PDF Text Extractor

A Python utility to extract text from PDF files, including scanned PDFs using OCR.

Features

• Extract text from regular (text-based) PDFs
• OCR support for scanned PDFs using Tesseract
• Conservative table data extraction to CSV format
• Handles OCR errors gracefully - leaves blank cells for manual entry
• Works with any product type, not just Apple products
• Table structure analysis with completion statistics
• Preserves original raw text for verification
• Batch processing for multiple PDFs
• Command-line interface
• Save output to files or print to console

Installation

1. Install Python dependencies:

pip install -r requirements.txt

2. Install Tesseract OCR:

macOS:

brew install tesseract

Ubuntu/Debian:

sudo apt-get install tesseract-ocr

Windows: Download and install from: https://github.com/UB-Mannheim/tesseract/wiki

Usage

Command Line Interface

Extract text from a single PDF:

python pdf_text_extractor.py path/to/document.pdf

Save extracted text to a file:

python pdf_text_extractor.py path/to/document.pdf -o output.txt

Process all PDFs in a directory:

python pdf_text_extractor.py path/to/pdf_directory --batch -o output_directory

Disable OCR (text-based PDFs only):

python pdf_text_extractor.py path/to/document.pdf --no-ocr

Extract table data to CSV (auto-names as document.csv):

python pdf_text_extractor.py path/to/document.pdf --csv

Or specify custom output name:

python pdf_text_extractor.py path/to/document.pdf --csv -o custom_name.csv

Analyze table structure:

python pdf_text_extractor.py path/to/document.pdf --analyze

Python API

from pdf_text_extractor import PDFTextExtractor

# Initialize extractor
extractor = PDFTextExtractor()

# Extract text from a PDF
text = extractor.extract_text_from_pdf('document.pdf')
print(text)

# Save to file
extractor.extract_to_file('document.pdf', 'output.txt')

# Batch process
processed = extractor.batch_extract('pdf_folder', 'output_folder')

# Extract table data to CSV (auto-names as invoice.csv)
extractor.extract_tables_to_csv('invoice.pdf', 'invoice.csv')

# Analyze table structure
analysis = extractor.analyze_table_structure('invoice.pdf')
print(f"Found {analysis['total_rows']} table rows")

Options

• -o, --output: Output file or directory path
• --csv: Extract table data to CSV format
• --analyze: Analyze table structure
• --no-ocr: Disable OCR for scanned pages
• --batch: Process all PDFs in directory
• --tesseract-cmd: Path to tesseract executable
• -v, --verbose: Enable verbose logging

Requirements

• Python 3.6+
• PyMuPDF (fitz)
• pytesseract
• Pillow (PIL)
• pandas
• Tesseract OCR engine

Windows Executable

Want to use this without installing Python? We can build a Windows executable:

Build Executable (Windows only)

1. Simple build:

   python build_simple.py

2. Full featured build:

   build_windows.bat

3. Find your executable in:

   dist/pdf-text-extractor.exe
   

Use Executable

# Extract table to CSV (auto-names as invoice.csv)
pdf-text-extractor.exe "invoice.pdf" --csv

# Analyze table structure
pdf-text-extractor.exe "invoice.pdf" --analyze

See BUILD_WINDOWS_EXE.md for detailed instructions.