Builds epub book out of Paul Graham's essays.

pgessays.py

# -*- coding: utf-8 -*-
"""
Builds epub book out of Paul Graham's essays: http://paulgraham.com/articles.html

Author: Ola Sitarska <[email protected]>

This script requires python-epub-library: http://code.google.com/p/python-epub-builder/
"""

import re, ez_epub, urllib2, genshi
from BeautifulSoup import BeautifulSoup

def addSection(link):
    page = urllib2.urlopen('http://www.paulgraham.com/'+link).read()
    soup = BeautifulSoup(page)
    soup.prettify()

    section = ez_epub.Section()
    try:
        section.title = soup.findAll('table', {'width':'455'})[0].find('img')['alt']
        print section.title
        for p in str(soup.findAll('table', {'width':'455'})[0].find('font')).split("<br /><br />"):
            section.text.append(genshi.core.Markup(p))
    except:
        pass
    
    return section


book = ez_epub.Book()
book.title = "Paul Graham's Essays"
book.authors = ['Paul Graham']

page = urllib2.urlopen('http://www.paulgraham.com/articles.html').read()
soup = BeautifulSoup(page)
soup.prettify()

links = soup.findAll('table', {'width': '455'})[1].findAll('a')
sections = []
for link in links:
    sections.append(addSection(link['href']))

book.sections = sections
book.make(book.title)