Develop Reference

Using Python and Selenium to scrape hard-to-find web tables

I've been using Python and Selenium to scrape data from specific state health web pages and output the table to a local CSV.
I've had a lot of success on several other states using similar code. But, I have hit a state that is using what appears to be R to create dynamic dashboards that I can't really access using my normal methods.
I've spent a great deal of time combing through StackOverflow . . . I've checked to see if there's an iframe to switch to, but, I'm just not seeing the data I want located in the iframe on the page.
I can find the table info easy enough using Chrome's "Inspect" feature. But, starting from the original URL, the data I need is not on that page and I can't find a source URL for the the table. I've even used Fiddler to see if there's a call somewhere.
So, I'm not sure what to do. I can see the data--but, I don't know where it is to tell Selenium and BS4 where to access it.
The page is here: https://coronavirus.utah.gov/case-counts/
The page takes a while to load . . . I've had other states have this issue and Selenium could work through it.
The table I need looks like this:
Any help or suggestions would be appreciated.
Here is the code I've been using . . . it doesn't work here, but, the structure is very similar to that which has worked for other states.
import pandas as pd
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.common.by import By
from selenium.webdriver.support import expected_conditions as EC
st = 'ut'
url = 'https://coronavirus.utah.gov/case-counts/'
timeout = 20
# Spawn the webpage using Selenium
driver = webdriver.Chrome(r'D:\Work\Python\utilities\chromedriver\chromedriver.exe')
driver.minimize_window()
driver.get(url)
# Let page load . . . it takes a while
wait = WebDriverWait(driver, timeout).until(EC.visibility_of_element_located()((By.ID, "total-number-of-lab-confirmed-covid-19-cases-living-in-utah")))
# Now, scrape table
html = driver.find_element_by_id("total-number-of-lab-confirmed-covid-19-cases-living-in-utah")
soup = BeautifulSoup(html, 'lxml')
table = soup.find_all('table', id='#DataTables_Table_0')
df = pd.read_html(str(table))
exec(st + "_counts = df[0]")
tmp_str = f"{st}_counts.to_csv(r'D:\Work\Python\projects\Covid_WebScraping\output\{st}_covid_cnts_' + str(datetime.now().strftime('%Y_%m_%d_%H_%M_%S')) + '.csv'"
file_path = tmp_str + ", index=False)"
exec(file_path)
# Close the chrome web driver
driver.close()

I found another way to the get the information I needed.
Thanks to Julian Stanley for letting me know about the Katalon Recorder product. That allowed me to see what the iframe was where the table was.
Using my old method of finding an element by CSS or XPATH was causing a Pickle error due to locked thread. I have no clue how to deal with that . . . but, it caused the entire project to just hang.
But, I was able to get the text/HTML of the table via attribute. After that, I just read it with BS4 as usual.
import pandas as pd
from datetime import datetime
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.support.ui import WebDriverWait
st = 'ut'
url = 'https://coronavirus.utah.gov/case-counts/'
timeout = 20
# Spawn the webpage using Selenium
driver = webdriver.Chrome(r'D:\Work\Python\utilities\chromedriver\chromedriver.exe')
#driver.minimize_window()
driver.get(url)
# Let page load . . . it takes a while
wait = WebDriverWait(driver, timeout)
# Get name of frame (or use index=0)
frames = [frame.get_attribute('id') for frame in driver.find_elements_by_tag_name('iframe')]
# Switch to frame
#driver.switch_to_frame("coronavirus-dashboard")
driver.switch_to_frame(0)
# Now, scrape table
html = driver.find_element_by_css_selector('#DataTables_Table_0_wrapper').get_attribute('innerHTML')
soup = BeautifulSoup(html, 'lxml')
table = soup.find_all('table', id='DataTables_Table_0')
df = pd.read_html(str(table))
exec(st + "_counts = df[0]")
tmp_str = f"{st}_counts.to_csv(r'D:\Work\Python\projects\Covid_WebScraping\output\{st}_covid_cnts_' + str(datetime.now().strftime('%Y_%m_%d_%H_%M_%S')) + '.csv'"
file_path = tmp_str + ", index=False)"
exec(file_path)
# Close the chrome web driver
driver.close()

python parse evernote shared notebook

I am trying to get data from evernote 'shared notebook'.
For example, from this one: https://www.evernote.com/pub/missrspink/evernoteexamples#st=p&n=56b67555-158e-4d10-96e2-3b2c57ee372c
I tried to use Beautiful Soup:
url = 'https://www.evernote.com/pub/missrspink/evernoteexamples#st=p&n=56b67555-158e-4d10-96e2-3b2c57ee372c'
r = requests.get(url)
bs = BeautifulSoup(r.text, 'html.parser')
bs
The result doesn't contain any text information from the notebook, only some code.
I also seen an advice to use selenium and find elements by XPath.
For example I want to find the head of this note - 'Term 3 Week2'. In Google Chrome i found that it's XPath is '/html/body/div[1]/div[1]/b/span/u/b'.
So i tried this:
driver = webdriver.PhantomJS()
driver.get(url)
t = driver.find_element_by_xpath('/html/body/div[1]/div[1]/b/span/u/b')
But it also didn't work, the result was 'NoSuchElementException:... '.
I am a newbie in python and especially parsing, so I would be glad to receive any help.
I am using python 3.6.2 and jupiter-notebook.
Thanks in advance.

