I had my first opportunity yesterday to play around with the GeoAPI geo-referencing service. Although not quite as intuitive as Google Map’s forward and reverse geocoding services, GoeAPI gives you immediate access to geo-located feeds from Twitter, Flickr, and YouTube, plus the ability to create custom neighborhoods (what they called “guids”) for you to query.

Sounds nerdy – so here’s the English translation:

You now can, with a few lines of code, get Tweets happening real-time in your neighborhood.

Fun? I guess it depends on where you live. Mostly what I’ve been finding is that my neighbors are kind of pervs.

Here’s a quick and dirty Python code sample for you to get started. You’ll need your own API key first.

Also, you’ll need to know the latitude and longitude of a point in the neighborhood you’re interested in. Try (37.766 -122.417) for the Mission. Also, because sometimes you get garbage for location data (like Where: !!!!Frisco U noe!!!!), I’ve add a quick parser to find correctly formatted latitude and longitude pairs. Plenty more you can do with this with a few minutes of time.

– Nerdy Part –

import sys, os, json, urllib, re

#usage: python get_tweets.py [lat] [lng]

lat = sys.argv[1]
lng = sys.argv[2]

# Regex to find properly formatted lat-long pairs in the where response.

lat_lng_re = re.compile('^[\D]*[^-^\d](?P[\d\.-]+)\s*\,\s*(?P[\d\.-]+).*')
api = '[YOUR API KEY]'

parent_request = 'http://api.geoapi.com/v1/parents?lat=%s&lon=%s&apikey=%s' % (lat,lng,api)
response = urllib.urlopen(parent_request).read()

json1 = json.loads(response)

# Find your neighborhood 

guid = json1['result']['parents'][0]['guid']
print "\n\nYour neighborhood: %s\n\n" % json1['result']['parents'][0]['meta']['name']
print "--TWEETS--\n\n"

# Get Tweets in your neighborhood

tweets_request = 'http://api.geoapi.com/v1/e/%s/view/twitter?apikey=%s' % (guid,api)

response = urllib.urlopen(tweets_request).read()

json2 = json.loads(response)

for tweet in json2['result']:
	print "Tweet: %s" % tweet['text']
	print "When: %s" % tweet['created_at']
	loc = lat_lng_re.match(tweet['location'])
	if loc:
		print "Lat: %s Lng: %s" % (loc.group('lat'),loc.group('lng'))
	else:
		print "Unusable location: %s" % tweet['location']
	print "User: %s" % tweet['from_user']
	print "Profile image url: %s\n" % tweet['profile_image_url']

It was fun working with a great team of volunteers to launch an awesome site on a short timeframe. Its the most video-focused site we’ve worked on to date.

The team recently posted a few funny videos highlighting the serious topic of healthcare reform:

I like this guy, who debuts around a minute thirty into the report.  I haz home computer too.