simonw

import re

explicits = (
    u'\u202a', # LEFT-TO-RIGHT EMBEDDING
    u'\u202b', # RIGHT-TO-LEFT EMBEDDING
    u'\u202d', # LEFT-TO-RIGHT OVERRIDE
    u'\u202e', # RIGHT-TO-LEFT OVERRIDE
)
pdf = u'\u202c' # POP DIRECTIONAL FORMATTING

simonw

"""
Generator that crawls an Apache directory listing (automatically using the
?F=0 querystring parameter which causes Apache to return simpler HTML) and 
yields file URLs one by one.
"""

import urllib, urlparse, re

always = lambda x: True

simonw

"""
Given a list of scores, returns the index of one of those items at random, 
biased towards the items with the highest scores.

e.g. random_index_with_bias([84, 62, 61, 67, 45])

To show it works, run it 1,000,000 times:

>>> scores = {}
>>> for i in range(1000000):

simonw

/* Fade in/out a different background image every 60 seconds

Relies on the background image itself being applied to the html tag, 
using classes bg1 through bg4 for four different background images.
*/
jQuery.fx.step.bgfadeout = function(fx) {
    fx.elem.style.backgroundColor = 'rgba(255,255,255,' + fx.now + ')';
};
jQuery.fx.step.bgfadein = function(fx) {
    fx.elem.style.backgroundColor = 'rgba(255,255,255,' + (1 - fx.now) + ')';

simonw

"""
A ridiculously simple term extractor - extracts sequences of one or more Title Case 
Words, e.g. "Gordon Brown is the Prime Minister" would extract "Gordon Brown" and 
"Prime Minister".
"""

import re

# Stopwords from http://snowball.tartarus.org/algorithms/english/stop.txt
stop_words = set(map(lambda s: s.strip(),

simonw

// node-reverse-proxy

var sys = require('sys'),
    http = require('http');

function proxy(backend, request, response) {
    var bits = backend.split(':');
    var host = bits[0];
    if (bits.length == 2) {
        var port = parseInt(bits[1], 10);

simonw

# Fiddly.

def binary_search(haystack, needle, lo=0, hi=None, size=None):
    if hi is None:
        hi = len(haystack) - 1
    if size is None:
        size = len(haystack)
    if size == 0:
        return -1
    t = ((hi - lo) / 2)

simonw

"For generating fake IP addresses"

import random

def make_ip_issuer(chance_of_a_repeat = 0.0):
    issued_ips = []
    def issue():
        if issued_ips and random.uniform(0.0, 1.0) < chance_of_a_repeat:
            return random.choice(issued_ips)
        ip = '%s.%s.%s.%s' % (

simonw

"Extracts the text from every slide in a keynote presentation"

# Limitation: doesn't currently deal with markup nested inside a p tag, e.g. 
# the following:
# <ns2:p ns2:list-level="1" ns2:style="SFWPParagraphStyle-990">Pen knife photo: 
# herzogbr      <ns2:link
#    href="http://www.flickr.com/photos/80516279@N00/2274372747/"><ns2:span 
#    ns2:style="SFWPCharacterStyle-158"
# >http://www.flickr.com/photos/80516279@N00/2274372747/</ns2:span></ns2:link>
# </ns2:p>

simonw

"""
Simple timer class. Example usage:

t = Timer()
for url in ('http://www.google.com/', 'http://www.yahoo.com/'):
    t.start('fetch url')
    content = urllib.urlopen(url).read()
    t.stop('fetch url')
    t.start('reorder bytes')
    bytes = list(content)