class Xvfb(object):
def __init__(self, width=1366, height=768, visible=0):
self.__virtual_display = None
self.width = width
self.height = height
self.visible = visible
def __init_display(self):
if self.__virtual_display is None:
self.__virtual_display = Display(visible=self.visible, size=(self.width, self.height))
self.__virtual_display.start()
def __enter__(self):
self.__init_display()
def __exit__(self, exc_type, exc_val, exc_tb):
self._close_display()
def _close_display(self):
if self.__virtual_display:
try:
self.__virtual_display.close()
except:
pass
self.__virtual_display = None
@staticmethod
def run(func, *args, **kwargs):
runner = Xvfb()
with runner:
return func(*args, **kwargs)
class Spider(scrapy.Spider):
name = "mayors"
allowed_domains = ["www.cec.gov.tw"]
start_urls = ["https://www.cec.gov.tw/pc/zh_TW/IDX/indexC.html",]
download_delay = 1
def __init__(self, ad=None, *args, **kwargs):
super(Spider, self).__init__(*args, **kwargs)
self.display = Display(visible=0, size=(800, 600))
self.display.start()
self.driver = webdriver.Chrome("/var/chromedriver/chromedriver")
def spider_closed(self, spider):
self.display.close()
def parse(self, response):
self.driver.get(response.url)
nodes = scrapy.Selector(text=self.driver.page_source).xpath('//a[@target="_top"]')
for node in nodes:
county = node.xpath('text()').extract_first()
print county
yield response.follow(node, callback=self.parse_list, meta={'meta': county})
def parse_list(self, response):
for tr in response.css(u'table.tableT tr.trT'):
d = {}
d['type'] = 'mayors'
d['county'] = response.meta['meta']
d['constituency'] = 0
d['elected'] = tr.xpath('td[1]/text()').extract_first().strip()
d['number'] = int(tr.xpath('td[2]/text()').extract_first())
d['votes'] = int(re.sub('\D', '', tr.xpath('td[5]/text()').extract_first()))
d['votes_percentage'] = tr.xpath('td[6]/text()').extract_first()
yield d
class Xvfb(object):
def __init__(self, width=1366, height=768, visible=0):
self.__virtual_display = None
self.width = width
self.height = height
self.visible = visible
def __init_display(self):
if self.__virtual_display is None:
self.__virtual_display = Display(visible=self.visible,
size=(self.width, self.height))
self.__virtual_display.start()
def __enter__(self):
self.__init_display()
def __exit__(self, exc_type, exc_val, exc_tb):
self._close_display()
def _close_display(self):
if self.__virtual_display:
try:
self.__virtual_display.close()
except:
pass
self.__virtual_display = None
@staticmethod
def run(func, *args, **kwargs):
runner = Xvfb()
with runner:
return func(*args, **kwargs)
class Spider(scrapy.Spider):
name = "mayors"
allowed_domains = ["www.cec.gov.tw"]
start_urls = [
"https://www.cec.gov.tw/pc/zh_TW/IDX/indexT.html",
]
download_delay = 1
def __init__(self, ad=None, *args, **kwargs):
super(Spider, self).__init__(*args, **kwargs)
self.display = Display(visible=0, size=(800, 600))
self.display.start()
self.driver = webdriver.Chrome("/var/chromedriver/chromedriver")
def spider_closed(self, spider):
self.display.close()
def parse(self, response):
self.driver.get(response.url)
nodes = scrapy.Selector(
text=self.driver.page_source).xpath('//a[@target="_top"]')
for node in nodes:
constituency = int(
re.sub('\D', '',
node.xpath('text()').extract_first()))
yield response.follow(node,
callback=self.parse_list,
meta={'meta': constituency})
def parse_list(self, response):
for tr in response.css(u'table.tableT tr.trT'):
d = {}
d['type'] = 'councilors'
d['county'] = response.xpath(
u'//img[@alt="搜尋結果"]/following-sibling::b[1]/text()'
).extract_first().split()[0]
d['constituency'] = response.meta['meta']
d['elected'] = tr.xpath('td[1]/text()').extract_first().strip()
d['number'] = int(tr.xpath('td[2]/text()').extract_first())
d['votes'] = int(
re.sub('\D', '',
tr.xpath('td[5]/text()').extract_first()))
d['votes_percentage'] = tr.xpath('td[6]/text()').extract_first()
yield d
class Spider(scrapy.Spider):
name = "lis_by_ad"
allowed_domains = ["lis.ly.gov.tw"]
start_urls = [
"http://lis.ly.gov.tw/lylgmeetc/lgmeetkm_lgmem",
]
download_delay = 1
def __init__(self, ad=None, *args, **kwargs):
super(Spider, self).__init__(*args, **kwargs)
self.display = Display(visible=0, size=(800, 600))
self.display.start()
self.driver = webdriver.Chrome("/var/chromedriver/chromedriver")
self.ad = ad
def spider_closed(self, spider):
self.display.close()
def parse(self, response):
yield FormRequest.from_response(
response,
formdata={
'_20_8_T':
str(self.ad).zfill(2),
'INFO':
