osdupy 0.4.0

A simple python client for interacting with the OSDU API.

osdupy

A simple python client for the OSDU data platform.

Contents

• Clients
  • SimpleOsduClient
  • AwsServicePrincipalOsduClient
  • AwsOsduClient
• Currently supported methods
• Installation
• Tests
• Usage
  • Instantiating the SimpleOsduClient
  • Instantiating the AwsServicePrincipalOsduClient
  • Instantiating the AwsOsduClient
  • Using the client
    • Search for records by query
    • Search with paging
    • Get a record
    • Upsert records
    • List groupmembership for the current user
    • List membership of a particular group
    • Add a user to a particular group
• Release Notes

Clients

Choose the client that best meets your needs. The same methods are all supported for each.

SimpleOsduClient

BYOT: Bring your own token. Great for backend service or business logic that supplements a front-end application.

This client assumes you are obtaining a token yourself (e.g. via your application's login form or otheer mechanism. With this SimpleOsduClient, you simply provide that token. With this simplicity, you are also then respnsible for reefreeshing the token as needed and re-instantiating the client with the new token.

AwsOsduServicePrincipalClient

Requires: boto3==1.15.*

Good for batch tasks that don't have an interactive front-end. Token management is handled with the boto3 library directly through the Cognito service. You have to supply additional arguments for this.

For OSDU on AWS, this client is usually simpler than the AwsOsduClient as long as you have IAM credentials to access the necessary resources. You only need to provide the OSDU resource_prefix, region, and profile.

AwsOsduClient

Requires: boto3==1.15.*

Good for batch tasks that don't have an interactive front-end. Token management is handled with the boto3 library directly through the Cognito service. You have to supply additional arguments for this.

For OSDU on AWS, this client is useful in the case where you may want to perform actions as a specific OSDU user rather than as the ServicePrinicpal.

Currently supported methods

• search
  • query
  • query_with_paging
• storage
  • query_all_kinds
  • get_record
  • get_records
  • get_all_record_versions
  • get_record_version
  • store_records
  • delete_record
  • purge_record
• dataset
  • get_dataset_registry
  • get_dataset_registries
  • get_storage_instructions
  • register_dataset
  • get_retrieval_instructions
• entitlements
  • get_groups
  • get_group_members
  • add_group_member
  • delete_group_member
  • create_group

Installation

pip install osdupy

Tests

Run unit tests

python -m unittest -v tests.unit

Run integration tests

python -m unittest -v tests.integration

Usage

Instantiating the SimpleOsduClient

If environment variable OSDU_API_URL is set, then it does not need to be passed as an argument. Otherwise it must be passed as keyword argument.

from osdu.client import SimpleOsduClient

data_partition = 'osdu'
token = 'token-received-from-front-end-app'

# With env var `OSDU_API_URL` set in current environment.
osdu_client = SimpleOsduClient(data_partition, token)

# Without env var set.
api_url = 'https://your.api.base_url.com'
osdu_client = SimpleOsduClient(data_partition, token, api_url=api_url)

Instantiating the AwsServicePrincipalOsduClient

from osdu.client import AwsOsduClient

data_partition = 'osdu'
resource_prefix = 'osdur3mX'

osdu_client = AwsServicePrincipalOsduClient(
    data_partition,
    resource_prefix,
    profile=os.environ['AWS_PROFILE'],
    region=os.environ['AWS_DEFAULT_REGION']
)

Instantiating the AwsOsduClient

The only required argument is data_partition. If your environment variables (below) have been set, then client can be instantiated with only data_partition as an argument. Environment variables:

1. OSDU_API_URL
2. OSDU_CLIENT_ID
3. OSDU_USER
4. OSDU_PASSWORD
5. AWS_PROFILE
6. AWS_SECRETHASH

from osdu.client import AwsOsduClient

data_partition = 'osdu'

osdu_client = AwsOsduClient(data_partition)

If you have not set the above environment variales—or you have only set some—then you will need to pass any undefined as args when instantiating the client.

from getpass import getpass
from osdu.client import AwsOsduClient

api_url = 'https://your.api.url.com'  # Must be base URL only
client_id = 'YOURCLIENTID'
user = '[e-mail]'
password = getpass()
data_partition = 'osdu'
profile = 'osdu-dev'

message = user + client_id
dig = hmac.new(client_secret.encode('UTF-8'), msg=message.encode('UTF-8'),
               digestmod=hashlib.sha256).digest()
secretHash = base64.b64encode(dig).decode()



osdu_client = AwsOsduClient(data_partition,
    api_url=api_url,
    client_id=client_id,
    user=user,
    password=password,
    secret_hash=secretHash,
    profile=profile)

Using the client

Below are just a few usage examples. See integration tests for more comprehensive usage examples.

Search for records by query

query = {
    "kind": f"osdu:wks:*:*"
}
result = osdu_client.search.query(query)
# { results: [ {...}, .... ], totalCount: ##### }

Search with paging

For result sets larger than 1,000 records, use the query with paging method.

page_size = 100 # Number of records per page (1-1000)
query = {
    "kind": f"osdu:wks:*:*",
    "limit": page_size
}
result = osdu_client.search.query_with_paging(query)

# Iterate over the pages to do something with the results.
for page, total_count in result:
    for record in page:
        # Do stuff with record...

Get a record

record_id = 'opendes:doc:123456789'
result = osdu_client.storage.get_record(record_id)
# { 'id': 'opendes:doc:123456789', 'kind': ..., 'data': {...}, 'acl': {...}, .... }

Upsert records

new_or_updated_record = './record-123.json'
with open(new_or_updated_record, 'r') as _file:
    record = json.load(_file)

result = osdu_client.storage.store_records([record])

List groupmembership for the current user

result = osdu_client.entitlements.get_groups()
# {
#  "desId": "[e-mail]",
#  "groups": [
#    {
#      "description": "Datalake Plugin-Manager users",
#      "email": "[e-mail]",
#      "name": "service.plugin.user"
#    },
#    {
#      "description": "Datalake csv-parser admins",
#      "email": "[e-mail]",
#      "name": "service.csv-parser.admin"
#    },
#    #...
#       {
#     "description": "The viewer of the datalake csv-parser service",
#     "email": "[e-mail]",
#     "name": "service.csv-parser.viewer"
#   }
# ],
# "memberEmail": "[e-mail]"
# }

List membership of a particular group

result = osdu_client.entitlements.get_group_members('[e-mail]')
#{
#  "members": [
#    {
#      "email": "[e-mail]",
#      "role": "OWNER"
#    },
#    {
#      "email": "[e-mail]",
#      "role": "OWNER"
#    },
#    {
#      "email": "[e-mail]",
#      "role": "OWNER"
#    }
#  ]
#}

Add a user to a particular group

Add a user ([e-mail]) to groups to give entitlement to search for and retrieve data.

query =  {
     "email": "[e-mail]",
     #OWNER or MEMBER
     "role": "MEMBER",
 }
result = osdu_client.entitlements.add_group_member('[e-mail]',query)
query =  {
     "email": "[e-mail]",
     #OWNER or MEMBER
     "role": "OWNER",
 }
result = osdu_client.entitlements.add_group_member('[e-mail]',query)

Delete user from a particular group

Remove a user ([e-mail]) from a group.

query =  {
     "email": "[e-mail]",
     #OWNER or MEMBER
     "role": "MEMBER",
 }
result = osdu_client.entitlements.delete_group_member('[e-mail]',query)
query =  {
     "email": "[e-mail]",
     #OWNER or MEMBER
     "role": "OWNER",
 }
result = osdu_client.entitlements.delete_group_member('[e-mail]',query)