dataspace-sdk 0.3.0

Python SDK for DataSpace API

DataSpace Python SDK

A Python SDK for programmatic access to DataSpace resources including Datasets, AI Models, and Use Cases.

Installation

From PyPI (once published)

pip install dataspace-sdk

From Source

git clone https://github.com/CivicDataLab/DataExchange.git
cd DataExchange/DataExBackend
pip install -e .

For Development

pip install -e ".[dev]"

Quick Start

from dataspace_sdk import DataSpaceClient

# Initialize the client
client = DataSpaceClient(base_url="https://api.dataspace.example.com")

# Login with Keycloak token
user_info = client.login(keycloak_token="your_keycloak_token")
print(f"Logged in as: {user_info['user']['username']}")

# Search for datasets
datasets = client.datasets.search(
    query="health data",
    tags=["public-health"],
    page=1,
    page_size=10
)

# Get a specific dataset
dataset = client.datasets.get_by_id("dataset-uuid")
print(f"Dataset: {dataset['title']}")

# Get organization's resources
org_id = user_info['user']['organizations'][0]['id']
org_datasets = client.datasets.get_organization_datasets(org_id)

Features

• Authentication: Login with Keycloak tokens and automatic token refresh
• Datasets: Search, retrieve, and list datasets with filtering and pagination
• AI Models: Search, retrieve, and list AI models with filtering
• Use Cases: Search, retrieve, and list use cases with filtering
• Organization Resources: Get resources specific to your organizations
• GraphQL & REST: Supports both GraphQL and REST API endpoints

Authentication

Login with Keycloak

from dataspace_sdk import DataSpaceClient

client = DataSpaceClient(base_url="https://api.dataspace.example.com")

# Login with Keycloak token
response = client.login(keycloak_token="your_keycloak_token")

# Access user information
print(response['user']['username'])
print(response['user']['organizations'])

Token Refresh

# Refresh access token when it expires
new_token = client.refresh_token()

Check Authentication Status

if client.is_authenticated():
    print("Authenticated!")
    print(f"User: {client.user['username']}")

Working with Datasets

Search Datasets

# Basic search
results = client.datasets.search(query="education")

# Advanced search with filters
results = client.datasets.search(
    query="health",
    tags=["public-health", "covid-19"],
    sectors=["Health"],
    geographies=["India", "Karnataka"],
    status="PUBLISHED",
    access_type="OPEN",
    sort="recent",
    page=1,
    page_size=20
)

# Access results
print(f"Total results: {results['total']}")
for dataset in results['results']:
    print(f"- {dataset['title']}")

Get Dataset by ID

# Get detailed dataset information
dataset = client.datasets.get_by_id("550e8400-e29b-41d4-a716-446655440000")

print(f"Title: {dataset['title']}")
print(f"Description: {dataset['description']}")
print(f"Organization: {dataset['organization']['name']}")
print(f"Resources: {len(dataset['resources'])}")

List All Datasets

# List with pagination
datasets = client.datasets.list_all(
    status="PUBLISHED",
    limit=50,
    offset=0
)

for dataset in datasets:
    print(f"- {dataset['title']}")

Get Trending Datasets

trending = client.datasets.get_trending(limit=10)
for dataset in trending['results']:
    print(f"- {dataset['title']} (views: {dataset['view_count']})")

Get Organization Datasets

# Get datasets for your organization
org_id = client.user['organizations'][0]['id']
org_datasets = client.datasets.get_organization_datasets(
    organization_id=org_id,
    limit=20,
    offset=0
)

Working with AI Models

Search AI Models

# Basic search
results = client.aimodels.search(query="language model")

# Advanced search
results = client.aimodels.search(
    query="llm",
    tags=["nlp", "text-generation"],
    model_type="LLM",
    provider="OPENAI",
    status="ACTIVE",
    sort="recent",
    page=1,
    page_size=10
)

Get AI Model by ID

# Using REST endpoint
model = client.aimodels.get_by_id("model-uuid")

# Using GraphQL (more detailed)
model = client.aimodels.get_by_id_graphql("model-uuid")

print(f"Model: {model['displayName']}")
print(f"Type: {model['modelType']}")
print(f"Provider: {model['provider']}")
print(f"Endpoints: {len(model['endpoints'])}")

List All AI Models

models = client.aimodels.list_all(
    status="ACTIVE",
    model_type="LLM",
    limit=20,
    offset=0
)

Get Organization AI Models

org_id = client.user['organizations'][0]['id']
org_models = client.aimodels.get_organization_models(
    organization_id=org_id,
    limit=20,
    offset=0
)

