This is a FastAPI application that acts as an API for a natural language chatbot that interacts with a Paperless-ngx instance. The API parses user queries to extract relevant information (like document types, dates, and names), searches for documents in Paperless-ngx, and returns a formatted, user-friendly response with links to the found documents.
-
Natural Language Processing: Uses
spaCyto parse German text queries and identify key entities. -
Intelligent Search: Extracts document types, tags, contributors, and dates from queries to build a comprehensive search request for Paperless-ngx.
-
Dynamic Pagination: Fetches all matching documents from the Paperless-ngx API, handling pagination automatically.
-
User-Friendly Output: Formats the search results into a clean, readable message with direct links to the documents in the Paperless-ngx web interface.
-
Robust Error Handling: Includes basic error handling for API requests and JSON parsing.
Before running the application, ensure you have the following installed:
-
Python 3.7+
-
An active Paperless-ngx instance.