Install:
- Python 3, at least Python 3.6.
- Node.js, at least Node.js 18.
- One way to manage Node versions is using nvm.
- Black, if you plan on contributing changes to the Python backend.
In the root directory, run:
pip install -r requirements.txt
to install dependencies.npm install
to install dependencies.
There's the frontend, which is the website and the interface. Then there's the backend, or the schedules, which are the files that have information about class schedules.
To spin up the site, we need two steps:
(1) We need to update the backend to get the data. cd scrapers
then run python update.py
.
(2) We then can update the frontend, via running npm run dev
. This will start a developer server. Open a browser tab to http://localhost:5173/
, which will update live as you edit code.
If this is the first time you're spinning up the website, the two steps have to be taken in order: step (1), and then step (2). If not followed, you'll see a blank frontend.
After the first time, the step order doesn't matter to bring up the site; in fact, backend step (1) can even be skipped -- since you'd already have locally cached data. Though backend commands are still necessary if you'd like to keep the data updated.
Before making commits, run black .
(for the backend) and npm run format
(for the frontend) to ensure that your changes comply with the project's code style. (This will also get checked by CI when you open a pull request.) Both Black and Prettier have editor integrations as well.
Let's say you're updating from e.g. Spring 2023 to Fall 2023.
First, archive the old semester. Make sure you have updated schedule files. Then run mv public/latest.json public/s23.json
.
Then, update the new semester. Open public/latestTerm.json
, change urlName
to f23
, and update the dates per Registrar.
Finally, run the normal update process and commit the results to the repo.
The server's frontend updates based on the deploy
branch on GitHub, so any changes pushed there will become live.
The server's backend is updated through a cron script that runs update.py
every hour.
See deploy/README.md
for more info.
I want to change...
- ...the data available to Hydrant.
- The entry point is
scrapers/update.py
. - This goes through
src/components/App.tsx
, which looks for the data. - The exit point is through the constructor of
State
insrc/lib/state.ts
.
- The entry point is
- ...the way Hydrant behaves.
- The entry point is
src/lib/state.ts
. - The exit point is through
src/components/App.tsx
, which constructshydrant
and passes it down.
- The entry point is
- ...the way Hydrant looks.
- The entry point is
src/components/App.tsx
. - We use Chakra UI as our component library. Avoid writing CSS.
- The entry point is
Try to introduce as few technologies as possible to keep this mostly future-proof. If you introduce something, make sure it'll last a few years. Usually one of these is a sign it'll last:
- some MIT class teaches how to use it
- e.g. web.lab teaches React, 6.102 teaches Typescript
- it's tiny and used in only a small part of the app
- e.g. msgpack-lite is only used for URL encoding, nanoid is only used to make IDs
- it's a big, popular, well-documented project that's been around for several years
- e.g. FullCalendar has been around since 2010, Chakra UI has a large community