Install dependencies:
npm install
There are a few environment variables required for the app to function. You can find an overview under .env.template
. Youʼll need access credentials for GitHub, Pinata, Tenderly and a Gelato Relay API key for claiming projects. Youʼll also need to set up PUBLIC_NETWORK
, as described right below. Lastly, you'll need to sign up for a free Coinmarketcap API developer account and populate the COINMARKETCAP_API_KEY
var. Lastly, set PUBLIC_PINATA_GATEWAY_URL
to the URL of the IPFS/Pinata gateway to use. You may use Drips' public IPFS gateway at https://drips.mypinata.cloud
.
The Drips App depends on the custom Drips GraphQL API. You need to set GQL_URL
to a URL of the API's graphql
endpoint, and GQL_ACCESS_TOKEN
for the Authorization: Bearer
token that should be used.
For most development tasks, you may use our Sepolia deployment of the API, hosted at https://drips-api-sepolia-s1.up.railway.app/. To use this API deployment, set GQL_URL
to https://drips-api-sepolia-s1.up.railway.app/
, and GQL_ACCESS_TOKEN
to afdb8b7e-8fa7-4de9-bd95-b650b839e745
. Ensure PUBLIC_NETWORK
is set to 11155111
, so that both the app and API are talking to the Sepolia testnet.
You can also use the local E2E env to easily spin up a fully-fledged deployment of the Drips GraphQL API, complete with event processing and a local, blank testnet to transact against. To do so, set GQL_URL
to http://localhost:8080/graphql
, GQL_ACCESS_TOKEN
to afdb8b7e-8fa7-4de9-bd95-b650b839e745
, and PUBLIC_NETWORK
to 5
. Then, follow the instructions below under "🌐 Run app locally with a local testnet".
To run the app, youʼll need to configure the PUBLIC_NETWORK
environment variable. This should be the chainId of the chain you want to run the app for, and can currently be either 1 or 11155111. The app will only allow connecting wallets that are set to this network, and all server-side requests will be made for this network's subgraph.
For your convenience, weʼve deployed production mirrors of the app set to allow testnet connections:
https://sepolia.drips.network/ # PUBLIC_NETWORK set to 11155111
Before you run the dev server or build the app, you're going to need to build TypeScript types for GraphQL fragments and queries within the app. These auto-generated files, which are all situated in folders called __generated__
adjacent to the .ts
or .svelte
files that define the GQL operation, are deliberately not tracked in Git to ensure the app's GQL types are always up-to-date with the GQL endpoint it is connected to.
To build GraphQL types, ensure you've configured the GraphQL endpoint as described above, and run npm run build:graphql
. This command scans the entire repo for GQL definitions tagged with gql
, and generates typings in adjacent __generated__
folders.
Whenever you make a change to a GraphQL definition, like fragments or queries, ensure to run npm run build:graphql
to make sure all definitions are valid and the typings are up-to-date.
Assuming you've fully configured your local environment and built GraphQL types as described above, you can run a local development server as follows:
npm run dev
# or start the server and open the app in a new browser tab
npm run dev -- --open
To create a production version of the app:
npm run build
You can preview the production build with npm run preview
.
Tests are setup with vitest
as a runner, and playwright
as the E2E test environment. Components can be unit-tested using jsdom
(check $lib/components/example/example.test.ts
for an example). Ensure unit tests follow the name format *.unit.test.ts
, while E2E tests should be named like *.e2e.test.ts
.
With all dependencies installed, simply run:
npm run test:unit
We provide a docker-compose configuration which runs the following services and components locally:
- A Ganache testnet with the Drips Protocol contracts pre-deployed
- Drips Events Processor
- Drips GraphQL API
- Drips Subgraph
- Graph Node
- IPFS Kubo Node (For Graph Node)
- Fake Pinata (Simple local mock of Pinata API)
- 2 Postgres Databases for Events Processor and Subgraph
- One Redis instance for the app's GitHub API caching
You can start and stop this environment with one command, and easily connect a local dev instance of the app to it. This allows entirely gasless local development on a testnet the state of which you can reset at will.
To get started, run:
npm run local-env:start
This command starts the local environment and starts logging all component's output to the console. When you run this for the first time, it'll take a while, but Docker will cache all build steps for subsequent usages.
Stop the local environment by running:
npm run local-env:stop
Docker is configured to always build the Drips Event Processor and Drips GraphQL API images from the latest main
branch commit. When you run the local env, it will automatically check for newer commits and rebuild the images if necessary. If you want to use a different branch of either of these services, you can customize the branch used by specifying environment variables:
GQL_API_BRANCH=main EVENTS_PROCESSOR_BRANCH=main
To start a dev server that has the app talk to the locally-running environment, simply run:
npm run dev:local-env
To run the E2E test suite, ensure the local mock environment is down, and then simply run:
npm run e2e
This command will start the local mock environment, transfer your local source code of the app into a new container, build the app within it, and run tests.
For debugging tests, it's useful to be able to see what's going on within the Playwright browser. To do so, you can run the tests outside of Docker by simply starting the local mock environment (see "Running the mocked local environment"), and execute tests outside of docker by running npm run test:e2e
.
In order to make the Playwright browser headful, add E2E_HEADLESS=0
to your .env
file.
Please note that data is persisted in the local environment unless you explicitly stop and restart it. You will need to do this in-between test runs, because subsequent tests will fail unless the environment is is its initial state.
In order to connect the app to the local E2E network during an E2E test, add this beforeAll
block in your test fixture:
import configureAppForTest from './helpers/configure-app-for-test';
// …
beforeAll(async () => {
server = await preview({ preview: { port: 3000 } });
browser = await chromium.launch();
page = await browser.newPage();
// …
await configureAppForTest(page);
});
This utility sets a window variable isPlaywrightTest
to true
, causing the app to connect to the local testnet and subgraph.
When starting your E2E test like this, user 0xf39fd6e51aad88f6f4ce6ab8827279cfffb92266
is automatically and immediately connected to the app, and has a balance of 100 TEST.
As part of E2E environment setup, a mock ERC-20 token is deployed at 0xefbF81372aBC3723463746a89CEb42080563684C
, and automatically grants address 0x433220a86126eFe2b8C98a723E73eBAd2D0CbaDc
a balance of 100000000000000000000
wei. In order to be able to use this token within the app in your tests, you can append a "custom token" to localstorage, which will be picked up by tokens.store
upon startup.
Simply run the following in a page.addInitScript
block:
localStorage.setItem(
'custom-tokens',
JSON.stringify([
{
source: 'custom',
banned: false,
info: {
chainId: 5,
address: '0xefbF81372aBC3723463746a89CEb42080563684C',
name: 'Testcoin',
decimals: 18,
symbol: 'TEST',
},
},
]),
);
After doing this, the app will display the token within all token pickers. Alternatively, you can also run through adding the custom token via the UI within your test.
Unfortunately, two major differences in app logic for E2E tests couldnʼt be avoided: Firstly, the app uses a mock wallet store that connects to the local testnet instead of the real one, and IPFS access is mocked using localstorage. The logic checks for an env variable PUBLIC_TEST_MODE
being true. The mock wallet store also checks for window.playwrightAddress
, and initializes itself to be connected to that address. In order to make use of these adjustments, call page.addInitScript
and set the two variables.
When a project profile is loaded, a project is added to a Splits editor, or a GitHub repo URL is entered into the Project Claim Flow, a request for the repo's details to the GitHub API is made. To avoid excessive requests, the responses can be cached. To enable caching, set the CACHE_REDIS_CONNECTION_STRING
env variable to a valid Redis connection string.