import datetime
import requests
from bs4 import BeautifulSoup

 
# just do current day for illustration
numDays = 1
for j in range(0, numDays):

    nEvents = 0
    
    # get string with date
    date = datetime.date.today() + datetime.timedelta(days = j) 
    dateString = date.strftime("%Y-%m-%d")
    print "DATE:"
    print dateString
    
    
    baseUrl = 'http://www.tucsonweekly.com/tucson/EventSearch?narrowByDate='
    
    # while loop to crawl over pages
    # Set maximum possible number of pages of events: 
    # in practice would be a huge number, and the code should break before this
    npageMax = 2 # for illustration, we'll stop after one page

    
    # base URL for this date
    url = baseUrl + dateString
    ii=0
    for pageNum in xrange(1, npageMax):
        print "On page =", pageNum
    
        # crawl over pages via appending initial date url
        if (pageNum>1):
            url = baseUrl + dateString + "&page=" + str(pageNum)
        
        # parse webpage with BeautifulSoup
        text = requests.get(url).text  # get raw html
        soup = BeautifulSoup(text)     # input it into BeautifulSoup to parse it
        
        # get all "div" tags that corresond to a single event
        eventTags = soup.find_all('div', 'EventListing clearfix')
        print "There are", len(eventTags), "events on this page"
        
        nEvents += len(eventTags)
        

 
# loop over each event and attempt to extract the html link for the page of the event
for i, event in enumerate(eventTags):

    print "EVENT", i ,"START:"
        
    # find all the anchor ('a') tags within the 'div' tag for the evetn
    anchor_tags = event.find_all('a')
        
        
    # iterate over each anchor tag looking for the FIRST one that has a 'href' attribute
    # but does NOT have a 'class' attribute
    isFound = False
    for j,mb in enumerate(anchor_tags):
        
        if (mb.has_attr('href') and (not isFound) and (not mb.has_attr('class')) ):
            event_name = mb.get_text().strip() # this should be the event name! (type unicode)
            event_link = mb.attrs['href']      # this should be the event webpage! (type string)
            
            print "Event name =", event_name
            print "Event link =", event_link
            
            isFound = True

import re

# function to scrape info from a given link on site
def scrape(link):

    # parse webpage with BeautifulSoup
    text = requests.get(link).text
    soup = BeautifulSoup(text)
    
    # extract venue and genre(s) from <title>stuff|stuff|stuff|</title>
    # at top of html
    title = soup.title.string
    title = title.split(' | ')
    name = title[0]
    venue = title[1]
    genre = title[2]
    print "Event name =", name
    print "Venue =", venue 
    print "Kind =", genre ,"\n"
    
    
    ### Get Description
    description = soup.find_all("meta", {"name":"description"} )[0]["content"]
    print "Description =", description ,"\n"
    
    
    ### Get Address stuff
    address = ["og:street-address", "og:latitude", "og:longitude", "og:locality",
               "og:region", "og:postal-code"]
    addr_stuff = []
    for ad in address:
        result = soup.find_all("meta", {"property": ad})
        if (len(result)>0):
            addr_stuff.append(result[0]["content"])
        else:
            addr_stuff.append(None)
        
    
    street = addr_stuff[0]
    lat = addr_stuff[1]
    lon = addr_stuff[2]
    locale = addr_stuff[3]
    region = addr_stuff[4]
    zipc = addr_stuff[5]
    
    print "Address =", street, locale, region, zipc
    print "Lat, lon =", lat, lon ,"\n"
    

    ### Get event meta data
    metaData = soup.find_all("div", {"class":"MainColumn Event ", "id":"EventMetaData"} )
    spans = metaData[0].find_all("span")
    
    ### Look for text indicating event property within each span tag
    foundWhen = False
    foundPrice = False
    textPrice = 'none'
    for i, cont in enumerate(spans):

        if (cont.text=="When:"):
            when = cont.next_sibling.strip()
            foundWhen = True
            
        if (cont.text=="Price:"):
            textPrice = cont.next_sibling.strip()
            foundPrice = True
            

            
    ### Some events don't have a 'when' property, 
    #   e.g. they just say 'every 4th Friday within event description
    if (not foundWhen):
        print "returning none"
        return None
            

    ### Parse price value
    
    # use regular expressions (re) to scrape multiple prices
    # ()    look for groups of the character sequence inside the brackets
    # \$    look for a dollar sign at the front (needs escape character \)
    # \d    look for decimal digit [0-9]
    # +     greedy, make longest sequence possible (match one or more of the preciding RE)
    # \.    look for '.' (needs escape character \)
    # ?     Causes the resulting RE to match 0 or 1 repetitions of the preceding RE. 
    #       ab? will match either 'a' or 'ab'. i.e. doesn't HAVE to match the thing immediately proceeding it
    
    dollarValues = re.findall('(\$?\d+\.?\d?\d?)', textPrice)
    
    # checks if price contains free 
    if (textPrice.count('free') != 0) or (textPrice.count('Free') != 0) or (textPrice.count('FREE') != 0): 
        minprice = maxprice = 0.00
    elif len(dollarValues) == 0:
        minprice = maxprice = None
    else:
        for i in range(0, len(dollarValues)):
            dollarValues[i] = float(dollarValues[i].strip('$'))
        minprice = min(dollarValues)
        maxprice = max(dollarValues)

        
    if (maxprice != minprice):
        print "Price range =", minprice, "to", maxprice
        price = str(minprice) + " to " + str(maxprice)
    else:
        price = minprice
        print "Price =", minprice
    
    print "\nWhen =", when ,"\n"
    
    return name, description, street + locale + region + zipc, price, when

    
# returns tuple of name, description, street + locale + region + zipc, price, when
print scrape(event_link)