A search engine for math contest problems. Deployed at search.maatester.com
You need to have nodejs, yarn, go, and docker installed on your system. On the
other hand, if you're a nix user, you can just use the provided flake.nix and
get everything going with a simple nix develop (except docker, which you
should have on your system).
Run cd ./frontend && yarn
- Start Meilisearch by running either
docker-compose up -dordocker compose up -d. - Start the backend server. In the project root, run
go run ./backend/main.go server. - Start the frontend vite server by running
cd ./frontend && yarn dev.
Run go run ./backend/main.go dump -C contests/(preference.json) (output json), which will pick a specific set of contests to fetch.
On the other hand, if you are not working with the scraper, you can always download it from the releases page.
Run go run ./backend/main.go load (json file); this
will populate the redis database with problems from the json file invoked.
Use the included ./docker-compose.example.yml. The production docker container
is monolithic and includes all the dependencies that it requires.
You must specify the runtime environment variable SOURCES with a set of
space-separated URL's that dictate where to download the dataset files.
The dataset that search.maatester.com uses is updated weekly by GitHub Actions. It can be accessed at https://github.com/MAA-Contest-Tester/search/releases/download/dataset/main.json
main.json contains a list of 17,000 problems from various short-answer and
olympiad contests. Each entry contains the following fields:
- The source of the problem and its number (i.e. 2023 IMO 1),
- A link to its source on C&P (e.g. https://artofproblemsolving.com/community/c3381519)
- A link to its solution/discussion on C&P (e.g. https://artofproblemsolving.com/community/c6h3106752p28097575)
- Its problem statement
- Any associated tags (e.g. "combinatorics," "functional equations," inequalities," ...)