def main(argv=sys.argv[1:]):
parser = argparse.ArgumentParser()
parser.add_argument('--url', default='http://127.0.0.1:8000/static/index.html')
args = parser.parse_args(argv)
url = args.url
browser = WebDriver()
browser.get(url)
tags = browser.find_elements_by_css_selector('li')
for tag in tags:
print(tag.text)
browser.close()
def main(argv=sys.argv[1:]):
parser = argparse.ArgumentParser()
parser.add_argument('--url',
default='http://127.0.0.1:8000/static/index.html')
args = parser.parse_args(argv)
url = args.url
browser = WebDriver()
browser.get(url)
tags = browser.find_elements_by_css_selector('li')
for tag in tags:
print(tag.text)
browser.close()
class Client:
def __init__(self, ig_id):
self.b = PhantomJS()
self.ig_id = ig_id
self.b.get('https://instagram.com/%s' % ig_id)
def close(self):
self.b.close()
def get_media(self) -> list:
js = self.b.execute_script('return window._sharedData;')
ed = js['entry_data']
pp = ed['PostPage'][0]
g = pp['graphql']
sc = g['shortcode_media']
if sc['__typename'] == 'GraphSidecar':
edges = sc['edge_sidecar_to_children']['edges']
medias = list(
map(
lambda x: {
'id': x['node']['id'],
'url': x['node']['display_url'],
'caption': x['node']['accessibility_caption']
}, edges))
elif sc['__typename'] == 'GraphImage':
medias = [{
'id': sc['id'],
'url': sc['display_url'],
'caption': sc['accessibility_caption']
}]
return list(
filter(
lambda x: 'person' in x['caption'] or 'people' in x['caption'],
medias))
def get_user(self) -> dict:
js = self.b.execute_script('return window._sharedData;')
ed = js['entry_data']
pp = ed['ProfilePage'][0]
g = pp['graphql']
return g['user']
def get_posts(self) -> set:
ps = self.b.find_elements_by_css_selector('a[href^="/p/"]')
return set(map(lambda x: x.get_attribute('href'), ps))
def scroll(self):
self.b.execute_script('window.scroll(0, document.body.scrollHeight);')
class WeixinPhantomjs(Base):
all_uids = {docs['uid'] for docs in in_collection.find({}, {'uid': 1}) if 'uid' in docs}
def __init__(self):
self.start_page = START_PAGE
self.end_page = END_PAGE
self.weixin_url = REFER_FIRST
# self.driver = Firefox()
if hasattr(config, 'PHANTOMJS_PATH'):
self.driver = PhantomJS(executable_path=getattr(config, 'PHANTOMJS_PATH'))
else:
self.driver = PhantomJS()
def open_weixin_browser(self, word):
try:
self.driver.get(self.weixin_url)
self.driver.set_page_load_timeout(3)
self.driver.find_element_by_id('upquery').send_keys(word)
self.driver.find_element_by_class_name('swz').click()
time.sleep(3)
urls_uids = self.extract_urls_uids(word=word)
Article(urls_uids=urls_uids, word=word).extract()
except Exception as e:
storage_word.append([word, 0])
self.logger.info('Open weixin error: type <{}>, mag <{}>'.format(e.__class__, e))
self.close_browser()
return True
return False
def get_total_pages_to_word(self):
pages = []
page_id_css = 'pagebar_container'
try:
e = self.driver.find_element_by_id(page_id_css)
for _p in e.text.split():
_p = _p.strip()
if not _p.isdigit():
return DEFAULT_PAGES if DEFAULT_PAGES <= pages[-1] else pages[-1]
else:
pages.append(int(_p))
return 1
except (NoSuchElementException, NoSuchWindowException, TypeError, IndexError):
pass
return 1
def extract_urls_uids(self, word):
urls_uids = []
timestamp = [_t.get_attribute('t') for _t in self.driver.find_elements_by_css_selector('div.s-p')]
urls_tits = [(t.get_attribute('href'), self.trim(t.text))
for t in self.driver.find_elements_by_css_selector('h4 a')]
if len(urls_tits) != len(timestamp):
return urls_uids
for index, url_tit in enumerate(urls_tits):
try:
uid = self.md5(timestamp[index] + url_tit[1] + word)
if uid not in self.__class__.all_uids:
