Skip to content

Tool for building tasks for informatics competitions, with support for cache, distributed computations and more

License

Notifications You must be signed in to change notification settings

edomora97/task-maker-rust

Repository files navigation

Workflow Status

task-maker-rust

The new cmsMake!

asciicast

Installation

For Ubuntu and Debian users you can find the .deb file in the Releases page. Install the package using sudo dpkg -i the_file.deb and it's dependencies (if you need to) with sudo apt install -f. There is a good chance that you have already all the dependencies already installed.

For Ubuntu it's also possible to install task-maker-rust using a PPA:

sudo add-apt-repository ppa:dariop1/task-maker-rust
sudo apt update
sudo apt install task-maker-rust

For ArchLinux users you can find the packages in the AUR: task-maker-rust (the stable release) and task-maker-rust-git (the version based on master).

For MacOS users you can install the package using Homebrew: brew install bortoz/bortoz/task-maker-rust.

For the other operating systems the recommended way to use task-maker-rust is the following:

  • Install the latest stable rust version (and cargo). For example using rustup
  • Install the system dependencies: libseccomp or libseccomp-dev on Ubuntu
  • Clone this repo: git clone https://github.com/edomora97/task-maker-rust
  • Build task-maker: cargo build --release

The executable should be located at target/release/task-maker. Due to limitations of cargo (the build system), cargo install should not be used since it doesn't copy some required files. For the same reason you should not delete or move the cloned repository after the build. If you need a package for your operating system/distro open an issue please!

The supported operating systems are Linux (with libseccomp support), OSX and Windows under WSL2. It should be possible to build task-maker using musl but it may be hard to link libseccomp!

Usage

Simple local usage

Run task-maker-rust in the task folder to compile and run everything.

Specifying no option all the caches are active, the next executions will be very fast, actually doing only what's needed.

Disable cache

If you really want to repeat the execution of something provide the --no-cache option:

task-maker-rust --no-cache

Without any options --no-cache won't use any caches.

If you want, for example, just redo the evaluations (maybe for retrying the timings), use --no-cache=evaluation. The available options for --no-cache can be found with --help.

Test only a subset of solutions

Sometimes you only want to test only some solutions, speeding up the compilation and cleaning a bit the output:

task-maker-rust sol1.cpp sol2.py

Note that you may or may not specify the folder of the solution (sol/ or solution/). You can also specify only the prefix of the name of the solutions you want to check.

Using different task directory

By default the task in the current directory is executed, if you want to change the task without cd-ing away:

task-maker-rust --task-dir ~/tasks/poldo
Extracting executable files

All the compiled files are kept in an internal folder but if you want to use them, for example to debug a solution, passing --copy-exe all the useful files are copied to the bin/ folder inside the task directory.

task-maker-rust --copy-exe
Statement

If you don't want to build the statement files (and the booklet) just pass --no-statement.

task-maker-rust --no-statement

If you want just to build the statement you can use:

task-maker-tools booklet

This tool can also be used to build the contest's booklet.

Clean the task directory

If you want to clean everything, for example after the contest, simply run:

task-maker-tools clear

This will remove the files that can be regenerated from the task directory. Note that the internal cache is not pruned by this command.

Remote evaluation

On a server (a machine accessible from clients and workers) run

task-maker-tools server

This will start task-maker in server mode, listening for connections from clients and workers respectively on port 27182 and 27183.

Then on the worker machines start a worker with

task-maker-tools worker server_addr num

This will start a worker on that machine (using a single core), connecting to the server and executing the jobs the server assigns. The num parameter can be used to distinguish between multiple workers in the same machine.

For running a remote computation on your machine just add the --evaluate-on option, like:

task-maker-rust --evaluate-on server_addr

Using docker

You can easily spawn a task-maker server and a set of workers in your local machine without having to install all the compilers.

docker run --rm -it \
    --name task-maker \
    -p 27183:27183 \
    -p 27182:27182 \
    --privileged \
    edomora97/task-maker-rust:latest

Then you can use task-maker locally adding --evaluate-on localhost.

--privileged is required to run the worker sandboxes.

License: MPL-2.0