response.xpath('//input[@name="INFO"]/@value').extract_first()
},
callback=self.parse_max_per_page)
def parse_max_per_page(self, response):
href = response.xpath(
'//select[@onchange="instback(this)"]/option[re:test(text(), "^\d+$")]/@value'
).extract()
yield Request(urljoin(response.url, href[-1]),
callback=self.parse_law_bill_list,
dont_filter=True)
def parse_law_bill_list(self, response):
self.driver.get(response.url)
while (True):
try:
element = WebDriverWait(self.driver, 10).until(
EC.presence_of_element_located((By.ID, "block30")))
except:
continue
sleep(randint(1, 2))
nodes = Selector(
text=self.driver.page_source).xpath('//a[@class="link02"]')
for node in nodes[1::2]:
href = node.xpath('@href').extract_first()
yield Request(urljoin(response.url, href),
callback=self.parse_law_bill,
dont_filter=True)
try:
next_page = self.driver.find_element_by_xpath(
'//input[@name="_IMG_次頁"]')
next_page.click()
except:
break
self.driver.close()
def parse_law_bill(self, response):
trs = response.xpath('//tr[@class="rectr"]')
item = {
tr.xpath('td[1]/nobr/text()').extract_first(): first_or_list(
tr.xpath('td[1]/nobr/text()').extract_first(),
tr.xpath('td[2]//text()').extract())
for tr in trs
}
item.pop(u"關係文書", None) # this one not proper info, parse below
has_motions = response.xpath(
u'//img[@src="/lylegis/images/ref4.png"]/parent::a/@href'
).extract_first()
bill_ref_pdf = response.xpath(
u'//img[@src="/lylgmeet/img/view.png"]/parent::a/@href'
).extract_first()
bill_ref_doc = response.xpath(
u'//img[@src="/lylgmeet/img/doc_icon.png"]/parent::a/@href'
).extract_first()
if bill_ref_pdf:
bill_ref = urljoin(
response.url,
'/lgcgi/lgmeetimage?%s' % bill_ref_pdf.split('^')[-1])
elif bill_ref_doc:
bill_ref = urljoin(response.url, bill_ref_doc)
else:
bill_ref = ''
item['links'] = {
u'關係文書': bill_ref,
u'審議進度':
urljoin(response.url, has_motions) if has_motions else None
}
if has_motions:
yield Request(item['links'][u'審議進度'],
callback=self.parse_law_bill_motions,
dont_filter=True,
meta={'item': item})
else:
item['motions'] = []
yield item
def parse_law_bill_motions(self, response):
item = response.request.meta['item']
motions = []
for node in response.xpath('//tr[@class="onetr0"]/parent::table'):
motion = {}
for tr in node.xpath('.//tr[@class="onetr1"]'):
motion[tr.xpath(
'td[1]/text()').extract_first()] = first_or_list(
tr.xpath('td[2]//text()').extract())
motion.pop(u"影像", None)
motions.append(motion)
item['motions'] = motions
yield item
def getstatement(self):
from selenium import webdriver
from selenium.webdriver.common.keys import Keys
from selenium.webdriver.common.action_chains import ActionChains
from pyvirtualdisplay import Display
from driver_builder import DriverBuilder
"""Setup options for chrome web browser"""
mn = Main()
mid_to_search = mn.get_mid_search()
display = Display(visible=0, size=(800, 600))
display.start()
driver_builder = DriverBuilder()
self.browser = driver_builder.get_driver(dwn, headless=False)
browser = self.browser
browser.get("https://www.youraccessone.com")
username = browser.find_element_by_id("txtUserName")
password = browser.find_element_by_id("txtPassword")
username.send_keys(os.environ["username"])
password.send_keys(os.environ["password"])
browser.find_element_by_name("uxLogin").click()
# go to chargebacks
browser.get("https://www.youraccessone.com/64_rpt_Statements.aspx")
# open filter
browser.find_element_by_xpath("//*[text()='FILTER']").click()
for item in mid_to_search:
# check my merchant
print(item)
for x in range(5):
try:
browser.find_element_by_id(
"ctl00_ContentPage_uxFiltering_uxReportFilter_ctl00"
).click()
break
except NoSuchElementException:
time.sleep(2)
time.sleep(2)
browser.find_element_by_xpath(
"//*[text()='Merchant Number']").click()
time.sleep(.5)
mid = browser.find_element_by_id(
"ctl00_ContentPage_uxFiltering_uxReportFilter_inMERCHANTNUMBER"
)
mid.clear()
mid.send_keys(item)
# Click search
time.sleep(.5)
try:
browser.find_element_by_name(
"ctl00$ContentPage$uxFiltering$uxReportFilter$btSubmit"
).click()
my_val = browser.find_element_by_name(
"ctl00$ContentPage$uxdroplist").get_attribute("value")
my_val = datetime.strptime(my_val, '%m/%d/%Y')
now = datetime.now() - relativedelta(months=1)
print(my_val.month, now.month, my_val.year, now.year)
if my_val.month == now.month and my_val.year == now.year:
browser.find_element_by_name(
"ctl00$ContentPage$uxSearch").click()
except NoSuchElementException:
continue
browser.quit()
try:
display.close()
except AttributeError:
print("The Display is already closed?")
