+- Python Forum (https://python-forum.io)
+-- Forum: Python Coding (https://python-forum.io/forum-7.html)
+--- Forum: Web Scraping & Web Development (https://python-forum.io/forum-13.html)
+--- Thread: Problem With Simple Multiprocessing Script (/thread-9445.html)
Pages:
1
2
Problem With Simple Multiprocessing Script - digitalmatic7 - Apr-09-2018
Is it possible to multiprocess this script without breaking it up into functions?
I'm trying to keep it as barebones and simple as possible.
# EXTREMELY SIMPLE SCRAPING SCRIPT
from time import sleep
from bs4 import BeautifulSoup
import re
import requests
from multiprocessing import Pool
exceptions = []
list1 = ["http://www.wallstreetinvestorplace.com/2018/04/cvs-health-corporation-cvs-to-touch-7-54-earnings-growth-for-next-year/",
"https://macondaily.com/2018/04/06/cetera-advisors-llc-lowers-position-in-cvs-health-cvs.html",
"http://www.thesportsbank.net/football/liverpool/jurgen-klopp-very-positive-about-mo-salah-injury/",
"https://www.moneyjournals.com/trump-wasting-time-trying-bring-amazon/",
"https://www.pmnewsnigeria.com/2018/04/06/fcta-targets-800000-children-for-polio-immunisation/",
"http://toronto.citynews.ca/2018/04/06/officials-in-canada-braced-for-another-spike-in-illegal-border-crossings/",
"https://www.pmnewsnigeria.com/2018/04/04/pdp-describes-looters-list-as-plot-to-divert-attention/",
"https://beyondpesticides.org/dailynewsblog/2018/04/epa-administrator-pruitt-colluding-regulated-industry/",
"http://thyblackman.com/2018/04/06/robert-mueller-is-searching-for/",
"https://www.theroar.com.au/2018/04/06/2018-commonwealth-games-swimming-night-2-finals-live-updates-results-blog/",
"https://medicalresearch.com/pain-research/migraine-linked-to-increased-risk-of-heart-disease-and-stroke/40858/",
"http://www.investingbizz.com/2018/04/amazon-com-inc-amzn-stock-creates-investors-concerns/",
"https://stocknewstimes.com/2018/04/06/convergence-investment-partners-llc-grows-position-in-amazon-com-inc-amzn.html",
"https://factsherald.com/old-food-rules-needs-to-be-updated/",
"https://www.nextadvisor.com/blog/2018/04/06/the-facebook-scandal-evolves/",
"http://sacramento.cbslocal.com/2018/04/04/police-family-youtube-shooter/",
"http://en.brinkwire.com/245768/why-does-stress-lead-to-weight-gain-study-sheds-light/",
"https://www.marijuana.com/news/2018/04/monterey-bud-jeff-sessions-is-on-the-wrong-side-of-history-science-and-public-opinion/",
"http://www.stocksgallery.com/2018/04/06/jpmorgan-chase-co-jpm-noted-a-price-change-of-0-80-and-amazon-com-inc-amzn-closes-with-a-move-of-2-92/",
"https://stocknewstimes.com/2018/04/06/front-barnett-associates-llc-has-2-41-million-position-in-cvs-health-corp-cvs.html",
"http://www.liveinsurancenews.com/colorado-mental-health-insurance-bill-to-help-consumers-navigate-the-system/",
"http://newyork.cbslocal.com/2018/04/04/youtube-headquarters-shooting-suspect/",
"https://ledgergazette.com/2018/04/06/liberty-interactive-co-series-a-liberty-ventures-lvnta-shares-bought-by-brandywine-global-investment-management-llc.html",
"http://bangaloreweekly.com/2018-04-06-city-holding-co-invests-in-cvs-health-corporation-cvs-shares/",
"https://www.thenewsguru.com/didnt-know-lawyer-paid-prostitute-130000-donald-trump/",
"http://www.westlondonsport.com/chelsea/football-wls-conte-gives-two-main-reasons-chelseas-loss-tottenham",
"https://registrarjournal.com/2018/04/06/amazon-com-inc-amzn-shares-bought-by-lenox-wealth-management-inc.html",
"http://www.businessdayonline.com/1bn-eca-withdrawal-commence-action-president-buhari-pdp-tasks-nass/",
