Understanding Asynchronous/Multiprocessing in Pyth

Question

Lets say I have a function:

from time import sleep

def doSomethingThatTakesALongTime(number):
  print number
  sleep(10)

and then I call it in a for loop

for number in range(10):
  doSomethingThatTakesALongTime(number)

How can I set this up so that it only takes 10 seconds TOTAL to print out:

$ 0123456789

Instead of taking 100 seconds. If it helps, I'm going to use the information YOU provide to do asynchronous web scraping. i.e. I have a list of sites I want to visit, but I want to visit them simultaneously, rather than wait for each one to complete.

Answer 1

asyncoro supports asynchronous, concurrent programming. It includes asynchronous (non-blocking) socket implementation. If your implementation does not need urllib/httplib etc. (that don't have asynchronous completions), it may fit your purpose (and easy to use, as it is very similar to programming with threads). Your above problem with asyncoro:

import asyncoro

def do_something(number, coro=None):
    print number
    yield coro.sleep(10)

for number in range(10):
    asyncoro.Coro(do_something, number)

Answer 2

Try to use Eventlet — the first example of documentation shows how to implement simultaneous URL fetching:

urls = ["http://www.google.com/intl/en_ALL/images/logo.gif",
     "https://wiki.secondlife.com/w/images/secondlife.jpg",
     "http://us.i1.yimg.com/us.yimg.com/i/ww/beta/y3.gif"]

import eventlet
from eventlet.green import urllib2

def fetch(url):
  return urllib2.urlopen(url).read()

pool = eventlet.GreenPool()
for body in pool.imap(fetch, urls):
  print "got body", len(body)

I can also advise to look toward Celery for more flexible solution.

Answer 3

Take a look at scrapy framework. It's intended specially for web scraping and is very good. It is asynchronus and built on twisted framework.

http://scrapy.org/

Answer 4

Just in case, this is the exact way to apply green threads to your example snippet:

from eventlet.green.time import sleep
from eventlet.greenpool import GreenPool

def doSomethingThatTakesALongTime(number):
    print number
    sleep(10)

pool = GreenPool()

for number in range(100):
    pool.spawn_n(doSomethingThatTakesALongTime, number)

import timeit
print timeit.timeit("pool.waitall()", "from __main__ import pool")
# yields : 10.9335260363