How bfiliks/comment-rule-labeling works
Project Type and Purpose
• This is a full‐stack, single‐page annotation tool built with Streamlit (Python) that lets multiple annotators label YouTube comments against community rules.
• The “frontend” and “backend” live in the same codebase (app.py) and run on Streamlit Cloud, while Google Sheets serves as the shared, collaborative data store.
File Structure Analysis
• Top level:
• app.py – the monolithic Streamlit application (UI + business logic)
• requirements.txt – Python dependencies (streamlit, gspread, pandas, etc.)
• example_data.csv – sample input data
• .streamlit/config.toml – theming, favicon, and Streamlit settings
• assets/ – static banner image
• No separate “frontend” or “backend” directories; the pattern is a server‐driven UI (Streamlit) with external data integration.
• Configuration & deployment: devcontainer for local dev, GitHub templates, Streamlit secrets for gspread credentials.
Key Architecture Insights from README
• Annotator authentication is name‐based (no heavy IAM).
• Data flow: CSV upload → in‐memory DataFrame → write/read to Google Sheets via gspread API → display/progress tracking.
• Real‐time multi‐annotator sync: uses Google Sheets locking or row‐status checks to avoid duplication.