"http://www.thesportsbank.net/football/manchester-united/pep-guardiola-asks-for-his-fans-help-vs-united-in-manchester-derby/",
"https://www.pakistantoday.com.pk/2018/04/06/three-palestinians-martyred-as-new-clashes-erupt-along-gaza-border/",
"http://www.nasdaqfortune.com/2018/04/06/risky-factor-of-cvs-health-corporation-cvs-is-observed-at-1-03/",
"https://stocknewstimes.com/2018/04/06/cetera-advisor-networks-llc-decreases-position-in-cvs-health-cvs.html",
"http://nasdaqjournal.com/index.php/2018/04/06/planet-fitness-inc-nyseplnt-do-analysts-think-you-should-buy/",
"http://www.tv360nigeria.com/apc-to-hold-national-congress/",
"https://www.pmnewsnigeria.com/2018/04/03/apc-governors-keep-sealed-lips-after-meeting-with-buhari/",
"https://www.healththoroughfare.com/diet/healthy-lifestyle-best-foods-you-should-eat-for-weight-loss/7061",
"https://stocknewstimes.com/2018/04/05/amazon-com-inc-amzn-shares-bought-by-west-oak-capital-llc.html",
"http://www.current-movie-reviews.com/48428/dr-oz-could-you-be-a-victim-of-sexual-assault-while-on-vacation/",
"https://www.brecorder.com/2018/04/07/410124/world-health-day-to-be-observed-on-april-7/",
"http://www.coloradoindependent.com/169637/trump-pruitt-emissions-epa-pollution",
"https://thecrimereport.org/2018/04/05/will-sessions-new-justice-strategy-turn-the-clock-back-on-civil-rights/",
"http://en.brinkwire.com/245490/pasta-unlikely-to-cause-weight-gain-as-part-of-a-healthy-diet/"]
list_counter = 0
p = Pool(10) # process count
records = p.map(,list1[list_counter]) # argument required
p.terminate()
p.join()
print()
print('Total URLS:', len(list1), "- Starting Task...")
print()
for items in list1:
try:
scrape = requests.get(list1[list_counter],
headers={"user-agent": "Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.90 Safari/537.36"},
timeout=10)
if scrape.status_code == 200:
html = scrape.content
soup = BeautifulSoup(html, 'html.parser')
""" --------------------------------------------- """
# ---------------------------------------------------
''' --> SCRAPE ALEXA RANK: <-- '''
# ---------------------------------------------------
""" --------------------------------------------- """
sleep(0.15)
scrape = requests.get("http://data.alexa.com/data?cli=10&dat=s&url=" + list1[list_counter],
headers={"user-agent": "Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.90 Safari/537.36"})
html = scrape.content
soup = BeautifulSoup(html, 'lxml')
rank = re.findall(r'<popularity[^>]*text="(\d+)"', str(soup))
print("Server Status:", scrape.status_code, '-', u"\u2713", '-', list_counter, '-', list1[list_counter], '-', "Rank:", rank[0])
list_counter = list_counter + 1
else:
print("Server Status:", scrape.status_code)
list_counter = list_counter + 1
pass
except BaseException as e:
exceptions.append(e)
print()
print(e)
print()
list_counter = list_counter + 1
pass
if len(exceptions) > 0:
print("OUTPUT ERROR LOGS:", exceptions)
else:
print("No Errors To Report")
RE: Problem With Simple Multiprocessing Script - nilamo - Apr-09-2018
No, you need to have a callable object that you can pass to the process pool. Or, you can rewrite it so it only handles one url, then store the urls in a different file, and let your operating system handle the multiprocessing with something like
cat urls.txt | parallel my_file.py {} (https://www.gnu.org/software/parallel/).
RE: Problem With Simple Multiprocessing Script - Larz60+ - Apr-09-2018
You should also consider saving your URL list to a file, rather than hard coding.
RE: Problem With Simple Multiprocessing Script - digitalmatic7 - Apr-10-2018
Hmmmm. I got it working.. but.. now I'm so confused..