The easiest way to interface with Evernote is to use their official Python API.
After you've configured your API key and can generally connect, you can then download and reference Notes and Notebooks.
Evernote Notes use their own template language called ENML (EverNote Markup Language) which is a subset of HTML. You'll be able to use BeautifulSoup4 to parse the ENML and extract the elements you're looking for.
If you're trying to extract information against a local installation (instead of their web app) you may also be able to get what you need from the executable. See how to pass arguments to the local install to extract data. For this you're going to need to use the Python3 subprocess module.
HOWEVER
If you want to use selenium, this will get you started:
import selenium.webdriver.support.ui as ui
from selenium.webdriver import Chrome
from selenium.webdriver.common.by import By
from selenium.webdriver.support import expected_conditions as EC
# your example URL
URL = 'https://www.evernote.com/pub/missrspink/evernoteexamples#st=p&n=56b67555-158e-4d10-96e2-3b2c57ee372c'
# create the browser interface, and a generic "wait" that we can use
# to intelligently block while the driver looks for elements we expect.
# 10: maximum wait in seconds
# 0.5: polling interval in seconds
driver = Chrome()
wait = ui.WebDriverWait(driver, 10, 0.5)
driver.get(URL)
# Note contents are loaded in an iFrame element
find_iframe = By.CSS_SELECTOR, 'iframe.gwt-Frame'
find_html = By.TAG_NAME, 'html'
# .. so we have to wait for the iframe to exist, switch our driver context
# and then wait for that internal page to load.
wait.until(EC.frame_to_be_available_and_switch_to_it(find_iframe))
wait.until(EC.visibility_of_element_located(find_html))
# since ENML is "just" HTML we can select the top tag and get all the
# contents inside it.
doc = driver.find_element_by_tag_name('html')
print(doc.get_attribute('innerHTML')) # <-- this is what you want
# cleanup our browser instance
driver.quit()

Get visible content of a page using selenium and BeautifulSoup

I want to retrieve all visible content of a web page. Let say for example this webpage. I am using a headless firefox browser remotely with selenium.
The script I am using looks like this
driver = webdriver.Remote('http://0.0.0.0:xxxx/wd/hub', desired_capabilities)
driver.get(url)
dom = BeautifulSoup(driver.page_source, parser)
f = dom.find('iframe', id='dsq-app1')
driver.switch_to_frame('dsq-app1')
s = driver.page_source
f.replace_with(BeautifulSoup(s, 'html.parser'))
with open('out.html', 'w') as fe:
fe.write(dom.encode('utf-8'))
This is supposed to load the page, parse the dom, and then replace the iframe with id dsq-app1 with it's visible content. If I execute those commands one by one via my python command line it works as expected. I can then see the paragraphs with all the visible content. When instead I execute all those commands at once, either by executing the script or by pasting all this snippet in my interpreter, it behaves differently. The paragraphs are missing, the content still exists in json format, but it's not what I want.
Any idea why this may happening? Something to do with replace_with maybe?

Sounds like the dom elements are not yet loaded when your code try to reach them.
Try to wait for the elements to be fully loaded and just then replace.
This works for your when you run it command by command because then you let the driver load all the elements before you execute more commands.

To add to Or Duan's answer I provide what I ended up doing. The problem of finding whether a page or parts of a page have loaded completely is an intricate one. I tried to use implicit and explicit waits but again I ended up receiving half-loaded frames. My workaround is to check the readyState of the original document and the readyState of iframes.
Here is a sample function
def _check_if_load_complete(driver, timeout=10):
elapsed_time = 1
while True:
if (driver.execute_script('return document.readyState') == 'complete' or
elapsed_time == timeout):
break
else:
sleep(0.0001)
elapsed_time += 1
then I used that function right after I changed the focus of the driver to the iframe
driver.switch_to_frame('dsq-app1')
_check_if_load_complete(driver, timeout=10)

