Skip to content

text-extractor

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

text-extractor

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
screenshots
screenshots
 
 
README.md
README.md
 
 

README.md

PDF Text Extractor Documentation

The PDF Text Extractor allows you to extract text from PDF documents using multiple methods: Python-based extraction, OCR for scanned documents, and Azure AI Document Intelligence for high-quality extraction.

Features

  • Multiple Extraction Methods:
    • Python (PyMuPDF): Fast, offline extraction for digital PDFs
    • OCR (OCRmyPDF): Extract text from scanned documents
    • Azure AI: High-quality extraction with layout preservation
  • Multiple Output Formats: Text (.txt), Markdown (.md), JSON
  • Drag and Drop: Drop PDF files directly into the application
  • Batch Processing: Process multiple PDFs at once
  • Folder Processing: Process entire folders with optional recursion
  • Layout Preservation: Maintains document structure where possible

Screenshots

Main Window

Main interface showing extraction methods, output formats, and drag & drop zone

Screenshot Description
01-main-window.png Main interface
02-drag-drop.png Drag and drop zone
03-file-processing.png File being processed
04-results.png Extraction results
05-azure-config.png Azure configuration dialog

Usage

Opening Files

  1. Drag and Drop: Drag PDF files onto the drop zone
  2. Select Files: Click "Select PDF Files" to browse
  3. Select Folder: Click "Select Folder" to process all PDFs in a directory

Choosing Extraction Method

Select the appropriate method based on your PDF type:

Method Best For Requirements
Python (PyMuPDF) Digital PDFs with embedded text None (included)
OCR Scanned documents, image-based PDFs Tesseract OCR
Azure AI Complex layouts, high accuracy needed Azure subscription

Output Formats

  • Text (.txt): Plain text with page markers
  • Markdown (.md): Structured text with headings and formatting
  • JSON: Full structured data with metadata

Azure AI Configuration

To use Azure AI Document Intelligence:

  1. Click "⚙️ Azure Config" in the options panel
  2. Enter your Azure Document Intelligence endpoint
  3. Enter your API key
  4. Click "Test Connection" to verify
  5. Click "Save" to store settings

Alternatively, set environment variables:

  • AZURE_DOC_INTEL_ENDPOINT: Your endpoint URL
  • AZURE_DOC_INTEL_API_KEY: Your API key

Command Line Usage

# Extract text from a single PDF
python src/pdf_text_extractor.py document.pdf

# Extract using OCR
python src/pdf_text_extractor.py --method ocr scanned.pdf

# Extract using Azure AI
python src/pdf_text_extractor.py --method azure document.pdf

# Output as Markdown
python src/pdf_text_extractor.py --format markdown document.pdf

# Process folder recursively
python src/pdf_text_extractor.py --recursive folder/

# Force overwrite existing files
python src/pdf_text_extractor.py --force document.pdf

# Start GUI mode
python src/pdf_text_extractor.py --gui

Command Line Options

Option Description
--method python Use PyMuPDF extraction (default)
--method ocr Use OCR-based extraction
--method azure Use Azure AI extraction
--format text Output as plain text (default)
--format markdown Output as Markdown
--format json Output as JSON
--recursive, -r Process folders recursively
--force Overwrite existing output files
--output-dir, -o Custom output directory
--gui Start GUI mode
--debug Enable debug output

Technical Details

  • Source File: src/pdf_text_extractor.py
  • Launch Script: launch_pdf_text_extractor.bat / launch_pdf_text_extractor.sh
  • Framework: tkinter with ttk widgets, tkinterdnd2 for drag-and-drop
  • Display Name: PDF Text Extractor (in launcher)
  • PDF Libraries:
    • PyMuPDF (fitz) for Python extraction
    • OCRmyPDF + Tesseract for OCR
    • Azure Document Intelligence for AI extraction

Dependencies

Required

  • pymupdf - PDF reading and text extraction
  • pillow - Image processing
  • pyyaml - Configuration files

Optional

  • ocrmypdf - OCR support (requires Tesseract OCR)
  • requests - Azure API calls
  • azure-identity - Azure authentication
  • tkinterdnd2 - Drag and drop support

Keyboard Shortcuts

Shortcut Action
Ctrl+O Open file
Escape Cancel operation

Tips

  • For digital PDFs: Use Python method for fastest results
  • For scanned PDFs: Use OCR method to extract text from images
  • For complex layouts: Use Azure AI for best layout preservation
  • For batch processing: Select a folder and enable recursive search
  • To skip existing: Leave "Overwrite existing files" unchecked

Troubleshooting

"PyMuPDF not available"

Install with: pip install pymupdf

"OCRmyPDF not available"

  1. Install Tesseract OCR on your system
  2. Install package: pip install ocrmypdf

"Azure AI not configured"

  1. Create an Azure Document Intelligence resource
  2. Configure endpoint and API key in the Azure Config dialog
  3. Or set environment variables

OCR quality is poor

  • Ensure the scanned document is high resolution
  • Try preprocessing the PDF (deskew, enhance contrast)
  • Use Azure AI for better results