Table of Contents
- 🏄 Get Started
- 🦑 Environment variables
- 🦀 Data Sources
- 🎨 Storybook
- ✨ Code Style
- 👩🔬 Testing
- 🛳 Production
- ⬆️ Deployment
- 💖 Contributing
- 🍴 Forking
- 💻 Advanced Features
- ✅ GDPR Compliance
- 🏛 License
The app is a React app built with Gatsby.js + TypeScript + CSS modules and will connect to Ocean remote components by default.
To start local development:
git clone git@github.com:oceanprotocol/market.git
cd market
# when using nvm to manage Node.js versions
nvm use
npm install
npm start
This will start the development server under
http://localhost:8000
.
To explore the generated GraphQL data structure fire up the accompanying GraphiQL IDE under
http://localhost:8000/__graphql
.
If you prefer to connect to locally running components instead of remote connections, you can spin up barge
and use a local Ganache network in another terminal before running npm start
:
git clone git@github.com:oceanprotocol/barge.git
cd barge
# startup with local Ganache node
./start_ocean.sh
Barge will deploy contracts to the local Ganache node which will take some time. At the end the compiled artifacts need to be copied over to this project into node_modules/@oceanprotocol/contracts/artifacts
. This script will do that for you:
./scripts/copy-contracts.sh
Finally, set environment variables to use this local connection in .env
in the app:
# modify env variables
cp .env.example .env
npm start
To use the app together with MetaMask, importing one of the accounts auto-generated by the Ganache container is the easiest way to have test ETH available. All of them have 100 ETH by default. Upon start, the ocean_ganache_1
container will print out the private keys of multiple accounts in its logs. Pick one of them and import into MetaMask.
To fully test all The Graph integrations, you have to run your own local Graph node with our ocean-subgraph
deployed to it. Barge does not include a local subgraph so by default, the subgraphUri
is hardcoded to the Rinkeby subgraph in our getDevelopmentConfig
function.
Cleaning all Docker images so they are fetched freshly is often a good idea to make sure no issues are caused by old or stale images:
docker system prune --all --volumes
The app.config.js
file is setup to prioritize environment variables for setting each Ocean component endpoint. By setting environment variables, you can easily switch between Ocean networks the app connects to, without directly modifying app.config.js
.
For local development, you can use a .env
file:
# modify env variables, Rinkeby is enabled by default when using those files
cp .env.example .env
All displayed data in the app is presented around the concept of one data set, which is a combination of:
- metadata about a data set
- the actual data set files
- the datatoken which represents the data set
- financial data connected to this datatoken, either a pool or a fixed rate exchange contract
- calculations and conversions based on financial data
- metadata about publishers
All this data then comes from multiple sources:
All initial data sets and their metadata (DDO) is retrieved client-side on run-time from the Aquarius instance, defined in app.config.js
. All app calls to Aquarius are done with 2 internal methods which mimic the same methods in ocean.js, but allow us:
- to cancel requests when components get unmounted in combination with axios
- hit Aquarius as early as possible without relying on any ocean.js initialization
Aquarius runs Elasticsearch under the hood so its stored metadata can be queried with Elasticsearch queries like so:
import { QueryResult } from '@oceanprotocol/lib/dist/node/metadatacache/MetadataCache'
import { queryMetadata } from '../../utils/aquarius'
const queryLatest = {
query: {
// https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-query-string-query.html
query_string: { query: `-isInPurgatory:true` }
},
sort: { created: 'desc' }
}
function Component() {
const { appConfig } = useSiteMetadata()
const [result, setResult] = useState<QueryResult>()
useEffect(() => {
if (!appConfig.metadataCacheUri) return
const source = axios.CancelToken.source()
async function init() {
const result = await queryMetadata(query, source.token)
setResult(result)
}
init()
return () => {
source.cancel()
}
}, [appConfig.metadataCacheUri, query])
return <div>{result}</div>
}
For components within a single data set view the useAsset()
hook can be used, which in the background gets the respective metadata from Aquarius.
import { useAsset } from '../../../providers/Asset'
function Component() {
const { ddo } = useAsset()
return <div>{ddo}</div>
}
Most financial data in the market is retrieved with GraphQL from our own subgraph, rendered on top of the initial data coming from Aquarius.