Try to get the Page Source after detecting the required ID/CSS_SELECTOR/CLASS or LINK.
You can always use explicit wait of Selenium WebDriver.
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.by import By
driver = webdriver.Remote('http://0.0.0.0:xxxx/wd/hub', desired_capabilities)
driver.get(url)
f = WebDriverWait(driver,10).until(EC.presence_of_element_located((By.ID,idName)
# here 10 is time for which script will try to find given id
# provide the id name
dom = BeautifulSoup(driver.page_source, parser)
f = dom.find('iframe', id='dsq-app1')
driver.switch_to_frame('dsq-app1')
s = driver.page_source
f.replace_with(BeautifulSoup(s, 'html.parser'))
with open('out.html', 'w') as fe:
fe.write(dom.encode('utf-8'))
Correct me if this not work

How to retrieve the values of dynamic html content using Python

I'm using Python 3 and I'm trying to retrieve data from a website. However, this data is dynamically loaded and the code I have right now doesn't work:
url = eveCentralBaseURL + str(mineral)
print("URL : %s" % url);
response = request.urlopen(url)
data = str(response.read(10000))
data = data.replace("\\n", "\n")
print(data)
Where I'm trying to find a particular value, I'm finding a template instead e.g."{{formatPrice median}}" instead of "4.48".
How can I make it so that I can retrieve the value instead of the placeholder text?
Edit: This is the specific page I'm trying to extract information from. I'm trying to get the "median" value, which uses the template {{formatPrice median}}
Edit 2: I've installed and set up my program to use Selenium and BeautifulSoup.
The code I have now is:
from bs4 import BeautifulSoup
from selenium import webdriver
#...
driver = webdriver.Firefox()
driver.get(url)
html = driver.page_source
soup = BeautifulSoup(html)
print "Finding..."
for tag in soup.find_all('formatPrice median'):
print tag.text
Here is a screenshot of the program as it's executing. Unfortunately, it doesn't seem to be finding anything with "formatPrice median" specified.

Assuming you are trying to get values from a page that is rendered using javascript templates (for instance something like handlebars), then this is what you will get with any of the standard solutions (i.e. beautifulsoup or requests).
This is because the browser uses javascript to alter what it received and create new DOM elements. urllib will do the requesting part like a browser but not the template rendering part. A good description of the issues can be found here. This article discusses three main solutions:
parse the ajax JSON directly
use an offline Javascript interpreter to process the request SpiderMonkey, crowbar
use a browser automation tool splinter
This answer provides a few more suggestions for option 3, such as selenium or watir. I've used selenium for automated web testing and its pretty handy.
EDIT
From your comments it looks like it is a handlebars driven site. I'd recommend selenium and beautiful soup. This answer gives a good code example which may be useful:
from bs4 import BeautifulSoup
from selenium import webdriver
driver = webdriver.Firefox()
driver.get('http://eve-central.com/home/quicklook.html?typeid=34')
html = driver.page_source
soup = BeautifulSoup(html)
# check out the docs for the kinds of things you can do with 'find_all'
# this (untested) snippet should find tags with a specific class ID
# see: http://www.crummy.com/software/BeautifulSoup/bs4/doc/#searching-by-css-class
for tag in soup.find_all("a", class_="my_class"):
print tag.text
Basically selenium gets the rendered HTML from your browser and then you can parse it using BeautifulSoup from the page_source property. Good luck :)

I used selenium + chrome
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
url = "www.sitetotarget.com"
options = Options()
options.add_argument('--headless')
options.add_argument('--disable-gpu')
options.add_argument('--no-sandbox')
options.add_argument('--disable-dev-shm-usage')`

Building off another answer. I had a similar issue. wget and curl do not work well anymore to get the content of a web page. It's particularly broken with dynamic and lazy content. Using Chrome (or Firefox or Chromium version of Edge) allows you to deal with redirects and scripting.
Below will launch an instance of Chrome, increase the timeout to 5 sec, and navigate this browser instance to a url. I ran this from Jupyter.
import time
from tqdm.notebook import trange, tqdm
from PIL import Image, ImageFont, ImageDraw, ImageEnhance
from selenium import webdriver
driver = webdriver.Chrome('/usr/bin/chromedriver')
driver.set_page_load_timeout(5)
time.sleep(1)
driver.set_window_size(2100, 9000)
time.sleep(1)
driver.set_window_size(2100, 9000)
## You can manually adjust the browser, but don't move it after this.
## Do stuff ...
driver.quit()
Example of grabbing dynamic content and screenshots of the anchored (hence the "a" tag) HTML object, another name for hyperlink:
url = 'http://www.example.org' ## Any website
driver.get(url)
pageSource = driver.page_source
print(driver.get_window_size())
locations = []
for element in driver.find_elements_by_tag_name("a"):
location = element.location;
size = element.size;
# Collect coordinates of object: left/right, top/bottom
x1 = location['x'];
y1 = location['y'];
x2 = location['x']+size['width'];
y2 = location['y']+size['height'];
locations.append([element,x1,y1,x2,y2, x2-x1, y2-y1])
locations.sort(key = lambda x: -x[-2] - x[-1])
locations = [ (el,x1,y1,x2,y2, width,height)
for el,x1,y1,x2,y2,width,height in locations
if not (
## First, filter links that are not visible (located offscreen or zero pixels in any dimension)
x2 <= x1 or y2 <= y1 or x2<0 or y2<0
## Further restrict if you expect the objects to be around a specific size
## or width<200 or height<100
)
]
for el,x1,y1,x2,y2,width,height in tqdm(locations[:10]):
try:
print('-'*100,f'({width},{height})')
print(el.text[:100])
element_png = el.screenshot_as_png
with open('/tmp/_pageImage.png', 'wb') as f:
f.write(element_png)
img = Image.open('/tmp/_pageImage.png')
display(img)
except Exception as err:
print(err)
Installation for mac+chrome:
pip install selenium
brew cask install chromedriver
brew cask install google-chrome
I was using Mac for the original answer and Ubuntu + Windows 11 preview via WSL2 after updating. Chrome ran from Linux side with X service on Windows to render the UI.
Regarding responsibility, please respect robots.txt on each site.