self.__class__.all_uids.add(uid)
urls_uids.append({'url': url_tit[0], 'uid': uid})
except (TypeError, IndexError):
pass
return urls_uids
@property
def is_forbidden(self):
css_id = 'seccodeForm'
try:
if self.driver.find_element_by_id(css_id):
return True
except NoSuchElementException:
pass
return False
def appear_element(self, by):
try:
# Have `click` function to specified element
tag = WebDriverWait(self.driver, 20).until(lambda driver: driver.find_element_by_id(by))
tag.click()
return True
except (TimeoutException, NoSuchWindowException, NoSuchElementException):
pass
return False
def crawl_single(self, word=None, go=0):
is_go = True
go_page = int(go)
next_page_css = 'sogou_page_%s'
is_break = self.open_weixin_browser(word)
pages = self.get_total_pages_to_word()
for page in range(self.start_page + 1, (pages or self.end_page) + 1):
if is_go and page < go_page:
continue
else:
is_go = False
if not self.appear_element(by=next_page_css % page):
is_break = True
msg = '\tNot appear next page element, will break'
elif self.is_forbidden:
is_break = True
msg = '\tSpider was forbidden, crawling again after sleeping a moment!'
if is_break:
storage_word.append([word, page])
self.logger.info(msg)
break
urls_uids = self.extract_urls_uids(word=word)
Article(urls_uids=urls_uids, word=word).extract()
# self.driver.find_element_by_id(next_page_css % page).click()
# wt = randint(10, 40) if page % 5 == 0 else randint(5, 18)
wt = randint(1, 5)
self.logger.info('Word <{}>, Page <{}> Done, sleeping {}s!'.format(word, page, wt))
# self.driver.implicitly_wait(wt)
time.sleep(wt)
self.close_browser()
@classmethod
def crawl_with_threads(cls):
pool = ThreadPool(4)
total_words = QueryWords().get_query_words()
for bulk_words in total_words:
try:
pool.map(lambda w: cls().crawl_single(w), bulk_words)
except Exception as e:
cls.logger.info('Threads crawl error: type <{}>, msg <{}>'.format(e.__class__, e))
pool.close()
pool.join()
in_client.close()
def close_browser(self):
try:
self.driver.close()
except (NoSuchWindowException,):
pass
class WeixinPhantomjs(Base):
def __init__(self):
self.start_page = START_PAGE
self.end_page = END_PAGE
self.weixin_url = REFER_FIRST
# self.driver = Firefox()
if hasattr(config, 'PHANTOMJS_PATH'):
self.driver = PhantomJS(executable_path=getattr(config, 'PHANTOMJS_PATH'))
else:
self.driver = PhantomJS()
self.client = MongoClient(HOST, PORT)
self.collection = self.client[DB][COLLECTION]
self.all_uids = self.uids
def open_weixin_browser(self, word):
try:
self.driver.get(self.weixin_url)
self.driver.set_page_load_timeout(3)
self.driver.find_element_by_id('upquery').send_keys(word)
self.driver.find_element_by_class_name('swz').click()
time.sleep(3)
urls_uids = self.extract_urls_uids(word=word)
Article(urls_uids=urls_uids, word=word).extract()
except Exception as e:
storage_word.append([word, 0])
self.logger.info('Open weixin error: type <{}>, mag <{}>'.format(e.__class__, e))
self.close_browser()
return True
return False
def get_total_pages_to_word(self):
pages = []
page_id_css = 'pagebar_container'
try:
e = self.driver.find_element_by_id(page_id_css)
for _p in e.text.split():
_p = _p.strip()
if not _p.isdigit():
return DEFAULT_PAGES if DEFAULT_PAGES <= pages[-1] else pages[-1]
else:
pages.append(int(_p))
return 1
except (NoSuchElementException, NoSuchWindowException, TypeError, IndexError):
pass
return 1
def get_query_words(self, word):
query_words = []
for docs in self.collection.find({}, {'rel': 1, 'conp': 1}).sort([('_id', 1)]):
w = docs['conp']
if w not in query_words:
query_words.append(w)
for item in docs['rel']:
if item not in query_words:
query_words.append(item)
self.client.close()
