This project implements a simple text annotation application using FastHTML, showcasing dynamic updates, database integration, and a user-friendly interface for annotating text data.

Contributed by: @altryne

• View and navigate through text samples
• Annotate texts as correct or incorrect
• Add notes to each text sample
• Real-time updates without page reloads
• SQLite database integration
• Automatic loading of dummy data

• FastHTML: A Python framework for building dynamic web applications
• HTMX: For seamless client-side updates without full page reloads
• SQLite: For persistent data storage
• DaisyUI and Tailwind CSS: For styling and UI components

The app is set up using FastHTML's fast_app function, which creates the application, routing, database connection, and data model in a single line:

app, rt, texts_db, Item = fast_app('texts.db', hdrs=(tlink, dlink, picolink, MarkdownJS(), HighlightJS()), live=True, id=int, messages=list, feedback=bool, notes=str, pk='id', render=render)

The app comes with a dummy data file dummy_data.jsonl containing sample text data. This file is loaded automatically when the app starts. If you want to load your own data, you'll need to replace this file with a new one in the same format.

The dummy data file contains JSON lines (JSONL) with the following structure on each line:

[{"role":"user","content":"User's message"},{"role":"assistant","content":"Assistant's response"}]

To run the app, execute the following command:

python main.py

Open your web browser and navigate to http://localhost:5001 to access the app.

To use your own data instead of the dummy data:

1. Create a JSONL file with your data, following the structure described above
2. Replace the dummy_data.jsonl file in the data directory with your file
3. Ensure your file is named dummy_data.jsonl or update the file path in main.py