The app has Urql Client setup to query the respective subgraph based on network. In any component this client can be used like so:
import { gql, useQuery } from 'urql'
const query = gql`
query PoolLiquidity($id: ID!, $shareId: ID) {
pool(id: $id) {
id
totalShares
}
}
`
function Component() {
const { data } = useQuery(query, {}, pollInterval: 5000 })
return <div>{data}</div>
}
Publishers can create a profile on 3Box Hub and when found, it will be displayed in the app.
For this our own 3box-proxy is used, within the app the utility method get3BoxProfile()
can be used to get all info:
import get3BoxProfile from '../../../utils/profile'
function Component() {
const [profile, setProfile] = useState<Profile>()
useEffect(() => {
if (!account) return
const source = axios.CancelToken.source()
async function get3Box() {
const profile = await get3BoxProfile(account, source.token)
if (!profile) return
setProfile(profile)
}
get3Box()
return () => {
source.cancel()
}
}, [account])
return (
<div>
{profile.emoji} {profile.name}
</div>
)
}
Based on list-purgatory some data sets get additional data. Within most components this can be done with the internal useAsset()
hook which fetches data from the market-purgatory endpoint in the background.
For asset purgatory:
import { useAsset } from '../../../providers/Asset'
function Component() {
const { isInPurgatory, purgatoryData } = useAsset()
return isInPurgatory ? <div>{purgatoryData.reason}</div> : null
}
For account purgatory:
import { useWeb3 } from '../../../providers/Web3'
import { useAccountPurgatory } from '../../../hooks/useAccountPurgatory'
function Component() {
const { accountId } = useWeb3()
const { isInPurgatory, purgatoryData } = useAccountPurgatory(accountId)
return isInPurgatory ? <div>{purgatoryData.reason}</div> : null
}
All displayed chain & network metadata is retrieved from https://chainid.network
on build time and integrated into Gatsby's GraphQL layer. This data source is a community-maintained GitHub repository under ethereum-lists/chains.
Within components this metadata can be queried for under allNetworksMetadataJson
. The useWeb3()
hook does this in the background to expose the final networkDisplayName
for use in components:
export default function NetworkName(): ReactElement {
const { networkId, isTestnet } = useWeb3()
const { networksList } = useNetworkMetadata()
const networkData = getNetworkDataById(networksList, networkId)
const networkName = getNetworkDisplayName(networkData, networkId)
return (
<>
{networkName} {isTestnet && `(Test)`}
</>
)
}
TODO: this is broken for most components. See oceanprotocol#128
Storybook is set up for this project and is used for UI development of components. Stories are created inside src/components/
alongside each component in the form of ComponentName.stories.tsx
.
To run the Storybook server, execute in your Terminal:
npm run storybook
This will launch the Storybook UI with all stories loaded under localhost:4000.
Code style is automatically enforced through ESLint & Prettier rules:
- Git pre-commit hook runs
prettier
on staged files, setup with Husky - VS Code suggested extensions and settings for auto-formatting on file save
- CI runs a linting & TypeScript typings check with
npm run lint
, and fails if errors are found
For running linting and auto-formatting manually, you can use from the root of the project:
# linting check, also runs Typescript typings check
npm run lint
# auto format all files in the project with prettier, taking all configs into account
npm run format
TODO: this is broken and never runs in CI. See oceanprotocol#128
Test suite for unit tests is setup with Jest as a test runner and:
- react-testing-library for all React components
To run all linting and unit tests:
npm test
For local development, you can start the test runner in a watch mode.
npm run test:watch
For analyzing the generated JavaScript bundle sizes you can use:
npm run analyze
To create a production build, run from the root of the project:
npm run build
# serve production build
npm run serve
Every branch or Pull Request is automatically deployed to multiple hosts for redundancy and emergency reasons:
A link to a preview deployment will appear under each Pull Request.
