Home → SCRIPTS → PDF Annotation Extractor

PDF Annotation Extractor

Python script that extracts annotations (highlights, comments, underlines, etc.) from PDF files. By default, it shows the annotated text. Use --with-context to also show surrounding text.

Setup

# Create virtual environment
python -m venv pdf_env

# Activate virtual environment:
# Linux/Mac:
source pdf_env/bin/activate
# Windows:
pdf_env\Scripts\activate

# Install required package
pip install PyMuPDF

# Optional: Set alias for python3 (if needed)
# Linux/Mac (add to ~/.bashrc or ~/.bash_profile):
alias python=python3
alias pip=pip3
# Windows (PowerShell profile):
Set-Alias python python3
Set-Alias pip pip3

Usage

# Basic usage (shows annotations only):
python pdf.py document.pdf

# With context (shows 100 characters before/after each annotation):
python pdf.py document.pdf --with-context

Output

• Shows page number, annotation type, and annotated text
• Optional context around annotations with --with-context flag
• Separators: === between annotations, --- between sections within same annotation

Script Code

import fitz  # PyMuPDF
import sys
import argparse

def extract_annotations(pdf_path, show_context=False):
    try:
        doc = fitz.open(pdf_path)
        annotations = []

        # Iterate through each page
        for page_num in range(len(doc)):
            page = doc.load_page(page_num)
            annots = page.annots()

            if annots:
                # Get all text from the page
                full_page_text = page.get_text()

                for annot in annots:
                    annot_type = annot.type[1] if annot.type else "Unknown"
                    content = annot.info.get("content", "")
                    rect = annot.rect

                    if annot_type in ["Text", "FreeText", "Highlight", "Underline", "StrikeOut", "Squiggly"]:
                        # Get text directly under the annotation
                        annotated_text = page.get_text("text", clip=rect).strip()

                        # Find context by searching in full page text
                        context = ""
                        if show_context and annotated_text and annotated_text in full_page_text:
                            # Find the position of annotated text in full page text
                            start_pos = full_page_text.find(annotated_text)
                            if start_pos != -1:
                                # Extract context (100 chars before and after)
                                context_start = max(0, start_pos - 100)
                                context_end = min(len(full_page_text), start_pos + len(annotated_text) + 100)
                                context = full_page_text[context_start:context_end].strip()

                        # If we couldn't find it and context is requested, try text blocks
                        if show_context and not context:
                            text_blocks = page.get_text("blocks")
                            for block in text_blocks:
                                block_rect = fitz.Rect(block[:4])
                                if rect.intersects(block_rect):
                                    context = block[4].strip()
                                    break

                        # Format the output
                        result = f"Page {page_num + 1}: [{annot_type}]"

                        if content:
                            result += f"\n---\nComment: {content}"

                        if annotated_text:
                            result += f"\n---\nAnnotated Text: {annotated_text}"

                        if context and show_context:
                            result += f"\n---\nContext: {context}"

                        annotations.append(result)

        # Print the results
        if annotations:
            for i, annot in enumerate(annotations):
                print(annot)
                if i < len(annotations) - 1:  # Don't print separator after last annotation
                    print("=" * 80)
        else:
            print("No annotations found in the PDF.")

    except Exception as e:
        print(f"Error opening PDF: {e}")

if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("pdf_file", help="PDF file path")
    parser.add_argument("--with-context", action="store_true", help="Include context around annotations")
    args = parser.parse_args()

    extract_annotations(args.pdf_file, args.with_context)

Last updated: 2025-08-26 20:00 UTC