I know this is an old question, but sometimes there is a better solution than using heavy selenium.
This request module for python comes with JS support (in the background it is still chromium) and you can still use beautifulsoup like normal.
Though, sometimes if you have to click elements or sth, I guess selenium is the only option.

python selenium - how can I get all visible text on the page (i.e., not the page source) [duplicate]

I've been googling this all day with out finding the answer, so apologies in advance if this is already answered.
I'm trying to get all visible text from a large number of different websites. The reason is that I want to process the text to eventually categorize the websites.
After a couple of days of research, I decided that Selenium was my best chance. I've found a way to grab all the text, with Selenium, unfortunately the same text is being grabbed multiple times:
from selenium import webdriver
import codecs
filen = codecs.open('outoput.txt', encoding='utf-8', mode='w+')
driver = webdriver.Firefox()
driver.get("http://www.examplepage.com")
allelements = driver.find_elements_by_xpath("//*")
ferdigtxt = []
for i in allelements:
if i.text in ferdigtxt:
pass
else:
ferdigtxt.append(i.text)
filen.writelines(i.text)
filen.close()
driver.quit()
The if condition inside the for loop is an attempt at eliminating the problem of fetching the same text multiple times - it does not however, only work as planned on some webpages. (it also makes the script A LOT slower)
I'm guessing the reason for my problem is that - when asking for the inner text of an element - I also get the inner text of the elements nested inside the element in question.
Is there any way around this? Is there some sort of master element I grab the inner text of? Or a completely different way that would enable me to reach my goal? Any help would be greatly appreciated as I'm out of ideas for this one.
Edit: the reason I used Selenium and not Mechanize and Beautiful Soup is because I wanted JavaScript tendered text

Using lxml, you might try something like this:
import contextlib
import selenium.webdriver as webdriver
import lxml.html as LH
import lxml.html.clean as clean
url="http://www.yahoo.com"
ignore_tags=('script','noscript','style')
with contextlib.closing(webdriver.Firefox()) as browser:
browser.get(url) # Load page
content=browser.page_source
cleaner=clean.Cleaner()
content=cleaner.clean_html(content)
with open('/tmp/source.html','w') as f:
f.write(content.encode('utf-8'))
doc=LH.fromstring(content)
with open('/tmp/result.txt','w') as f:
for elt in doc.iterdescendants():
if elt.tag in ignore_tags: continue
text=elt.text or ''
tail=elt.tail or ''
words=' '.join((text,tail)).strip()
if words:
words=words.encode('utf-8')
f.write(words+'\n')
This seems to get almost all of the text on www.yahoo.com, except for text in images and some text that changes with time (done with javascript and refresh perhaps).

Here's a variation on #unutbu's answer:
#!/usr/bin/env python
import sys
from contextlib import closing
import lxml.html as html # pip install 'lxml>=2.3.1'
from lxml.html.clean import Cleaner
from selenium.webdriver import Firefox # pip install selenium
from werkzeug.contrib.cache import FileSystemCache # pip install werkzeug
cache = FileSystemCache('.cachedir', threshold=100000)
url = sys.argv[1] if len(sys.argv) > 1 else "https://stackoverflow.com/q/7947579"
# get page
page_source = cache.get(url)
if page_source is None:
# use firefox to get page with javascript generated content
with closing(Firefox()) as browser:
browser.get(url)
page_source = browser.page_source
cache.set(url, page_source, timeout=60*60*24*7) # week in seconds
# extract text
root = html.document_fromstring(page_source)
# remove flash, images, <script>,<style>, etc
Cleaner(kill_tags=['noscript'], style=True)(root) # lxml >= 2.3.1
print root.text_content() # extract text
I've separated your task in two:
get page (including elements generated by javascript)
extract text
The code is connected only through the cache. You can fetch pages in one process and extract text in another process or defer to do it later using a different algorithm.