How does it manage to pull "url" from list1? I was surprised this script actually works lol.
from multiprocessing import Lock, Pool
from time import sleep
from bs4 import BeautifulSoup
import re
import requests
exceptions = []
lock = Lock()
def scraper(url):
"""
Testing multiprocessing and requests
"""
lock.acquire()
try:
scrape = requests.get(url,
headers={"user-agent": "Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.90 Safari/537.36"},
timeout=10)
if scrape.status_code == 200:
""" --------------------------------------------- """
# ---------------------------------------------------
''' --> SCRAPE ALEXA RANK: <-- '''
# ---------------------------------------------------
""" --------------------------------------------- """
sleep(0.1)
scrape = requests.get("http://data.alexa.com/data?cli=10&dat=s&url=" + url,
headers={"user-agent": "Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.90 Safari/537.36"})
html = scrape.content
soup = BeautifulSoup(html, 'lxml')
rank = re.findall(r'<popularity[^>]*text="(\d+)"', str(soup))
print("Server Status:", scrape.status_code, '-', u"\u2713", '-', url, '-', "Rank:", rank[0])
else:
print("Server Status:", scrape.status_code)
pass
except BaseException as e:
exceptions.append(e)
print()
print(e)
print()
pass
finally:
lock.release()
if __name__ == '__main__':
list1 = ["http://www.wallstreetinvestorplace.com/2018/04/cvs-health-corporation-cvs-to-touch-7-54-earnings-growth-for-next-year/",
"https://macondaily.com/2018/04/06/cetera-advisors-llc-lowers-position-in-cvs-health-cvs.html",
"http://www.thesportsbank.net/football/liverpool/jurgen-klopp-very-positive-about-mo-salah-injury/",
"https://www.moneyjournals.com/trump-wasting-time-trying-bring-amazon/",
"https://www.pmnewsnigeria.com/2018/04/06/fcta-targets-800000-children-for-polio-immunisation/",
"http://toronto.citynews.ca/2018/04/06/officials-in-canada-braced-for-another-spike-in-illegal-border-crossings/",
"https://www.pmnewsnigeria.com/2018/04/04/pdp-describes-looters-list-as-plot-to-divert-attention/",
"https://beyondpesticides.org/dailynewsblog/2018/04/epa-administrator-pruitt-colluding-regulated-industry/",
"http://thyblackman.com/2018/04/06/robert-mueller-is-searching-for/",
"https://www.theroar.com.au/2018/04/06/2018-commonwealth-games-swimming-night-2-finals-live-updates-results-blog/",
"https://medicalresearch.com/pain-research/migraine-linked-to-increased-risk-of-heart-disease-and-stroke/40858/",
"http://www.investingbizz.com/2018/04/amazon-com-inc-amzn-stock-creates-investors-concerns/",
"https://stocknewstimes.com/2018/04/06/convergence-investment-partners-llc-grows-position-in-amazon-com-inc-amzn.html",
"https://factsherald.com/old-food-rules-needs-to-be-updated/",
"https://www.nextadvisor.com/blog/2018/04/06/the-facebook-scandal-evolves/",
"http://sacramento.cbslocal.com/2018/04/04/police-family-youtube-shooter/",
"http://en.brinkwire.com/245768/why-does-stress-lead-to-weight-gain-study-sheds-light/",
"https://www.marijuana.com/news/2018/04/monterey-bud-jeff-sessions-is-on-the-wrong-side-of-history-science-and-public-opinion/",
"http://www.stocksgallery.com/2018/04/06/jpmorgan-chase-co-jpm-noted-a-price-change-of-0-80-and-amazon-com-inc-amzn-closes-with-a-move-of-2-92/",
"https://stocknewstimes.com/2018/04/06/front-barnett-associates-llc-has-2-41-million-position-in-cvs-health-corp-cvs.html",
"http://www.liveinsurancenews.com/colorado-mental-health-insurance-bill-to-help-consumers-navigate-the-system/",
"http://newyork.cbslocal.com/2018/04/04/youtube-headquarters-shooting-suspect/",
"https://ledgergazette.com/2018/04/06/liberty-interactive-co-series-a-liberty-ventures-lvnta-shares-bought-by-brandywine-global-investment-management-llc.html",
"http://bangaloreweekly.com/2018-04-06-city-holding-co-invests-in-cvs-health-corporation-cvs-shares/",
"https://www.thenewsguru.com/didnt-know-lawyer-paid-prostitute-130000-donald-trump/",
"http://www.westlondonsport.com/chelsea/football-wls-conte-gives-two-main-reasons-chelseas-loss-tottenham",
"https://registrarjournal.com/2018/04/06/amazon-com-inc-amzn-shares-bought-by-lenox-wealth-management-inc.html",
