Skip to content

entities_reference.md

Latest commit

 

History

History
92 lines (68 loc) · 2.34 KB

entities_reference.md

File metadata and controls

92 lines (68 loc) · 2.34 KB

Entities Reference

This section describes the main data models (Entities) used in the RAGFlow SDK.
Entities represent resources like datasets, documents, chunks, files, chats, agents, and sessions.
They provide convenient accessors, serialization methods, and encapsulate the raw API data.

Common Entity Methods

All entities share several helper methods:

Method Description
.id Returns the unique identifier of the entity.
.to_dict(export_fields=...) Convert entity to a dictionary. Optional export_fields to select specific keys.
.to_json(pretty=False) Serialize entity to JSON string. pretty=True formats the output nicely.
._raw Access the raw API response dictionary for advanced usage.

Examples

Dataset

dataset = await client.datasets.get_dataset(dataset_id="123")

# Access attributes
print(dataset.id, dataset.name)

# Convert to dictionary
dataset_dict = dataset.to_dict(export_fields=["id", "name"])
print(dataset_dict)

# Convert to JSON
dataset_json = dataset.to_json(pretty=True)
print(dataset_json)

# Access raw response
raw_data = dataset._raw

Document

documents, total = await client.documents.list_documents(dataset_id="123")
doc = documents[0]

print(doc.id, doc.name)
print(doc.to_dict())
print(doc.to_json(pretty=True))

Chunk

chunks, count = await client.chunks.list_chunks(dataset_id="123", document_id="doc_456")
chunk = chunks[0]

print(chunk.id, chunk.content)
print(chunk.to_dict(export_fields=["id", "content"]))

ChatCompletionResult

result = await client.chats.ask(chat_id="chat_123", session_id="sess_456", prompt="Hello!")

print(result.text)  # Generated text
print(result.to_dict())
print(result.to_json(pretty=True))

AgentCompletionResult

result = await client.agents.ask(agent_id="agent_123", session_id="sess_456", prompt="Run analysis")

print(result.text)
print(result.to_dict())

Session

sessions, total = await client.chats.list_sessions(chat_id="chat_123")
session = sessions[0]

print(session.id, session.name)
print(session.to_dict())

Notes:

  • Not every entity class is shown, but the usage pattern is the same for all.
  • ._raw is useful for inspecting additional data returned by the server that may not have dedicated attributes.