Dictionary-based HTML hyphenation using soft hyphen

hyphenation.py

import os
import re

from htmlentitydefs import name2codepoint

# Must install the `hyphenator` library from PyPi!
from hyphenator import Hyphenator

# Firefox comes with an English hyphenation dictionary
path = os.popen('locate hyph_en_US.dic').readlines()[0].strip()

hyphenator = Hyphenator(path)
spaces = re.compile(r'\s+')
shy = unichr(name2codepoint['shy'])

def hyphenate(string, max_length=20):
    return " ".join(
        len(word) > max_length and
        shy.join(hyphenator.wrap(word, max_length) or (word, )).\
        replace('-', '')
        or word for word in spaces.split(string.strip())
        )