Heuristically splits a book pdf into chapters

pdf-separator.py

# This script takes a PDF book, splits it into chapters, and saves each chapter as a separate PDF file.

import PyPDF2
import re

# Open the PDF file
pdf_file = open('book.pdf', 'rb')
pdf_reader = PyPDF2.PdfReader(pdf_file)

# Create a PDF writer object
pdf_writer = PyPDF2.PdfWriter()

# Regular expression to match the chapter headings
chapter_pattern = re.compile(r'^Chapter \d+$')



page_lengths = []
for page_num in range(len(pdf_reader.pages)):
    page = pdf_reader.pages[page_num]
    text = page.extract_text()
    print(f'Processing page {page_num}...')
    page_lengths.append((page_num, len(text.split(" ")), "chapter" in text.lower()))

print(page_lengths)

# Average path length
average_length = sum([page[1] for page in page_lengths]) / len(page_lengths)

# Filter pages that have chapter
chapter_pages = [page for page in page_lengths if page[2]]
print(chapter_pages)

print("Average length: ", average_length)

# Find chapter pages that are shorter than %60 percent of average length 
short_chapter_pages = [page for page in chapter_pages if page[1] < (average_length * 0.6)]

print(short_chapter_pages)


# Partition pdf by chapter pages
chapter_ranges = []
start = 0
for i in range(len(short_chapter_pages)):
    if i == len(short_chapter_pages) - 1:
        chapter_ranges.append((start, short_chapter_pages[i][0]))
    else:
        chapter_ranges.append((start, short_chapter_pages[i][0]))
        start = short_chapter_pages[i][0]

# Add the last page
chapter_ranges.append((start, len(pdf_reader.pages)))

print(chapter_ranges)

# Save each chapter as a separate PDF file
for i, (start, end) in enumerate(chapter_ranges):
    pdf_writer = PyPDF2.PdfWriter()
    for page_num in range(start, end):
        pdf_writer.add_page(pdf_reader.pages[page_num])
    with open(f'book/chapter_{i+1}.pdf', 'wb') as pdf_output_file:
        pdf_writer.write(pdf_output_file)

# Close the PDF file
pdf_file.close()

print('Chapters separated successfully.')