Working with Use Cases

Search Use Cases

# Basic search
results = client.usecases.search(query="health monitoring")

# Advanced search
results = client.usecases.search(
    query="covid",
    tags=["health", "monitoring"],
    sectors=["Health"],
    status="PUBLISHED",
    running_status="COMPLETED",
    sort="completed_on",
    page=1,
    page_size=10
)

Get Use Case by ID

usecase = client.usecases.get_by_id(123)

print(f"Title: {usecase['title']}")
print(f"Summary: {usecase['summary']}")
print(f"Status: {usecase['runningStatus']}")
print(f"Datasets used: {len(usecase['datasets'])}")
print(f"Organizations: {len(usecase['organizations'])}")

List All Use Cases

usecases = client.usecases.list_all(
    status="PUBLISHED",
    running_status="COMPLETED",
    limit=20,
    offset=0
)

Get Organization Use Cases

org_id = client.user['organizations'][0]['id']
org_usecases = client.usecases.get_organization_usecases(
    organization_id=org_id,
    limit=20,
    offset=0
)

Error Handling

from dataspace_sdk import (
    DataSpaceAPIError,
    DataSpaceAuthError,
    DataSpaceNotFoundError,
    DataSpaceValidationError,
)

try:
    dataset = client.datasets.get_by_id("invalid-uuid")
except DataSpaceNotFoundError as e:
    print(f"Dataset not found: {e.message}")
except DataSpaceAuthError as e:
    print(f"Authentication error: {e.message}")
    # Try to refresh token
    client.refresh_token()
except DataSpaceValidationError as e:
    print(f"Validation error: {e.message}")
    print(f"Details: {e.response}")
except DataSpaceAPIError as e:
    print(f"API error: {e.message}")
    print(f"Status code: {e.status_code}")

Advanced Usage

Pagination

# Manual pagination
page = 1
page_size = 20
all_datasets = []

while True:
    results = client.datasets.search(
        query="health",
        page=page,
        page_size=page_size
    )

    all_datasets.extend(results['results'])

    if len(results['results']) < page_size:
        break

    page += 1

print(f"Total datasets fetched: {len(all_datasets)}")

Working with Multiple Organizations

# Get user's organizations
user_info = client.get_user_info()

for org in user_info['organizations']:
    print(f"\nOrganization: {org['name']} (Role: {org['role']})")

    # Get resources for each organization
    datasets = client.datasets.get_organization_datasets(org['id'])
    models = client.aimodels.get_organization_models(org['id'])
    usecases = client.usecases.get_organization_usecases(org['id'])

    print(f"  Datasets: {len(datasets)}")
    print(f"  AI Models: {len(models)}")
    print(f"  Use Cases: {len(usecases)}")

Combining Search Results

# Search across all resource types
query = "health"

datasets = client.datasets.search(query=query, page_size=5)
models = client.aimodels.search(query=query, page_size=5)
usecases = client.usecases.search(query=query, page_size=5)

print(f"Found {datasets['total']} datasets")
print(f"Found {models['total']} AI models")
print(f"Found {usecases['total']} use cases")

API Reference

DataSpaceClient

Main client for interacting with DataSpace API.

Methods:

• login(keycloak_token: str) -> dict: Login with Keycloak token
• refresh_token() -> str: Refresh access token
• get_user_info() -> dict: Get current user information
• is_authenticated() -> bool: Check authentication status

Properties:

• datasets: DatasetClient instance
• aimodels: AIModelClient instance
• usecases: UseCaseClient instance
• user: Current user information
• access_token: Current access token

DatasetClient

Client for dataset operations.

Methods:

• search(...): Search datasets with filters
• get_by_id(dataset_id: str): Get dataset by UUID
• list_all(...): List all datasets with pagination
• get_trending(limit: int): Get trending datasets
• get_organization_datasets(organization_id: str, ...): Get organization's datasets

AIModelClient

Client for AI model operations.

Methods:

• search(...): Search AI models with filters
• get_by_id(model_id: str): Get AI model by UUID (REST)
• get_by_id_graphql(model_id: str): Get AI model by UUID (GraphQL)
• list_all(...): List all AI models with pagination
• get_organization_models(organization_id: str, ...): Get organization's AI models

UseCaseClient

Client for use case operations.

Methods:

• search(...): Search use cases with filters
• get_by_id(usecase_id: int): Get use case by ID
• list_all(...): List all use cases with pagination
• get_organization_usecases(organization_id: str, ...): Get organization's use cases

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

AGPL-3.0 License

Support

For issues and questions, please open an issue on GitHub.