The latest deployment of the main
branch is automatically aliased to market.oceanprotocol.com
, where the deployment on Netlify is the current live deployment.
We welcome contributions in form of bug reports, feature requests, code changes, or documentation improvements. Have a look at our contribution documentation for instructions and workflows:
We encourage you to fork this repository and create your own data marketplace. When you publish your forked version of this market there are a few elements that you are required to change for copyright reasons:
- The typeface is copyright protected and needs to be changed unless you purchase a license for it.
- The Ocean Protocol logo is a trademark of the Ocean Protocol Foundation and must be removed from forked versions of the market.
- The name "Ocean Market" is also copyright protected and should be changed to the name of your market.
Additionally, we would also advise that your retain the text saying "Powered by Ocean Protocol" on your forked version of the marketplace in order to give credit for the development work done by the Ocean Protocol team.
Everything else is made open according to the apache2 license. We look forward to seeing your data marketplace!
Ocean Market also includes a number of advanced features that are suitable for an enterprise data market, such as:
- Role based access control
- Allow and deny lists
- Free pricing
See our seperate guide on advanced features
Ocean Market comes with prebuilt components for you to customize to cover GDPR requirements. Find additional information on how to use them below.
Feel free to adopt our provided privacy policies to your needs. Per default we cover four different languages: English, German, Spanish and French. Please be advised, that you will need to adjust some paragraphs in the policies depending on your market setup (e.g. the use of cookies). You can easily add or remove policies by providing your own markdown files in the content/pages/privacy
directory. For guidelines on how to format your markdown files refer to our provided policies. The pre-linked content tables for these files are automatically generated.
Additionally, Ocean Market provides a privacy preference center for you to use. This feature is disabled per default since we do not use cookies requiring consent on our deployment of the market. However, if you need to add some functionality depending on cookies, you can simply enable this feature by changing the value of the GATSBY_PRIVACY_PREFERENCE_CENTER
environmental variable to "true"
in your .env
file. This will enable a customizable cookie banner stating the use of your individual cookies. The content of this banner can be adjusted within the content/gdpr.json
file. If no optionalCookies
are provided, the privacy preference center will be set to a simpler version displaying only the title
, text
and close
-button. This can be used to inform the user about the use of essential cookies, where no consent is needed. The privacy preference center supports two different styling options: 'small'
and 'default'
. Setting the style propertie to 'small'
will display a smaller cookie banner to the user at first, only showing the default styled privacy preference center upon the user's customization request.
Now your market users will be provided with additional options to toggle the use of your configured cookie consent categories. You can always retrieve the current consent status per category with the provided useConsent()
hook. See below, how you can set your own custom cookies depending on the market user's consent. Feel free to adjust the provided utility functions for cookie usage provided in the src/utils/cookies.ts
file to your needs.
import { CookieConsentStatus, useConsent } from '../../providers/CookieConsent'
import { deleteCookie, setCookie } from '../../utils/cookies'
// ...
const { cookies, cookieConsentStatus } = useConsent()
cookies.map((cookie) => {
const consent = cookieConsentStatus[cookie.cookieName]
switch (consent) {
case CookieConsentStatus.APPROVED:
// example logic
setCookie(`YOUR_COOKIE_NAME`, 'VALUE')
break
case CookieConsentStatus.REJECTED:
case CookieConsentStatus.NOT_AVAILABLE:
default:
// example logic
deleteCookie(`YOUR_COOKIE_NAME`)
break
}
})
The privacy preference centre has two styling options default
and small
. The default view shows all of the customization options on a full-height side banner. When the small
setting is used, a much smaller banner is shown which only reveals all of the customization options when the user clicks "Customize".
The style can be changed by altering the style
prop in the PrivacyPreferenceCenter
component in src/components/App.tsx
. For example:
<PrivacyPreferenceCenter style="small" />
Copyright 2021 Ocean Protocol Foundation Ltd.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.