diw-wc.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
Filling out forms is hard, let's do some Python.

This program determines the symbol and word count excluding markup for
a (X)HTML document, by passing a POST request to the count_words.php
script at the INF5270 web site.

Note that there is NO error handling whatsoever, so use at your own peril.

Usage: 
./diw-wc.py file.html
cat file.html | ./diw-wc.py

Dependencies:
BeautifulSoup; on Debian/Ubuntu, sudo aptitude install python-beautifulsoup,
or pull it from PyPi with pip install BeautifulSoup or
easy_install BeautifulSoup
"""
from sys import argv, stdin
from urllib import urlencode
from urllib2 import urlopen
from BeautifulSoup import BeautifulSoup
import re

def getCounts(html):
    """
    Fetch symbol and word counts, excluding (X)HTML tags from the
    script at the INF5270 web site.
    """
    site = urlopen("http://www.ifi.uio.no/~inf5270/ownstuff/count_words.php",
                   urlencode({"body": html}))
    soup = BeautifulSoup(site.read())
    return  tuple(soup.findAll(text=re.compile("^[0-9]+$")))

if __name__ == "__main__":
    if len(argv) == 1:
        inputHTML = "".join([line for line in stdin])
    else:
        inputHTML = "".join([line for line in open(argv[1])])

    counts = getCounts(inputHTML)
    print "{0} symbols, {1} words".format(counts[0], counts[1])

emacs integration

(defun diw-word-count ()
  "Count symbols and words in buffer, excluding (X)HTML markup"
  (interactive)
  (shell-command-on-region (point-min) (point-max) "diw-wc.py"))