email_crawler_bangalore.py

from settings import LOGGING
import logging, logging.config
import urllib, urllib2
import re, urlparse
import traceback
from databasebangalore import CrawlerDb
from googleapiclient.discovery import build
import pprint

my_api_key = 'API_KEY'
my_cse_id = 'CSE_ID'

def google_search(search_term, api_key, cse_id, **kwargs):
    service = build("customsearch", "v1", developerKey=api_key)
    res = service.cse().list(q=search_term, cx=cse_id, **kwargs).execute()
    return res['items']

# Debugging
# import pdb;pdb.set_trace()

# Logging
logging.config.dictConfig(LOGGING)
logger = logging.getLogger("crawler_logger")

google_adurl_regex = re.compile('adurl=(.*?)"')
google_url_regex = re.compile('url\?q=(.*?)&sa=')
email_regex = re.compile('([A-Z0-9._%+-]+@[A-Z0-9.-]+\.[A-Z]{2,4})', re.IGNORECASE)
url_regex = re.compile('<a\s.*?href=[\'"](.*?)[\'"].*?>')
# Below url_regex will run into 'Castrophic Backtracking'!
# http://stackoverflow.com/questions/8010005/python-re-infinite-execution
# url_regex = re.compile('<a\s(?:.*?\s)*?href=[\'"](.*?)[\'"].*?>')

# Maximum number of search results to start the crawl
MAX_SEARCH_RESULTS = 50

EMAILS_FILENAME = 'data/emails.csv'
DOMAINS_FILENAME = 'data/domains.csv'

# Set up the database
db = CrawlerDb()
db.connect()

def getHostName(url2):
	url = urlparse.urlparse(url2)
	hostname = url.hostname.split(".")
	# Simplistic assumeption of a domain. If 2nd last name is <4 char, then it has 3 parts eg. just2us.com.sg
	hostname = ".".join(len(hostname[-2]) < 4 and hostname[-3:] or hostname[-2:])
	print hostname
	return hostname
def crawl(keywords):
	"""
	This method will

	1) Google the keywords, and extract MAX_SEARCH_RESULTS
	2) For every result (aka website), crawl the website 2 levels deep.
		That is the homepage (level 1) and all it's links (level 2).
		But if level 1 has the email, then skip going to level 2.
	3) Store the html in /data/html/ and update the database of the crawled emails

	crawl(keywords):
		Extract Google search results and put all in database
		Process each search result, the webpage:
			Crawl webpage level 1, the homepage
			Crawl webpage level 2, a link away from the homepage
			Update all crawled page in database, with has_crawled = True immediately
			Store the HTML
	"""
	logger.info("-"*40)
	logger.info("Keywords to Google for: %s" % keywords.decode('utf-8'))
	logger.info("-"*40)

	# Step 1: Crawl Google Page
	# eg http://www.google.com/search?q=singapore+web+development&start=0
	# Next page: https://www.google.com/search?q=singapore+web+development&start=10
	# Google search results are paged with 10 urls each. There are also adurls
	# db.delete()
	# for i in range(1,MAX_SEARCH_RESULTS,10):
	# 	results = google_search(keywords, my_api_key, my_cse_id, num=10,start=i)
	# 	for result in results:
	# 		print result['link']
	# 		db.enqueue(unicode(result['link']))

	# for page_index in range(0, MAX_SEARCH_RESULTS, 10):
	# 	query = {'q': keywords}
	# 	url = 'https://www.google.com/search?' + urllib.urlencode(query) + '&start=' + str(page_index)
	# 	data = retrieve_html(url)
	# 	# 	print("data: \n%s" % data)
	# 	for url in google_url_regex.findall(data):
	# 		db.enqueue(unicode(url))
	# 	for url in google_adurl_regex.findall(data):
	# 		db.enqueue(unicode(url))

	# Step 2: Crawl each of the search result
	# We search till level 2 deep
	while (True):
		# Dequeue an uncrawled webpage from db
		uncrawled = db.dequeue()
		if (uncrawled == False):
			break
		if (getHostName(uncrawled.url) in ['zomato.com','naukri.com', 'facebook.com']):
			db.crawled(uncrawled, 'test@test.com');
		email_set = find_emails_2_level_deep(uncrawled.url)
		if (len(email_set) > 0):
			db.crawled(uncrawled, ",".join(list(email_set)))
		else:
			db.crawled(uncrawled, None)

def retrieve_html(url):
	"""
	Crawl a website, and returns the whole html as an ascii string.

