In production our APIs run via open-runtimes on self-hosted AppWrite functions.

Locally, you can test by running server.js

First, build whatever API Functions you wish to test. cd into the function's directory within functions containing package.json, and run npm install && npm run build.

Then to start the server:

• cd function_server && npm install
• Set the following environment variables
  • PGUSER - Postgres User of the Phoenix DB
  • PGHOST - Hostname of the Phoenix DB
  • PGPASSWORD - Password of the Phoenix DB
  • PGDATABASE - Database Name
  • PGPORT - Port of the Phoenix DB
• npm run start

The server will listen on localhost port 3000. Test an API by adding its path within functions to the end of the URL. For example, the following tests Case GET:

curl -d '{"payload": {"id": 120}}' localhost:3000/Case/GET/v2

Or to get all cases: curl -d '{}' localhost:3000/Case/GET/v2

Some REST objects are defined in models, which needs to be built and deployed separately here.

(Copied from the AppWrite documentation)

There are two ways of deploying the Appwrite function, both having the same results, but each using a different process. We highly recommend using CLI deployment to achieve the best experience.

Make sure you have Appwrite CLI installed, and you have successfully logged into your Appwrite server. To make sure Appwrite CLI is ready, you can use the command appwrite client --debug and it should respond with green text ✓ Success.

Make sure you are in the same folder as your appwrite.json file and run appwrite deploy function to deploy your function. You will be prompted to select which functions you want to deploy.

Manual deployment has no requirements and uses Appwrite Console to deploy the tag. First, enter the folder of your function. Then, create a tarball of the whole folder and gzip it. After creating .tar.gz file, visit Appwrite Console, click on the Deploy Tag button and switch to the Manual tab. There, set the entrypoint to src/index.js, and upload the file we just generated.