Built with Typescript connects to Supabase and runs on vercel.com.
🚨 Might become part of the giessdenkiez-de repo eventually.
- Vercel.com Account
- Supabase Account
- Supabase CLI install with brew
brew install supabase/tap/supabase - Docker Dependency for Supabase
git clone [email protected]:technologiestiftung/giessdenkiez-de-postgres-api.git
cd ./giessdenkiez-de-postgres-api
npm ci
# supabase needed for local development
supabase login
# Check if docker is running
docker --version
# then run
supabase start
# After a few minutes you will have a local supabase instance running with
# - Postgres DB at postgrsql://postgres:postgres@localhost:5432/postgres
# - Postgrest API at http://localhost:54321 a rest api for your db
# - Supabase Studio at http://localhost:54323 a gui for your db
# - Other cool things we currently don't use
# The Database will already have some seeded trees in Berlin
# Create .env file and populate with ENV variables from the supabase start command
# You can always get the values again by running `supabase status`
cp .env.example .env
# SERVICE_ROLE_KEY=...
# SUPABASE_URL=...
# SUPABASE_ANON_KEY=...
# SUPABASE_MAX_ROWS=1000In the example code above the Postgres database Postgrest API are run locally. You SHOULD NOT use production variables in your local or CI environments. The tests will modify the database and also truncate tables through the API and also with direct calls.
Again. Be a smart developer, read https://12factor.net/config, https://github.com/motdotla/dotenv#should-i-have-multiple-env-files and never ever touch production with your local code!
Setup your Vercel.com account. You might need to login. Run npx vercel login in your shell. You will have to link your local project to a vercel project by running npx vercel link and follow the instructions or deploy your application with npx vercel. This will create a new project on vercel.com and deploy the application.
Add all your environment variables to the Vercel project by running the commands below. The cli will prompt for the values as input and lets you select if they should be added to development, preview and production. For local development you can overwrite these value with an .env file in the root of your project. It is wise to have one Supabase project for production and one for preview. The preview will then be used in deployment previews on GitHub. You can connect your vercel project with your GitHub repository on the vercel backend.
# the master key for supabase
vercel env add SUPABASE_SERVICE_ROLE_KEY
# the url to your supabase project
vercel env add SUPABASE_URL
# the anon key for supabase
vercel env add SUPABASE_ANON_KEY
# the max rows allowed to fetch from supabase (default 1000)
vercel env add SUPABASE_MAX_ROWSTo let these variables take effect you need to deploy your application once more.
vercel --prodThere are 3 main routes /v3/get, /v3/post and /v3/delete.
On the /get route all actions are controlled by passing URL params. On the /post and /delete route you will have to work with additional POST bodies. For example to fetch a specific tree run the following command.
curl --request GET \
--url 'http://localhost:3000/get/byid&id=_123456789' \
You can see all the available routes in the docs/api.http file with all their needed URLSearchParams and JSON bodies or by inspecting the JSON Schema that is returned when you do a request to the /get, /post or /delete route.
Currently we have these routes
/v3/get |
/v3/post |
/v3/delete |
|---|---|---|
/byid |
/adopt |
/unadopt |
/treesbyids |
/water |
/unwater |
/adopted |
||
/istreeadopted |
||
/wateredandadopted |
||
/lastwatered |
||
/wateredbyuser |
You can sign up with the request below. You will get an access token to use in your requests.
curl --request POST \
--url http://localhost:54321/auth/v1/signup \
--header 'apikey: <SUPABASE ANON KEY>' \
--header 'content-type: application/json' \
--header 'user-agent: vscode-restclient' \
--data '{"email": "[email protected]","password": "1234567890"}'curl --request POST \
--url http://localhost:8080/post/adopt \
--header 'authorization: Bearer <ACCESS_TOKEN>' \
--header 'content-type: application/json' \
--data '{"tree_id":"_01","uuid": "<YOUR USERS ID>"}'
The user id will be removed in future versions since the supabase SDK can get the user id from the access token and each token is bound to a specific user.
Take a look into docs/api.http. The requests in this file can be run with the VSCode extension REST Client.
