tweetsForBeats.py

import sqlite3 as lite
import shapefile
from haversine import haversine
import pickle
import sys
from datetime import datetime

def main(beatFile, dbName, shapeFileName, outputFile):
	shapeFile = shapefile.Reader(shapeFileName)
	results = pickle.load(open(beatFile, "rb"))
	con = lite.connect(dbName)
	cur = con.cursor()
	cur.execute("Select * from TweetData")
	tweets = cur.fetchall()
	for i in range(len(shapeFile.records())):
		record = shapeFile.records()[i]
		shape = shapeFile.shapes()[i]
		beatNum = record[3]
		if beatNum == '3100':
			continue
		else:
			for tweet in tweets:
				lat = tweet[2]
				long1 = tweet[3]
				bbox = shape.bbox
				centerLat = (bbox[0] + bbox[2]) / 2
				centerLong = (bbox[1] + bbox[3]) / 2
				haversineDist = haversine(long1, lat, centerLong, centerLat) 
				if haversineDist < 1:
					if beatNum in results:
						tweetDay = datetime.strptime(tweet[4], "%Y-%m-%d %H:%M:%S.%f")
						key = str(tweetDay.month) + "/" + str(tweetDay.day)
						currRecord = results[beatNum][key]['tweets']
						currRecord.append([tweet[0], tweet[1].strip().encode('utf-8'), tweet[4].encode('utf-8')])
		print("processed " + str(i) + " beats")
	pickle.dump(results, open(outputFile, "wb"))
	return results

if __name__ == '__main__':
    main(sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4])