Sam Sargeant

I’ve subscribed to the email updates from our police for a while now. Given these updates were usually about things happening live, I thought it’d be useful to write something to post them into Twitter.

After about an hour of hacking, I had a quick and dirty script that did the job. There are three users you can follow if you are so inclined:

• Auckland Police
• Wellington Police
• Christchurch Police

The script uses BeautifulSoup, a particularly excellent python library that ZachTwo reminded me about. I’d prefer it if the police released their updates in a better format, but screen-scraping will do for now.

Read on for the script. Consider it under a GPL licence.

#!/usr/bin/python

from BeautifulSoup import BeautifulSoup, SoupStrainer
import re
import urllib2
import tinyurl
import pickle
import twitter
import logging

logFile = '/tmp/twitter-police.log'
logging.basicConfig(filename=logFile, level=logging.DEBUG)

twitterDistrict = dict()

pickleFile = '/home/sam/.twitter-police.pickle'
twitterPass = 'secret password goes here'

twitterDistrict['Auckland City'] = twitter.Twitter('nzpolice_akl', twitterPass)
twitterDistrict['Wellington'] = twitter.Twitter('nzpolice_wgtn', twitterPass)
twitterDistrict['Canterbury'] = twitter.Twitter('nzpolice_chch', twitterPass)

twitterDistrict['Counties Manukau'] = twitterDistrict['Auckland City']

try:
lastId = pickle.load(open(pickleFile))
except EOFError:
lastId = 7600

logging.debug('Loading events from ID %d' % lastId)

class InvalidUpdate(Exception):
pass

def simplifyUpdate(node):
if len(node) == 0:
return ''
return node.contents[0]

def makeEvent(url):
result = dict()
logging.debug('Loading %s' % url)
updateContent = urllib2.urlopen(url).read()
soup = BeautifulSoup(updateContent)
updateMessage = soup.findAll('td', { 'class': 'updateMessage' })
for i in range(0, len(updateMessage), 2):
key = updateMessage[i].contents[0].contents
if len(key) == 1:
key = key[0][:-1]
value = updateMessage[i+1].contents
if key == 'Time' and value[0].contents[0] == u'12:00  1-01-1970':
raise InvalidUpdate()
if key == 'What happened':
result[key] = ' '.join(map(simplifyUpdate, value))
else:
result[key] = value[0].contents[0]
return result

def loadUpdate(id):
url = 'http://police.govt.nz/news/updates/update.html?id=%d' % id
event = makeEvent(url)
logging.info('Found %s update for %s' % (event['Location'], event['District']))
if twitterDistrict.has_key(event['District']):
logging.info('Posting to twitter')
api = twitterDistrict[event['District']]
postUrl = tinyurl.create_one(url)
suffix = '... %s' % postUrl
updateLength = 140 - len(suffix)
postBody = event['What happened'][0:updateLength] + suffix
api.statuses.update(status=postBody)
return True

try:
while loadUpdate(lastId):
lastId = lastId + 1
except InvalidUpdate:
pass
finally:
pickle.dump(lastId, open(pickleFile,'w'))