Locally you will need supabase running and a .env file with the right values in it.
cd giessdenkiez-de-postgres-api
supabase start
# Once the backend is up and running, run the tests
# Make sure you have your .env file setup right
# with all the values from `supabase status`
npm testOn CI the Supabase is started automagically. See .github/workflows/tests.yml
- Run
supabase startto start the supabase stack - make changes to your db using sql and run
supabase db diff --file <MIGRATION FILE NAME> --schema public --use-migrato create migrations - Run
supabase gen types typescript --local > ./_types/database.tsto generate typescript types for your DB.
- Create a project on supabase.com
- Configure your GitHub actions to deploy all migrations to staging and production. See .github/workflows/deploy-to-supabase-staging.yml and .github/workflows/deploy-to-supabase-production.yml for an example. We are using actions environments to deploy to different environments. You can read more about it here: https://docs.github.com/en/actions/reference/environments.
- Needed variables are:
DB_PASSWORDPROJECT_IDSUPABASE_ACCESS_TOKEN
- Needed variables are:
- (Not recommended but possible) Link your local project directly to the remote
supabase link --project-ref <YOUR PROJECT REF>(will ask you for your database password from the creation process) - (Not recommended but possible) Push your local state directly to your remote project
supabase db push(will ask you for your database password from the creation process)
if you want to use the DWD Radolan harvester you need to prepare some data in your database
- Update the table
radolan_harvesterwith a time range for the last 30 days
INSERT INTO "public"."radolan_harvester" ("id", "collection_date", "start_date", "end_date")
VALUES (1, (
SELECT
CURRENT_DATE - INTEGER '1' AS yesterday_date),
(
SELECT
(
SELECT
CURRENT_DATE - INTEGER '31')::timestamp + '00:50:00'),
(
SELECT
(
SELECT
CURRENT_DATE - INTEGER '1')::timestamp + '23:50:00'));- Update the table
radolan_geometrywith sql file radolan_geometry.sql This geometry is Berlin only. - Populate the table radolan_data with the content of radolan_data.sql
This process is actually a little blackbox we need to solve.
There are 3 main routes /get, /post and /delete.
On the /get route all actions are controlled by passing URL params. On the /post and /delete route you will have to work with additional POST bodies. For example to fetch a specific tree run the following command.
curl --request GET \
--url 'http://localhost:8080/get/byid&id=_123456789' \
You can see all the available routes in the docs/api.http file with all their needed URLSearchParams and JSON bodies or by inspecting the JSON Schema that is returned when you do a request to the /get, /post or /delete route.
Currently we have these routes
/v3/get |
/v3/post |
/v3/delete |
|---|---|---|
/byid |
/adopt |
/unadopt |
/treesbyids |
/water |
/unwater |
/adopted |
||
/istreeadopted |
||
/wateredandadopted |
||
/lastwatered |
||
/wateredbyuser |
Some of the requests need a authorized user. You can create a new user using email password via the Supabase API.
curl --request POST \
--url http://localhost:54321/auth/v1/signup \
--header 'apikey: <SUPABASE_ANON_KEY>' \
--header 'content-type: application/json' \
--data '{"email": "[email protected]","password": "1234567890"}'This will give you in development already an aceess token. In production you will need to confirm your email address first.
A login can be done like this:
curl --request POST \
--url 'http://localhost:54321/auth/v1/token?grant_type=password' \
--header 'apikey: <SUPABASE_ANON_KEY>' \
--header 'content-type: application/json' \
--data '{"email": "[email protected]","password": "1234567890"}'See the docs/api.http file for more examples or take a look into the API documentation in your local supabase instance under http://localhost:54323/project/default/api?page=users
Locally you will need supabase running and a .env file with the right values in it.
cd giessdenkiez-de-postgres-api
supabase start
# Once the backend is up and running, run the tests
# Make sure you have your .env file setup right
# with all the values from `supabase status`
npm testOn CI the Supabase is started automagically. See .github/workflows/tests.yml
Thanks goes to these wonderful people (emoji key):
 Fabian Morón Zirfas 💻 📖 |
 Fabian 💻 📖 |
 warenix 💻 📖 |
 Daniel Sippel 📖 |
 Sebastian Meier 💻 |
 Lucas Vogel 📖 |
 Dennis Ostendorf 👀 |
 Julia Zet 👀 |
This project follows the all-contributors specification. Contributions of any kind welcome!
|
A project by:
|
Supported by:
|