julee.domain.repositories.document

Document repository interface defined as Protocol for the Capture, Extract, Assemble, Publish workflow.

This module defines the core document storage and retrieval repository protocol. The repository works with Document domain objects that use BinaryIO streams for efficient content handling.

All repository operations follow the same principles as the sample repositories:

  • Idempotency: All methods are designed to be idempotent and safe for retry. Multiple calls with the same parameters will produce the same result without unintended side effects.

  • Workflow Safety: All operations are safe to call from deterministic workflow contexts. Non-deterministic operations (like ID generation) are explicitly delegated to activities.

  • Domain Objects: Methods accept and return domain objects or primitives, never framework-specific types. Content streams are handled through the BinaryIO interface.

  • Content Streaming: Repository implementations should support both small content (via BytesIO) and large content (via file streams) through the unified ContentStream interface wrapping io.IOBase.

In Temporal workflow contexts, these protocols are implemented by workflow stubs that delegate to activities for durability and proper error handling.

Classes

DocumentRepository

Handles document storage and retrieval operations.

Module Contents

class julee.domain.repositories.document.DocumentRepository[source]

Bases: julee.domain.repositories.base.BaseRepository[julee.domain.models.Document], Protocol

Handles document storage and retrieval operations.

This repository manages the core document storage and metadata operations within the Capture, Extract, Assemble, Publish workflow.

Inherits common CRUD operations (get, save, generate_id) from BaseRepository. The save method handles both content and metadata storage atomically.