Form scraping using scrapy with Javascript embedded - javascript

I am trying to scrape this website. My spider is functional but the website has javascript embedded in the form to get the result, which I can t get through.
I've read about selenium and how I have to include a browser to get through it but I still with how to scrape after the dynamically generated HTML is loaded, while still passing form arguments.
Here is the code for my spider, any help, referrals or code snippets are welcome. I've navigated and read many threads to no avail.
from scrapy.spiders import BaseSpider
from scrapy.http import FormRequest
from scrapy.http import Request
from scrapy.selector import HtmlXPathSelector
from tax1.items import Tax1Item
from scrapy.contrib.linkextractors.sgml import SgmlLinkExtractor
from scrapy.contrib.spiders import CrawlSpider, Rule
from scrapy.selector import Selector
from selenium import webdriver
import time
class tax1Spider(BaseSpider):
name = "tax1Spider"
allowed_domains = ["http://taxesejour.impots.gouv.fr"]
start_urls = ["http://taxesejour.impots.gouv.fr/DTS_WEB/UK/"]
def parse(self, response):
yield FormRequest.from_response(response,
formname='PAGE_DELIBV2',
formdata={'A8':'05 - Hautes-alpes',
'A10':'AIGUILLES'},
callback = self.parse1)
rules = (
Rule(SgmlLinkExtractor(allow=(), restrict_xpaths=('//div[#class="lh0 dzSpan dzA15"]',)), callback="parse1", follow= True))
# lh0 dzSpan dzA15
def __init__(self):
CrawlSpider.__init__(self)
# use any browser you wish
self.browser = webdriver.Firefox()
def __del__(self):
self.browser.close()
def parse1(self, response):
#hxs = HtmlXPathSelector(response)
self.browser.get(response.url)
time.sleep(3)
Selector(text=self.browser.page_source)
items = []
#item = Tax1Item()
item['message'] = hxs.select('//td[#id="tzA18"]').extract()
print item['message']
return item

Related

Use Python to Parse HTML from dynamic js webpage with qt5 and Beautifulsoup [duplicate]

I am trying to get HTML of a page loaded in PyQT5 QWebEngineView. Here is a simple example:
import sys
from PyQt5.QtCore import *
from PyQt5.QtWebEngineWidgets import *
from PyQt5.QtWidgets import *
def callback_function(html):
print(html)
def on_load_finished():
web.page().runJavaScript("document.getElementsByTagName('html')[0]", callback_function)
app = QApplication(sys.argv)
web = QWebEngineView()
web.load(QUrl("https://stackoverflow.com"))
web.show()
web.loadFinished.connect(on_load_finished)
sys.exit(app.exec_())
I was hoping to be able to return html from the runJavaScript() call but i get a blank in the callback function.
What is incorrect in my code and what alternatives are available for obtaining HTML of a page?

Using my old answer written C++ and translating the solution to Python:
import sys
from PyQt5.QtCore import QUrl
from PyQt5.QtWebEngineWidgets import QWebEngineView
from PyQt5.QtWidgets import QApplication
def callback_function(html):
print(html)
def on_load_finished():
web.page().runJavaScript("document.documentElement.outerHTML", callback_function)
app = QApplication(sys.argv)
web = QWebEngineView()
web.load(QUrl("https://stackoverflow.com"))
web.show()
web.resize(640, 480)
web.loadFinished.connect(on_load_finished)
sys.exit(app.exec_())
Update:
The problem in your case is that getElementsByTagName() returns a list of js elements, and that element cannot be exported to python, what you should do is get the innerHTML:
import sys
from PyQt5.QtCore import QUrl
from PyQt5.QtWebEngineWidgets import QWebEngineView
from PyQt5.QtWidgets import QApplication
def callback_function(html):
print(html)
def on_load_finished():
web.page().runJavaScript(
"document.getElementsByTagName('html')[0].innerHTML", callback_function
)
# or document.getElementsByTagName('html')[0].outerHTML
app = QApplication(sys.argv)
web = QWebEngineView()
web.load(QUrl("https://stackoverflow.com"))
web.show()
web.resize(640, 480)
web.loadFinished.connect(on_load_finished)
sys.exit(app.exec_())

