Skip to content

meyfa/ka-mensa-api

Repository files navigation

ka-mensa-api

CI Test Coverage Maintainability

Introduction

ka-mensa-api is one component in a three-part project whose goal it is to aggregate, process and visualize the Studierendenwerk Karlsruhe's canteen plans in ways superior to the official sources.

Disclaimer: This project is neither affiliated with nor endorsed by the Studierendenwerk Karlsruhe or the Karlsruhe Institute of Technology.

The entire project is written in JavaScript+TypeScript and is composed as follows:

  • ka-mensa-fetch: library package responsible for the fetching of raw plan data and conversion into canonical, easily digestible JSON documents
  • ka-mensa-api: Node.js server that utilizes the fetcher to continuously collect meal plans and makes them available via REST API
  • ka-mensa-ui: single-page web app that loads meal plans from an API instance and displays them in a modern, responsive interface with filtering and color-coding capabilities

Setup (standard)

Prerequisites

Ensure Node and NPM are available on your system. Then clone this repository somewhere and run npm install to load dependencies.

Configuration

ka-mensa-api is configured exclusively via environment variables that need to be set before starting the server. The following variables are available:

  • CACHE_DIRECTORY: The path to a directory where downloaded plans should be stored, and where they should be served from. Defaults to a cache/ inside the current working directory.
  • SERVER_HOST: The address to listen on, e.g. :: (or 0.0.0.0) for all interfaces, ::1 (or 127.0.0.1) for localhost only. Defaults to ::.
  • SERVER_PORT: The port to listen on. Defaults to 8080.
  • CORS_ALLOWORIGIN: Set this to a specific URL to allow CORS requests from that URL, or set to * to allow all origins. Defaults to no CORS headers.
  • FETCH_INTERVAL: The time between runs of the fetcher. Default: 6 hours.
  • FETCH_SOURCE: The source to fetch plans from. Available options are simplesite and jsonapi. Default: 'simplesite'.

If using the simplesite source:

  • SIMPLESITE_DAYS: The number of days to fetch from the simplesite source. More may be fetched, depending on how many the server returns for the specific request. Default: 14.

If using the jsonapi source:

  • JSONAPI_AUTH_USERNAME: Credentials for JSON API authentication.
  • JSONAPI_AUTH_PASSWORD: Credentials for JSON API authentication.

Notice that network options are rather limited. If you want HTTPS support or advanced embedding into existing domain structures, you will need to set up a reverse proxy such as nginx.

Data Source

You might want to change the plan fetch source. This will not impact the API behavior but only how data is retrieved. There are two options available:

  • 'simplesite' (the default)
  • 'jsonapi'

The simplesite source fetches plans from the Studierendenwerk's website, which offers a view for the visually impaired that is quite easy to parse. It allows for fetching plans far into the future and is relatively stable. That said, this still breaks sometimes, e.g. when the Studierendenwerk renames lines or takes other unpredictable actions.

The JSON API source is more reliable. It will basically never break. The downside: it requires authentication (you will have to figure out how to obtain credentials on your own).

In summary, both sources provide the same information but differ in stability and auth requirements.

Startup

To start the server, run npm start. It will immediately fetch the most recent set of plans, then listen for API requests. Plan polling will continue indefinitely with the configured interval.

Note that npm start is shortcut for npm run build followed by npm run production, where the former compiles the TypeScript code and the latter executes it. To start the server without compiling again, use npm run production only.

A fixup job will run at startup. This fills in missing data that might have been missing at cache time but is now inferrable due to an updated data set. Any plans that are considered complete, or that cannot be inferred further, will not be changed.

Setup with Docker

This project is available as a Docker image. See also https://hub.docker.com/r/meyfa/ka-mensa-api.

Running the Container

If you just want to get going:

docker run --name mensa -p <host-port>:8080 -d meyfa/ka-mensa-api

If you would like to put the cache in a volume (highly recommended!) so it has enough room to grow, is available after restarting the container and/or can be accessed from the host: Mount to /usr/src/app/cache like so:

docker run \
        --name mensa \
        -p <host-port>:8080 \
        -v /path/on/host:/usr/src/app/cache \
        -d meyfa/ka-mensa-api

Any configuration can be done via environment variables (-e VAR=value). The available variables are the same as for the standard setup.

CORS Headers

Perhaps you need to configure CORS for the API server - to enable displaying the plans via an instance of ka-mensa-ui running on another domain, for example:

docker run \
        --name mensa \
        -p <host-port>:8080 \
        -v /path/on/host:/usr/src/app/cache \
        -e CORS_ALLOWORIGIN=https://example.com \
        -d meyfa/ka-mensa-api

Specify a regular URL to only allow that one origin. Use * to allow all origins.

Development

Contributions are welcome. Guidelines:

  • By contributing, you agree to make your changes available under the MIT license of this project.
  • Please write unit tests for as much code as possible.
    • To run: npm test
    • With coverage: npm run coverage
  • Make sure to adhere to JS standard style and proper usage of TypeScript.
    • Linter: npm run lint
    • Automatic fixing of most style issues: npm run lint-fix

Documentation

A document outlining the API provided by this software is available: API_DOCUMENTATION.md.