IdeTwo

README

IdeTwo

A (very bare-bones) clone of Ideone - the online IDE, compiler and debugging tool. It supports C, C++, Java and Python 3 out of the box, but can be extended to be used with any language.

Architecture

The program consists of an Angular front-end and a Gin backend. The backend uses MongoDB for data storage, Faktory to queue up execution requests and Docker to containerize source code compilation and execution.

Whenever the user makes an execution request, the request will be placed in the default queue of Faktory to be handled sequentially by the FaktoryWorker. The worker compiles (for C/C++/Java sources) and executes each request in Docker containers with the following specification:

• 1 CPU core
• 1 GB of RAM
• 30 second of run time

If the source code fails to compile, takes too long to compile or execute, or exits with a non-zero code, the job is canceled and the error log and exit code are stored in the database.

Installation

MongoDB, Faktory and Docker Engine are required to run this program. This can be easily achieved by installing Docker alone, then run a MongoDB and a Faktory container. Configuration to connect to these two services can be found in .env.

Since the front end is a separate project inside the views folder, you will need to compile the front end first before compiling and running the back end.

# Compile the front end.
cd views
npm install
ng build -c production
cd ..

# Compile the back end.
go build main.go

# The program uses Docket Engine's Golang SDK underneath, so we need to config
# environment variables as described in
# https://pkg.go.dev/github.com/docker/docker/client#FromEnv.
export DOCKER_HOST=tcp://localhost:2375

# Run in release mode. Skip this line if running in debug mode.
export GIN_MODE=release

# Start the server to listen to port 8080.
./main

Configuration

These field can be edited inside the file .env:

Field	Description
IDETWO_EXECUTIONS_DIR	The root temporary directory for executions. Usually there is no reason to configure this.
MONGODB_URI	The URI to connect to the MongoDB instance.
MONGODB_DB	The database name of the program.
FAKTORY_URL	The address of the Faktory server.
FAKTORY_POOL_CAPACITY	The number of Faktory connections to maintain.
FAKTORY_WORKER_CONCURRENCY	The number of goroutine used to handle worker requests.

Support for other programming languages

FaktoryWorker uses the JobExecutor interface as an abstraction of the underlying compilation + execution logic for each programming language. By extending this interface, you can add support for any programming languages.

Detailed steps to add support for a new programming languages:

1. Add the programming language to the ProgrammingLanguage enum inside source_code.go.
2. Implement a new struct that implements the JobExecutor interface.
3. Add the struct as a case inside the executeJob() function in FaktoryWorker.
4. Add the programming language to the ProgrammingLanguage enum inside source_code.ts, plus editing related functions in the file.
5. Profit 💰

Licence

MIT