Quickstart: Marketplace Flow

This batteries-included flow includes metadata, multiple services for one datatoken, and compute-to-data.

It focuses on Alice's experience as a publisher, and Bob's experience as a buyer & consumer. The rest are services used by Alice and Bob.

Here's the steps.

Initialize services
Alice publishes assets for data services (= publishes a datatoken contract and metadata)
Alice mints 100 tokens
Alice allows marketplace to sell her datatokens
Marketplace posts asset for sale
Value swap: Bob buys datatokens from marketplace
Bob uses a service he just purchased (download)

Let's go through each step.

0. Installation

If you haven't installed yet:

pip install ocean-lib

1. Initialize services

This quickstart treats the publisher service, metadata store, and marketplace as externally-run services. For convenience, we run them locally in default settings.

docker run @oceanprotocol/provider-py:latest
docker run @oceanprotocol/metadatastore:latest
docker run @oceanprotocol/marketplace:latest

2. Alice publishes assets for data services (= publishes a datatoken contract)

from ocean_lib import Ocean
from ocean_lib.web3_internal.utils import get_account

#Alice's config
config = {
   'network' : 'rinkeby',
   'privateKey' :'8da4ef21b864d2cc526dbdb2a120bd2874c36c9d0a1fb7f8c63d7f7a8b41de8f',
   'metadataStoreUri' : 'localhost:5000',
   'providerUri' : 'localhost:8030'
}
ocean = Ocean(config)
account = get_account(0)

data_token = ocean.create_data_token(ocean.config.metadata_store_url, account)
token_address = data_token.address

metadata =  {"main": {
    "type": "dataset", "name": "10 Monkey Species Small", "author": "Mario", 
    "license": "CC0: Public Domain", "dateCreated": "2012-02-01T10:55:11Z", 
    "files": [{ "index": 0, "contentType": "application/zip", "url": "https://s3.amazonaws.com/datacommons-seeding-us-east/10_Monkey_Species_Small/assets/training.zip"},
              { "index": 1, "contentType": "text/text", "url": "https://s3.amazonaws.com/datacommons-seeding-us-east/10_Monkey_Species_Small/assets/monkey_labels.txt"},
              { "index": 2, "contentType": "application/zip", "url": "https://s3.amazonaws.com/datacommons-seeding-us-east/10_Monkey_Species_Small/assets/validation.zip"}]}}
asset = ocean.assets.create(metadata, account, data_token_address=token_address)
assert token_address == asset._other_values['dataTokenAddress']

did = asset.did

3. Alice mints 100 tokens

data_token.mint(account.address, 100, account)

4. Alice allows marketplace to sell her datatokens

marketplace_address = '0x068ed00cf0441e4829d9784fcbe7b9e26d4bd8d0'
data_token.approve(marketplace_address, 20)

5. Marketplace posts asset for sale

Now, you're the marketplace:)

from ocean_lib import Ocean

#Market's config
config = {
   'network': 'rinkeby',
   'privateKey':'1234ef21b864d2cc526dbdb2a120bd2874c36c9d0a1fb7f8c63d7f7a8b41de8f',
}
market_ocean = Ocean(config)

asset = ocean.assets.resolve(did)
service1 = asset.get_service('download')
service2 = asset.get_service('access')
price = 10.0 #marketplace-set price of 10 USD / datatoken

#Display key asset information, such as the cost of each service
print(f"Service 1 costs {service1.get_num_dt_needed() * price} USD") # 1.5 * 10 = 15
print(f"Service 2 costs {service2.get_num_dt_needed() * price} USD") # 2.5 * 10 = 25

6. Value swap: Bob buys datatokens from marketplace

#Not shown: in marketplace GUI, Bob uses Stripe to send USD to marketplace (or other methods / currencies).

market_token = market_ocean.get_data_token(token_address)
market_token.transfer(dst_address=bob_address, 1.0)

7. Bob uses a service he just purchased (download)

Now, you're Bob:)


#Bob's config
config = {
   'network': 'rinkeby',
   'privateKey':'1234ef21b864d2cc526dbdb2a120bd2874c36c9d0a1fb7f8c63d7f7a8b41de8o',
}
market_ocean = Ocean(config)

service = asset.get_service('access')
file_path = bob_ocean.assets.download(asset.did, service.index, bob_account, '~/my-datasets')

4.0 KiB Raw Blame History