---
title: Atlas CSV Maker
emoji: "📄"
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: "4.44.0"
python_version: "3.11"
app_file: app.py
pinned: false
---

# Semantic Chunking CSV Builder

This Gradio app runs in Hugging Face Spaces and converts uploaded files into a CSV containing the filename, page number, and semantic text chunks.

## Features
- Upload individual files or an entire folder of files
- Extract machine-readable text from PDF, DOCX, TXT, and common image formats
- Chunk text using the Chonkie `SemanticChunker`
- Download a CSV with `filename`, `page_number`, and `chunk_text`

## Supported formats
- PDFs (`.pdf`)
- Word documents (`.docx`)
- Text files (`.txt`)
- Images (`.jpg`, `.jpeg`, `.png`, `.tif`, `.tiff`, `.bmp`, `.gif`)

## Notes for OCR
Image extraction relies on Tesseract. In Hugging Face Spaces, add the following to `apt.txt`:
- `tesseract-ocr`

## Local usage
1. Install dependencies from `requirements.txt`.
2. Run the app:

```bash
python app.py
```

Then open the local Gradio URL shown in the terminal.

## Tests
Install `requirements-dev.txt` and run pytest to execute the smoke test.