getting dynamic data using python

I'm new to Python and got interested in writing scripts. I'm currently building a crawler that goes on a page and extract copy from tags. Write now I can only list tags; I'm having trouble getting the text out of tags and I'm not sure why exactly. I'm also using BeautifulSoup and PyQt4 to get dynamic data(this might need a new question).
So based on this code below, I should be getting the "Images" copy from the Google homepage, or at least the span tag itself. I'm getting returned NONE
I tried reading the docs for BeautifulSoup and it was a little overwhelming. I'm still reading it, but I think I keep going down a rabbit hole. I can print all anchor tags or all divs, but targeting a specific one is where I'm struggling.
import urllib
import re
from bs4 import BeautifulSoup, Comment
import sys
from PyQt4.QtGui import *
from PyQt4.QtCore import *
from PyQt4.QtWebKit import *
class Render(QWebPage):
def __init__(self, url):
self.app = QApplication(sys.argv)
QWebPage.__init__(self)
self.loadFinished.connect(self._loadFinished)
self.mainFrame().load(QUrl(url))
self.app.exec_()
def _loadFinished(self, result):
self.frame = self.mainFrame()
self.app.quit()
url = 'http://google.com'
source = urllib.urlopen(url).read()
soup = BeautifulSoup(source, 'html.parser')
js_test = soup.find("a", class_="gb_P")
print js_test

How to scrape Javascript rendered data?

I have been trying to scrape this page for a while now and no matter what I try I can't seem to get the table data. The closest I've gotten is the table headers and titles.
I've tried using PhantomJS, Selenium and other methods but I'm stuck.
Site: http://marketwatch.dfm.ae/?isRedirected=true
import sys
from bs4 import BeautifulSoup
from PyQt4.QtGui import *
from PyQt4.QtCore import *
from PyQt4.QtWebKit import *
class Render(QWebPage):
def __init__(self, url):
self.app = QApplication(sys.argv)
QWebPage.__init__(self)
self.loadFinished.connect(self._loadFinished)
self.mainFrame().load(QUrl(url))
self.app.exec_()
def _loadFinished(self, result):
self.frame = self.mainFrame()
self.app.quit()
url = 'http://marketwatch.dfm.ae/?isRedirected=true'
r = Render(url)
soup = BeautifulSoup(unicode(r.frame.toHtml()))
print soup
And I've tried this
import platform
from bs4 import BeautifulSoup
from selenium import webdriver
PHANTOMJS_PATH='./phantomjs'
browser = webdriver.PhantomJS(PHANTOMJS_PATH)
browser.get('http://marketwatch.dfm.ae/')
# let's parse our html
soup = BeautifulSoup(browser.page_source, "html.parser")
# get all the table data
tabdata = soup.find_all('tr', {'id': 'mw'})
print tabdata
Any help is appreciated.
Thank you