	On any error, return.
	"""
	req = urllib2.Request(url)
	req.add_header('User-Agent', 'Just-Crawling 0.1')
	request = None
	status = 0
	try:
		logger.info("Crawling %s" % url)
		request = urllib2.urlopen(req)
	except urllib2.URLError, e:
		logger.error("Exception at url: %s\n%s" % (url, e))
	except urllib2.HTTPError, e:
		status = e.code
	except Exception, e:
		return
	if status == 0:
		status = 200

	try:
		data = request.read()
	except Exception, e:
		return

	return str(data)


def find_emails_2_level_deep(url):
	"""
	Find the email at level 1.
	If there is an email, good. Return that email
	Else, find in level 2. Store all results in database directly, and return None
	"""
	html = retrieve_html(url)
	email_set = find_emails_in_html(html)

	if (len(email_set) > 0):
		# If there is a email, we stop at level 1.
		print email_set
		return email_set

	else:
		# No email at level 1. Crawl level 2
		logger.info('No email at level 1.. proceeding to crawl level 2')

		link_set = find_links_in_html_with_same_hostname(url, html)
		for link in link_set:
			# Crawl them right away!
			# Enqueue them too
			html = retrieve_html(link)
			if (html == None):
				continue
			email_set = find_emails_in_html(html)
			print email_set
			db.enqueue(link, list(email_set))
			# if len(email_set) != 0:
			# 	break;

		# We return an empty set
		return set()


def find_emails_in_html(html):
	if (html == None):
		return set()
	email_set = set()
	for email in email_regex.findall(html):
		email_set.add(email)
	return email_set


def find_links_in_html_with_same_hostname(url, html):
	"""
	Find all the links with same hostname as url
	"""
	if (html == None):
		return set()
	url = urlparse.urlparse(url)
	links = url_regex.findall(html)
	link_set = set()
	for link in links:
		if link == None:
			continue
		try:
			link = str(link)
			if link.startswith("/"):
				link_set.add('http://'+url.netloc+link)
			elif link.startswith("http") or link.startswith("https"):
				if (link.find(url.netloc)):
					link_set.add(link)
			elif link.startswith("#"):
				continue
			else:
				link_set.add(urlparse.urljoin(url.geturl(),link))
		except Exception, e:
			pass

	return link_set




if __name__ == "__main__":
	import sys
	try:
		arg = sys.argv[1].lower()
		if (arg == '--emails') or (arg == '-e'):
			# Get all the emails and save in a CSV
			logger.info("="*40)
			logger.info("Processing...")
			emails = db.get_all_emails()
			logger.info("There are %d emails" % len(emails))
			file = open(EMAILS_FILENAME, "w+")
			file.writelines("\n".join(emails))
			file.close()
			logger.info("All emails saved to ./data/emails.csv")
			logger.info("="*40)
		elif (arg == '--domains') or (arg == '-d'):
			# Get all the domains and save in a CSV
			logger.info("="*40)
			logger.info("Processing...")
			domains = db.get_all_domains()
			logger.info("There are %d domains" % len(domains))
			file = open(DOMAINS_FILENAME, "w+")
			file.writelines("\n".join(domains))
			file.close()
			logger.info("All domains saved to ./data/domains.csv")
			logger.info("="*40)
		else:
			# Crawl the supplied keywords!
			crawl(arg)

	except KeyboardInterrupt:
		logger.error("Stopping (KeyboardInterrupt)")
		sys.exit()
	except Exception, e:
		logger.error("EXCEPTION: %s " % e)
		traceback.print_exc()