return self.query_index(query_words, word)
@property
def uids(self):
return {docs['uid'] for docs in in_collection.find({}, {'uid': 1}) if 'uid' in docs}
def extract_urls_uids(self, word):
urls_uids = []
timestamp = [_t.get_attribute('t') for _t in self.driver.find_elements_by_css_selector('div.s-p')]
urls_tits = [(t.get_attribute('href'), self.trim(t.text))
for t in self.driver.find_elements_by_css_selector('h4 a')]
if len(urls_tits) != len(timestamp):
return urls_uids
for index, url_tit in enumerate(urls_tits):
try:
uid = self.md5(timestamp[index] + url_tit[1] + word)
if uid not in self.all_uids:
self.all_uids.add(uid)
urls_uids.append({'url': url_tit[0], 'uid': uid})
except (TypeError, IndexError):
pass
return urls_uids
@staticmethod
def query_index(words, cut_word):
temp_words = words[START_INDEX:END_INDEX]
try:
index = temp_words.index(cut_word)
return temp_words[index:], index + START_INDEX
except ValueError:
pass
return temp_words, START_INDEX
@property
def is_forbidden(self):
css_id = 'seccodeForm'
try:
if self.driver.find_element_by_id(css_id):
return True
except NoSuchElementException:
pass
return False
def appear_element(self, by):
try:
# Have `click` function to specified element
tag = WebDriverWait(self.driver, 20).until(lambda driver: driver.find_element_by_id(by))
tag.click()
return True
except (TimeoutException, NoSuchWindowException, NoSuchElementException):
pass
return False
def crawl(self, word=None, go=0):
is_go = True
is_break = False
go_page = int(go)
next_page_css = 'sogou_page_%s'
query_words, ind = self.get_query_words(word)
for index, word in enumerate(query_words, 1):
next_ind = ind + index
is_break = self.open_weixin_browser(word)
pages = self.get_total_pages_to_word()
for page in range(self.start_page + 1, (pages or self.end_page) + 1):
if is_go and page < go_page:
continue
else:
is_go = False
if not self.appear_element(by=next_page_css % page):
is_break = True
msg = '\tNot appear next page element, will break'
elif self.is_forbidden:
is_break = True
msg = '\tSpider was forbidden, crawling again after sleeping a moment!'
if is_break:
storage_word.append([word, page])
self.logger.info(msg)
break
urls_uids = self.extract_urls_uids(word=word)
Article(urls_uids=urls_uids, word=word).extract()
# self.driver.find_element_by_id(next_page_css % page).click()
wt = randint(10, 40) if page % 3 == 0 else randint(5, 18)
self.logger.info('Index <{}>, Word <{}>, Page <{}> Done, sleeping {}s!'.format(next_ind, word, page, wt))
# self.driver.implicitly_wait(wt)
time.sleep(wt)
if is_break:
break
in_client.close()
self.close_browser()
def close_browser(self):
try:
self.driver.close()
except (NoSuchWindowException,):
pass
from selenium.webdriver import PhantomJS
import time
import json
dr = PhantomJS()
url = "https://www.v2ex.com/signin"
u = 'username'
p = 'password'
dr.get(url)
account = dr.find_element_by_name('u')
account.clear()
account.send_keys(u)
password = dr.find_element_by_name('p')
password.clear()
password.send_keys(p)
login_btn = dr.find_elements_by_css_selector('.super.normal.button')[1]
login_btn.click()
mission_daily_url = 'https://www.v2ex.com/mission/daily'
dr.get(mission_daily_url)
get_daily_award = dr.find_element_by_css_selector('.super.normal.button')
get_daily_award.click()
balance_url = 'https://www.v2ex.com/balance'
dr.get(balance_url)
intergal = dr.find_elements_by_css_selector('.positive')[0]
yue = dr.find_elements_by_css_selector('.balance_area')[0]
now = time.strftime("%y-%m-%d %H:%M:%S")
item = {"Time": now, "get": intergal.text, "all": yue.text}
with open('v2ex_sign.json', "a") as fp:
fp.write(json.dumps(item))
dr.close()