How to retrieve the exact HTML as in a browser

I'm using a Python script to render web pages and retrieve their HTML's. It works fine with most of the pages, but with some of them the HTML retrieved is incomplete. And I don't quite understand why. This is the script I'm using to scrap this page, for some reason, the link to every product is not in the HTML:
Link: http://www.pullandbear.com/es/es/mujer/vestidos-c29016.html
Python script:
import sys
from PyQt4.QtGui import *
from PyQt4.QtCore import *
from PyQt4.QtWebKit import *
from PyQt4 import QtNetwork
from PyQt4 import QtCore
url = sys.argv[1]
path = sys.argv[2]
class Render(QWebPage):
def __init__(self, url):
self.app = QApplication(sys.argv)
QWebPage.__init__(self)
self.loadFinished.connect(self._loadFinished)
self.request = QtNetwork.QNetworkRequest()
self.request.setUrl(QtCore.QUrl(url))
self.request.setRawHeader("Accept-Language", QtCore.QByteArray ("es ,*"))
self.mainFrame().load(self.request)
self.app.exec_()
def _loadFinished(self, result):
self.frame = self.mainFrame()
self.app.quit()
r = Render(url)
result = r.frame.toHtml()
html_file = open(path, "w")
html_file.write("%s" % result.encode("utf-8"))
html_file.close()
sys.exit(app.exec_())
This code was taken from here: https://impythonist.wordpress.com/2015/01/06/ultimate-guide-for-scraping-javascript-rendered-web-pages/
Am I missing something? What are the limitations of this framework?
Thanks in advance,

If you want headless browsing you can combine phantomjs with selenium, the following gets all the source:
url = "http://www.pullandbear.com/es/es/mujer/vestidos-c29016.html"
from selenium import webdriver
dr = webdriver.PhantomJS()
dr.get(url)
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
element = WebDriverWait(dr, 5).until(
EC.presence_of_element_located((By.CLASS_NAME, "grid_itemContainer"))
)
Just using selenium without the WebDriverWait did not always return the full source, adding the wait until the a tags with the grid_itemContainer class were visible makes sure the html has been generated, the xpath below returns all your links:
print([a.get_attribute('href') for a in dr.find_elements_by_xpath("//a[#class='grid_itemContainer']")])
[u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-detalle-crochet-pechera-c29016p100064004.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-bordado-escote-pico-c29016p100123006.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-manga-larga-espalda-abierta-c29016p100147503.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-hombros-descubiertos-beads-c29016p100182001.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-jacquard-capa-c29016p100255505.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-vaquero-eyelets-c29016p100336010.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-liso-oversized-c29016p100289013.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-liso-oversized-c29016p100289013.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-camisero-oversized-c29016p100036616.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-pico-c29016p100166506.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-estampado-rayas-c29016p100234507.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-manga-corta-liso-c29016p100262008.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-largo-cuello-halter-liso-c29016p100036162.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-capa-jacquard-%C3%A9tnico-c29016p100259002.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-largo-cuello-halter-rayas-c29016p100036161.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-capa-jacquard-tri%C3%A1ngulo-c29016p100255506.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-marinero-escote-bardot-c29016p100259003.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-rayas-escote-espalda-c29016p100262007.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cruzado-c29016p100216013.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-flores-canes%C3%BA-bordado-c29016p100203011.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-bordados-c29016p100037160.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-flores-volante-c29016p100216014.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-lencero-c29016p100104515.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuadros-detalle-encaje-c29016p100216016.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-drapeado-abertura-bajo-c29016p100129011.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-drapeado-abertura-bajo-c29016p100129011.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-vaquero-bolsillo-plastr%C3%B3n-c29016p100036822.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-rayas-bajo-desigual-c29016p100123010.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-camisero-vaquero-c29016p100036575.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-midi-estampado-rayas-c29016p100189011.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-midi-rayas-manga-3-4-c29016p100149507.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-midi-canal%C3%A9-ajustado-c29016p100149508.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-estampado-bolsillos-c29016p100212503.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-corte-evas%C3%A9-bolsillos-c29016p100189012.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-vaquero-camisero-cuadros-c29016p100036624.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/pichi-vaquero-c29016p100073526.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-estampado-geom%C3%A9trico-cuello-halter-c29016p100037021.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-perkins-manga-larga-c29016p100036882.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-perkins-manga-larga-c29016p100036882.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-perkins-manga-larga-c29016p100036882.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-perkins-manga-larga-c29016p100036882.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-jacquard-evas%C3%A9-c29016p100037207.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cr%C3%AApe-evas%C3%A9-estampado-flores-manga-3-4-c29016p100036932.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cr%C3%AApe-evas%C3%A9-estampado-flores-manga-3-4-c29016p100037280.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-perkins-parche-c29016p100037464.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cr%C3%AApe-evas%C3%A9-liso-manga-3-4-c29016p100036930.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cr%C3%AApe-evas%C3%A9-liso-manga-3-4-c29016p100036930.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-alto-liso-c29016p100037156.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-alto-estampado-flores-c29016p100036921.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-cuello-alto-estampado-corbatero-c29016p100037155.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-largo-manga-sisa-c29016p100170011.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-largo-manga-sisa-rayas-c29016p100170012.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-manga-acampanada-c29016p100149506.html', u'http://www.pullandbear.com/es/es/mujer/vestidos/vestido-punto-espalda-abierta-c29016p100195504.html']
If you want to write the source:
with open("out.html", "w") as f:
f.write(dr.page_source)

