count.py

#!/usr/bin python
import os, sys, oauth2, urlparse, pytumblr, json, time
from difflib import context_diff
from unicodedata import normalize
from pprint import pprint

class bcolors:
    HEADER = '\033[95m'
    OKBLUE = '\033[94m'
    OKGREEN = '\033[92m'
    WARNING = '\033[93m'
    FAIL = '\033[91m'
    ENDC = '\033[0m'
    BOLD = '\033[1m'
    UNDERLINE = '\033[4m'

# GLOBAL VARIABLES
storageDir = '.saves/'
REQUEST_TOKEN_URL = 'http://www.tumblr.com/oauth/request_token'
AUTHORIZATION_URL = 'http://www.tumblr.com/oauth/authorize'
ACCESS_TOKEN_URL  = 'http://www.tumblr.com/oauth/access_token'
CONSUMER_KEY = ''
CONSUMER_SECRET = ''

def getBlogUrl(blogurl=""):
    # CHECK FOR ARGV[1]
    try:
        blogurl = sys.argv[1]
    except:
        while not blogurl:
            sys.stderr.write("At least give me a blog URL: ")
            blogurl = str(raw_input(""))
    if not "." in blogurl:
        blogurl = blogurl+'.tumblr.com'
    return blogurl

def getFollowers(client, blogurl):
    setFollowers = set()
    followers = []
    offset = 0
    doneCounting = False
    firstLine = True
    counter = 0
    
    while True:
        try:
            response = client.followers(blogurl, offset=offset)['users']
            if firstLine:
                firstLine = False
                sys.stdout.write("Retrieving followers")
            offset = offset + 20
            sys.stdout.flush() #to clear output buffer
            sys.stdout.write(".")
            # sys.stdout.write("Got %d new followers\n" % len(response))

            if len(response) == 0:
                doneCounting = True

            for i in range(0,len(response)):
                try:
                    #extract name field and convert from unicode
                    user = response[i]['name'].encode('ascii', 'ignore')
                    if user not in setFollowers:
                        setFollowers.add(user)
                        followers.append(user)
                        counter = counter + 1
                except IndexError:
                    doneCounting = True
                    break
            if doneCounting:
                sys.stdout.write("\n")
                break
        except KeyError:
            sys.stdout.flush() #to clear output buffer
            sys.stderr.write("\nYou gave me an incorrect blog! I have had it up to here with you honestly.\n")
            blogurl = getBlogUrl()
        except Exception as e:
            sys.stdout.flush() #to clear output buffer
            sys.stderr.write("\nError retrieving. Try again in a minute.\n")
           # sys.stderr.write(e)
            exit()
    
    return followers, counter

def WriteToFile(followers,storageDir):
    # WRITE FOLLOWERS TO FILE
    fh = open(storageDir+'out', 'w')
    for user in followers:
        fh.write(user+"\n")
    fh.close()

def loadOldFollowers(storageDir):
    # If this is the first time this script is being run, there will be no old followers
    if not os.path.exists(storageDir+'old'):
        sys.stdout.write("Thanks for using. :) "
                     + "We will have something for you the next time you run it.\n")
        exit()

    # LOAD OLD FOLLOWERS
    oldfollowers = []
    fo = open(storageDir+'old', 'r')
    for line in fo:
        line = line.strip()
        oldfollowers.append(line)
    fo.close()
    #print '\n'.join(oldfollowers)
    return oldfollowers

def TumblrExists(url, client):
    statuscode = client.blog_info(url)
    try:
        if statuscode['meta']['status'] == 200:
            return True
        else:
            #print statuscode['msg']
            return False
    except KeyError:
        if len(statuscode['blog']) > 0:
            return True
    except:
        print "Error checking blog!!"
        
def analyzeFollowers(followers, oldfollowers, client):
    ## This will compare the two lists.
    ## it prints 5 lines of context around a change (unless it is near ends of file)
    ## shows a + for additions, - for deletions, ! for line changes
    ## The first two lines are useless
    
    ## Implementation note: urlchanges is special because we have an old url and the new url.
    ## We get both of those at separate, but consecutive lines. So once we hit one, 
    ## I set a toggle to wait for the next line which should contain the new URL.
    
