pharia-data-sdk

Formerly the intelligence_layer/connectors/data, intelligence_layer/connectors/document_index and intelligence_layer/connectors/retrievers modules.

Overview

This module provides connectors for interacting with the Pharia Data Platform and Pharia Search (a.k.a. Document Index), you can use it to semantically search, access and manage data, documents.

Installation

pip install pharia-data-sdk

Usage

Data Platform Connector

from pharia_data_sdk.connectors.data import DataClient

client = DataClient(token="<token>", base_url="<base_data_platform_url>")

repositories = client.list_repositories()
repository = repositories[0]
datasets = client.list_datasets(repository.repository_id)
dataset = datasets[0]

Document Index Connector

from pharia_data_sdk.connectors.document_index.document_index import DocumentIndexClient, SearchQuery

client = DocumentIndexClient(token="<token>", base_url="<base_document_index_url>")

namespaces = client.list_namespaces()
collections = client.list_collections(namespaces[0])
indexes = client.list_indexes(namespaces[0])
client.search(collections[0], indexes[0].index, SearchQuery(query="What fish is most common in swedish lakes?"))

Retrievers

from pharia_data_sdk.connectors.retrievers.document_index_retriever import DocumentIndexRetriever
from pharia_data_sdk.connectors.document_index.document_index import DocumentIndexClient

retriever = DocumentIndexRetriever(
    document_index=DocumentIndexClient(token="<token>", base_url="<base_document_index_url>"),
    index_name="<index_name>",
    namespace="<namespace>",
    collection="<collection>",
)

retriever.get_relevant_documents_with_scores("What fish is most common in swedish lakes?")

Contributing

We welcome contributions! Please see our Contributing Guide for details on how to set up the development environment and submit changes.