epub word counter

epub_word_counter.py

# Thanks to https://github.com/aerkalov/ebooklib
#!pip install EbookLib
import ebooklib
from ebooklib import epub
from bs4 import BeautifulSoup
from collections import Counter

book = epub.read_epub('ready_player_two.epub')
content = ''
for item in book.get_items():
    if item.get_type() == ebooklib.ITEM_DOCUMENT:
        soup = BeautifulSoup(item.get_content())
        content += soup.get_text()
        
content = content.replace('\n', ' ')
break_counter = 100
for word, count in Counter(content.split()).most_common():
    print(count, word)
    break_counter -= 1
    if break_counter < 0:
        break