class YoukuGhostDriver(object):
def __init__(self, host, port, timeout):
#url = 'http://111.161.35.198:12210/youku_ghost.html'
url = 'http://%s:%s/youku_ghost.html' % (host, port)
self.ghost = Ghost()
self.session = Session(self.ghost,
wait_timeout=timeout,
plugins_enabled=True)
self.session.open(url)
def parse(self, vid):
try:
res = []
self.session.evaluate('window.getPlayUrl("%s")' % vid)
success, resources = self.session.wait_for_selector('div[id="ck"]')
if success:
ck = self.session.evaluate(
'document.getElementById("ck").innerHTML')
res = ck[0]
except Exception, e:
log.app_log.error(traceback.format_exc())
finally:
content = ''.join([i if ord(i) < 128 else ' ' for i in content])
print(file_name)
with open(file_name, 'wt', encoding='utf-8') as file:
file.write(content)
searching = True
story_buffer = ''
next_url = initial_url
while searching:
try:
session.open(next_url, timeout=300)
lower_bound += 1
session.wait_for_selector(next_button, 60)
except:
break
story_data = session.evaluate(
'document.querySelector("{0}").innerText;'.format(wrapping_div))
text = story_data[0]
story_buffer += str(text)
if lower_bound > upper_bound:
save_story(series_name, lower_bound - chunks, upper_bound,
story_buffer)
story_buffer = ''
lower_bound = upper_bound
upper_bound += chunks