I think you can use http://ghost-py.readthedocs.org/en/latest/ for this case. It's loads web page like real browser and run JavaScript.
Also you can try PhantomJS for example, but it written on nodeJS.

Selenium Click() not working with scrapy spider

I am trying to scrape links to product pages from a listing page using a scrapy spider. The page shows the first 10 machines and has a button for 'show all machines' that calls some javascript. The javascript is reasonably complicated (i.e. I can't just look at the function and see the url that the button points to). I'm trying to use the selenium webdriver to simulate a click on the button but it isn't working for some reason. When I scrape the product links I only get the first 10, not the complete list.
Can anybody tell me why it doesn't work?
The page I'm trying to scrape is http://www.ncservice.com/en/second-hand-milling-machines
The spider is
from scrapy.selector import HtmlXPathSelector
from scrapy.contrib.linkextractors.sgml import SgmlLinkExtractor
from scrapy.contrib.spiders import CrawlSpider, Rule
from scrapy.http import Request, FormRequest
from scrapy import log
from scrapy.exceptions import DropItem
from scrapy import signals
from mtispider.items import MachineItem
import urlparse
import time
import MySQLdb
import unicodedata
import re
from mtispider import tools
from selenium import webdriver
class MachineSpider(CrawlSpider):
name = 'nc-spider'
allowed_domains = ['ncservice.com']
def start_requests(self):
requests = list(super(MachineSpider, self).start_requests())
requests.append(Request('http://www.ncservice.com/en/second-hand-milling-machines', callback=self.parsencmilllist))
return requests
def parsencmilllist(self,response):
hxs=HtmlXPathSelector(response)
driver= webdriver.Firefox()
driver.get(response.url)
try:
driver.FindElement(By.Id("mas-resultados-fresadoras")).Click()
except:
log.msg("Couldnt get all the machines", level=log.INFO)
ncmachs = hxs.select('//div[#id="resultados"]//a/#href').extract()
for ncmach in ncmachs:
yield Request(ncmach,
meta = {'type':'Milling'},
callback=self.parsencmachine)
driver.quit()
def parsencmachine(self,response):
#scrape the machine
return item
Thanks!

The main problem is that you need to initialize your Selector from the webdriver's page_source and not the response passed into the callback:
from scrapy.contrib.spiders import CrawlSpider
from scrapy.http import Request
from scrapy import Selector
from selenium import webdriver
class MachineSpider(CrawlSpider):
name = 'nc-spider'
allowed_domains = ['ncservice.com']
def start_requests(self):
yield Request('http://www.ncservice.com/en/second-hand-milling-machines',
callback=self.parsencmilllist)
def parsencmilllist(self, response):
driver = webdriver.Firefox()
driver.get(response.url)
driver.find_element_by_id("mas-resultados-fresadoras").click()
sel = Selector(text=driver.page_source)
driver.quit()
links = sel.xpath('//div[#id="resultados"]//a/#href').extract()
for link in links:
yield Request(link,
meta={'type': 'Milling'},
callback=self.parsencmachine)
def parsencmachine(self, response):
print response.url

Categories