Data Vector Search

The Data Vector Search API allows you to retrieve processed document chunks from your data stores. See the Data Vector Search API reference and response types for more details.

Get File Chunks

You can retrieve chunks from files in a data store using the get_file_chunks method. This allows you to access processed document chunks with their scores and metadata for analysis or further processing.

Synchronous Usage

from airia import AiriaClient

client = AiriaClient(api_key="your_api_key")

# Get file chunks with default pagination
chunks_response = client.data_vector_search.get_file_chunks(
    data_store_id="your_data_store_id",
    file_id="your_file_id"
)

# Access the chunks
for chunk in chunks_response.chunks:
    print(f"Chunk: {chunk.chunk}")
    print(f"Document: {chunk.document_name}")
    if chunk.score is not None:
        print(f"Score: {chunk.score}")
    if chunk.sequence_number is not None:
        print(f"Sequence: {chunk.sequence_number}")

Asynchronous Usage

from airia import AiriaAsyncClient
import asyncio

async def main():
    client = AiriaAsyncClient(api_key="your_api_key")

    # Get file chunks with default pagination
    chunks_response = await client.data_vector_search.get_file_chunks(
        data_store_id="your_data_store_id",
        file_id="your_file_id"
    )

    # Access the chunks
    for chunk in chunks_response.chunks:
        print(f"Chunk: {chunk.chunk}")
        print(f"Document: {chunk.document_name}")
        if chunk.score is not None:
            print(f"Score: {chunk.score}")
        if chunk.sequence_number is not None:
            print(f"Sequence: {chunk.sequence_number}")

    await client.close()

asyncio.run(main())