..