Python script to scrape the number of Issued U.S. utility patents

number_of_US_utility_patents.py

import urllib2
import re
import datetime
import pprint

# Start by Getting Date of Last Tuesday
today = datetime.date.today()
tuesday = today - datetime.timedelta(days=today.weekday()) + datetime.timedelta(days=1)
month = str(tuesday.month)
day = str(tuesday.day)
year = str(tuesday.year)

# Create Search Address with Correct Search Options
search_address = "http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&u=/netahtml/PTO/search-adv.htm"
options = "&r=0&p=1&f=S&l=50&Query=+ISD/%s/%s/%s+AND+APT/1&d=PTXT" % (month, day, year)
the_add = search_address + options

# Query USPTO search
response = urllib2.urlopen(the_add)
html = response.read()

# Extract Patent Numbers and Sort
nums = re.findall('(\d),(\d\d\d),(\d\d\d)', html)
nums2 = []
for num in nums:
  nums2.append(''.join(num))
sortednums = sorted(nums2)
highest = sortednums[len(sortednums)-1]

#Print out result
print "The Total Number of U.S. Patents issued is: %s" % str(highest)