How qiuyannnn/local-file-organizer works
Project type and purpose
• This is an open-source, local CLI tool (“Local File Organizer”) written in Python.
• Purpose: scan a user-specified directory, analyze file contents (text and images) via on-device AI (Nexa SDK LLM and VLM models), generate metadata (folder names, filenames), and reorganize files into a new directory tree—all without sending data off-device.
File structure
Top level:
• main.py ← CLI entry point / orchestrator
• data_processing_common.py ← shared helper functions (prompt templates, metadata structures)
• text_data_processing.py ← handles text/document analysis via NexaTextInference
• image_data_processing.py ← handles image analysis via NexaVLMInference
• file_utils.py ← scanning folders, creating target dirs, moving/renaming files
• output_filter.py ← optional post-processing or log filtering
• requirements.txt ← Python dependencies (nexaai, Pillow, pytesseract, PyMuPDF, python-docx, etc.)
• sample_data/… ← example inputs
• license & README ← metadata and usage
README for additional insights
• Processing pipeline stages described (“Scanning input directory → Content Understanding → Organize into new structure”).
• Three sorting modes (content, date, type), dry-run & silent modes, real-time progress bar, multiprocessing for performance.
• Relies on external native deps: Tesseract OCR, PyMuPDF.