from selenium import webdriver
from pyvirtualdisplay import Display
display = Display(visible=0, size=[800, 600])
display.start()
print("lancement")
driver = webdriver.Firefox()
driver.get("http://www.industrie-expo.com/liste-catalogue-exposants/")
print("")
pageid = 2
while True:
try:
driver.execute_script("searchExposant(" + str(pageid) + ", '#')")
pageid += 1
print(pageid)
except:
break
driver.close()
display.close()
#http://www.marinamele.com/selenium-tutorial-web-scraping-with-selenium-and-python
class Spider(scrapy.Spider):
name = "lis_by_ad"
allowed_domains = ["lis.ly.gov.tw"]
start_urls = [
"http://lis.ly.gov.tw/lylgmeetc/lgmeetkm_lgmem",
]
download_delay = 1
def __init__(self, ad=None, *args, **kwargs):
super(Spider, self).__init__(*args, **kwargs)
self.display = Display(visible=0, size=(800, 600))
self.display.start()
self.driver = webdriver.Chrome("/var/chromedriver/chromedriver")
self.ad = ad
def spider_closed(self, spider):
self.display.close()
def parse(self, response):
yield FormRequest.from_response(
response,
formdata={
'_20_8_T': str(self.ad).zfill(2),
'INFO': response.xpath('//input[@name="INFO"]/@value').extract_first()
},
callback=self.parse_max_per_page
)
def parse_max_per_page(self, response):
href = response.xpath('//select[@onchange="instback(this)"]/option[re:test(text(), "^\d+$")]/@value').extract()
yield Request(urljoin(response.url, href[-1]), callback=self.parse_law_bill_list, dont_filter=True)
def parse_law_bill_list(self, response):
self.driver.get(response.url)
while (True):
try:
element = WebDriverWait(self.driver, 10).until(
EC.presence_of_element_located((By.ID, "block30"))
)
except:
continue
sleep(randint(1, 2))
nodes = Selector(text=self.driver.page_source).xpath('//a[@class="link02"]')
for node in nodes[1::2]:
href = node.xpath('@href').extract_first()
yield Request(urljoin(response.url, href), callback=self.parse_law_bill, dont_filter=True)
try:
next_page = self.driver.find_element_by_xpath('//input[@name="_IMG_次頁"]')
next_page.click()
except:
break
self.driver.close()
def parse_law_bill(self, response):
trs = response.xpath('//tr[@class="rectr"]')
item = {tr.xpath('td[1]/nobr/text()').extract_first(): first_or_list(tr.xpath('td[1]/nobr/text()').extract_first(), tr.xpath('td[2]//text()').extract()) for tr in trs}
item.pop(u"關係文書", None) # this one not proper info, parse below
has_motions = response.xpath(u'//img[@src="/lylegis/images/ref4.png"]/parent::a/@href').extract_first()
bill_ref_pdf = response.xpath(u'//img[@src="/lylgmeet/img/view.png"]/parent::a/@href').extract_first()
bill_ref_doc = response.xpath(u'//img[@src="/lylgmeet/img/doc_icon.png"]/parent::a/@href').extract_first()
if bill_ref_pdf:
bill_ref = urljoin(response.url, '/lgcgi/lgmeetimage?%s' % bill_ref_pdf.split('^')[-1])
elif bill_ref_doc:
bill_ref = urljoin(response.url, bill_ref_doc)
else:
bill_ref = ''
item['links'] = {
u'關係文書': bill_ref,
u'審議進度': urljoin(response.url, has_motions) if has_motions else None
}
if has_motions:
yield Request(item['links'][u'審議進度'], callback=self.parse_law_bill_motions, dont_filter=True, meta={'item': item})
else:
item['motions'] = []
yield item
def parse_law_bill_motions(self, response):
item = response.request.meta['item']
motions = []
for node in response.xpath('//tr[@class="onetr0"]/parent::table'):
motion = {}
for tr in node.xpath('.//tr[@class="onetr1"]'):
motion[tr.xpath('td[1]/text()').extract_first()] = first_or_list(tr.xpath('td[2]//text()').extract())
motion.pop(u"影像", None)
motions.append(motion)
item['motions'] = motions
yield item
