def run(self):
totalTestCount = 0
totalWaitingTime = 0
successCount = 0
failCount = 0
# Not support non http protocol for now.
if self.proxy.protocol != 'http':
return
for i in range(0, self.totalTestNumber):
success, waitingTime = self.makeRequest()
if success:
successCount += 1
totalWaitingTime += waitingTime
else:
failCount += 1
totalWaitingTime += 10000
totalTestCount += 1
if failCount >= self.acceptableFailTestNumber:
break
time.sleep(1)
if successCount >= 1:
averageTime = float(totalWaitingTime) / successCount
else:
averageTime = 9999999
self.proxy.averageTime = averageTime
self.proxy.testCount = totalTestCount
self.proxy.successCount = successCount
log.debug('Finish single proxy test: ' +\
str(self.proxy.averageTime) + ' ' +\
str(self.proxy.successCount) + '/' + str(self.proxy.testCount) +\
' ' + self.proxy.getAllString())
def run(self):
try:
opener = urllib2.build_opener()
html = opener.open(self.url, timeout=10)
proxies = self.parser.parse(html, self.url)
for proxy in proxies:
self.importer.addProxy(proxy)
log.debug('Crawl proxies from ' + str(self.url) + ':')
for proxy in proxies:
log.debug('>>>>>' + proxy.getAllString())
except Exception, e:
log.warning('Crawling thread exception: ' + str(e))
def makeRequest(self):
success = False
startTime = time.time()
try:
protocol = self.proxy.protocol
if not protocol:
# Default to HTTP protocol
protocol = 'http'
proxy_handler = urllib2.ProxyHandler({
protocol.lower(): self.proxy.getProxyString()
})
opener = urllib2.build_opener(proxy_handler)
response = opener.open(self.url, timeout=5)
response.read()
success = True
except Exception, e:
log.debug('Fail on proxy test: ' + str(e) + ' ' +\
self.proxy.getAllString())
def parseProfileHtml(self, html):
"""Returns: {(UserInfo, ErrorCode)} the user information and errorcode.
"""
document = BeautifulSoup(html)
errorCode = ErrorCode.LACK_CRITICAL_INFO
pagetype = -1
info = None
if document.find('div', id = 'timeline'):
# Timeline profile page
log.debug('The page type is: Timeline')
pagetype = PageType.TIME_LINE
info = self.parseTimelineProfilePage(document)
else:
# Old profile page
if document.find('div', id = 'visitors'):
# With access permission
log.debug('The page type is: old with access')
pagetype = PageType.OLD_WITH_ACCESS
info = self.parseOldProfilePageWithAccess(document)
elif document.find('div', id = 'allFrdGallery'):
# Without access permission
log.debug('The page type is: old without access')
pagetype = PageType.OLD_WITHOUT_ACCESS
info = self.parseOldProfilePageWithoutAccess(document)
else:
# Unknown page tamplate
errorCode = ErrorCode.UNKNOWN_PAGE
if info:
info.pagetype = pagetype
info.html = html
return (info, ErrorCode.OK)
else:
# Parse error
return (None, errorCode)
def parseOldProfilePageWithAccess(self, document):
"""Returns: {UserInfo} the user information."""
info = UserInfo()
# Name & visitedNum
try:
holderNode = document.find('div', class_='status-holder')
info.name = holderNode.find('h1', class_='username').string
info.visitedNum = int(
holderNode.find('span', class_='count')
.find('span', class_='count').string)
except:
return None # Lack of critical information
summaryNode = document.find('div', class_='profile-summary')
# Gender
try:
# sayHiText is something like: '向他打招呼'
sayHiText = summaryNode.ul.find('a', stats='pf_poke').string
if re.compile(u'.*他.*').match(sayHiText):
info.gender = 'male'
elif re.compile(u'.*她.*').match(sayHiText):
info.gender = 'female'
except Exception, e:
log.debug('Exception in gender: ' + str(e))
pass
except:
pass
# Hometown
try:
hometownNode = ul.find('li', class_='hometown')
hometownString = ''
# TODO: Some times they don't user <a> to palce hometown string
for aTag in hometownNode.find_all('a'):
hometownString += aTag.string + ' '
if not hometownNode.find_all('a'):
# Use string to directly represent hometown
pattern = re.compile('\S*\s(.*)')
hometownString = pattern.match(hometownNode.string).group(1)
info.hometown = hometownString
except Exception, e:
log.debug('Exception in hometown: ' + str(e))
pass
# Residence
try:
residenceNode = ul.find('li', class_='address')
placeString = ''
# TODO: Some times they don't user <a> to palce hometown string
for aTag in residenceNode.find_all('a'):
placeString += aTag.string + ' '
if not residenceNode.find_all('a'):
# Use string to directly represent hometown
pattern = re.compile('\S*\s(.*)')
placeString = pattern.match(residenceNode.string).group(1)
info.residence = placeString
except Exception, e:
log.debug('Exception in Residence: ' + str(e))