    ## n in context_diff() sets number of lines of context to show. it fixes a case where two
    ## users appearing consecutively might fall in chargedurls, and get mixed up
    newfollows, unfollows, deactives, urlchanges = [],[],[],[]
    newurl = False
    for line in context_diff(oldfollowers, followers, n=0):
        if line.startswith('+ '):
            newfollows.append(line.strip().split(' ')[1])
        if line.startswith('- '):
            if TumblrExists(line.strip().split(' ')[1], client):
                unfollows.append(line.strip().split(' ')[1])
            else:
                deactives.append(line.strip().split(' ')[1])
        if line.startswith('! '): 
            if newurl:
                templist.append(line.strip().split(' ')[1]) #add to existing list
                urlchanges.append(templist)
                newurl = False
            else:
                templist = list() #make new list
                templist.append(line.strip().split(' ')[1])
                newurl = True     #signal that next line is going to be new url
    
    # DO THE PRINTING STUFF
    if newfollows:
        sys.stdout.write(bcolors.OKGREEN + str(len(newfollows))
                         + " new cultists!!\n"
                         + bcolors.ENDC)
        temp = [sys.stdout.write(" %s," % (newfollower)) for newfollower in newfollows]
        sys.stdout.write("\b  \n") #sketchy way to erase the last comma
    else:
        sys.stdout.write("No new people found you interesting.\n")
        
    if unfollows:
        sys.stdout.write(bcolors.FAIL + str(len(unfollows)) 
                         + " have chosen to pursue a career in the landfill business." 
                         + " Nothin but a bunch of LOSERS!!\n"
                         + bcolors.ENDC)
        temp = [sys.stdout.write(" %s," % (unfollower)) for unfollower in unfollows]
        sys.stdout.write("\b  \n") #erase the last comma
    else:
        sys.stdout.write("No one ran away from your stank. This must be your good week.\n")
    
    if deactives:
        sys.stdout.write(bcolors.FAIL + str(len(deactives)) 
                         + " people deactivated. A prayer for those freed.\n" 
                         + bcolors.ENDC)

        temp = [sys.stdout.write(" %s," % (deactive)) for deactive in deactives]
        sys.stdout.write("\b  \n") #erase the last comma

    if urlchanges:
        sys.stdout.write(bcolors.OKBLUE + str(len(urlchanges)) 
                         + " people changed their URL. Lord help us.\n" 
                         + bcolors.ENDC)
        temp = [sys.stdout.write("%13s -> %s\n" % (oldurl,newurl)) for oldurl,newurl in urlchanges]
        #13 is the average length of a url (based on scientific study done on my followers list)

########
#  MAIN 
########

blogurl = getBlogUrl()

# Authenticate via OAuth
client = pytumblr.TumblrRestClient(
  'VMXZOL8hIJ0iaFhjjErPWACD2hAGnWXwaTKskG67ZFuWAz5pfK',
  'knwMMzMlCmTZGvlXHhtfUFs1qm0NxMqsBEdbKtYDYVa3fuNw2U',
  'AjVTT65vgbioccIJ9I89Ee6Ug91UGtLK9REUjAlfvdtKmTtCsL',
  'BCADjechYrTsfZFK154mOohXkMP2kOpE0WW8gDIxlueU3viLzC'
)

start_time = time.time()
followers, followercnt = getFollowers(client, blogurl)
end_time = time.time()
if not followercnt:
    sys.stderr.write("Buddy, you don't have any followers.\nAre you sure this is the right blog?\n")
    exit()
elapsed_time = end_time - start_time
print (str(elapsed_time) + " seconds to retrieve " + str(followercnt) + " followers")
WriteToFile(followers,storageDir)
oldfollowers = loadOldFollowers(storageDir)
analyzeFollowers(followers, oldfollowers, client)
print ""