"http://www.businessdayonline.com/1bn-eca-withdrawal-commence-action-president-buhari-pdp-tasks-nass/",
"http://www.thesportsbank.net/football/manchester-united/pep-guardiola-asks-for-his-fans-help-vs-united-in-manchester-derby/",
"https://www.pakistantoday.com.pk/2018/04/06/three-palestinians-martyred-as-new-clashes-erupt-along-gaza-border/",
"http://www.nasdaqfortune.com/2018/04/06/risky-factor-of-cvs-health-corporation-cvs-is-observed-at-1-03/",
"https://stocknewstimes.com/2018/04/06/cetera-advisor-networks-llc-decreases-position-in-cvs-health-cvs.html",
"http://nasdaqjournal.com/index.php/2018/04/06/planet-fitness-inc-nyseplnt-do-analysts-think-you-should-buy/",
"http://www.tv360nigeria.com/apc-to-hold-national-congress/",
"https://www.pmnewsnigeria.com/2018/04/03/apc-governors-keep-sealed-lips-after-meeting-with-buhari/",
"https://www.healththoroughfare.com/diet/healthy-lifestyle-best-foods-you-should-eat-for-weight-loss/7061",
"https://stocknewstimes.com/2018/04/05/amazon-com-inc-amzn-shares-bought-by-west-oak-capital-llc.html",
"http://www.current-movie-reviews.com/48428/dr-oz-could-you-be-a-victim-of-sexual-assault-while-on-vacation/",
"https://www.brecorder.com/2018/04/07/410124/world-health-day-to-be-observed-on-april-7/",
"http://www.coloradoindependent.com/169637/trump-pruitt-emissions-epa-pollution",
"https://thecrimereport.org/2018/04/05/will-sessions-new-justice-strategy-turn-the-clock-back-on-civil-rights/",
"http://en.brinkwire.com/245490/pasta-unlikely-to-cause-weight-gain-as-part-of-a-healthy-diet/"]
p = Pool(10)
p.map(scraper, list1)
p.terminate()
p.join()
RE: Problem With Simple Multiprocessing Script - nilamo - Apr-10-2018
(Apr-10-2018, 06:40 PM)digitalmatic7 Wrote: p.map(scraper, list1)
map will call the function, scraper, once for each item in the iterable, list1. It might be a method of a Process Pool, but it works very similarly to map: https://docs.python.org/3/library/functions.html#map
RE: Problem With Simple Multiprocessing Script - digitalmatic7 - Apr-10-2018
(Apr-10-2018, 06:55 PM)nilamo Wrote:(Apr-10-2018, 06:40 PM)digitalmatic7 Wrote: p.map(scraper, list1)mapwill call the function,scraper, once for each item in the iterable,list1. It might be a method of a Process Pool, but it works very similarly to map: https://docs.python.org/3/library/functions.html#map
Thanks for the help! I really, really appreciate it. I think I almost have a grasp on it now.
def scraper(url):This is the last part I need some clarification on. url is just some name I made up, yet somehow it cycles through list1 items.
I don't really understand how that happens. Is map passing each individual list item into scraper function, and then it gets named what ever I call it in the function brackets?
RE: Problem With Simple Multiprocessing Script - nilamo - Apr-10-2018
When you call a function, you can pass it parameters. The function decides what variables those parameters are bound to, and what they're named. Nothing outside the function needs to know that it's called a "url", as far as the process.map is concerned, it's just an element of the list.
RE: Problem With Simple Multiprocessing Script - digitalmatic7 - Apr-12-2018
I've run into issues getting a counter to work inside the scraper function.
I just need a very basic counter that increments for each URL (iteration) that is processed. I tried using a global variable and it didn't work. It's assigning a counter to each individual process:
![[Image: LPIpPj58SNy5-jXFfi8y_A.png]](https://image.prntscr.com/image/LPIpPj58SNy5-jXFfi8y_A.png)
I tried passing the variable as an argument but couldn't get it to work.
What you guys think? Is it even possible to have a counter work with multiple processes?
Code here: https://pastebin.com/qnRbdaC2
RE: Problem With Simple Multiprocessing Script - nilamo - Apr-12-2018
You shouldn't use shared state in separate processes. The easy answer is to just tell each function which one it is, something like:
class Counter:
def __init__(self, start=0):
self.value = start
def inc(self, value=1):
self.value += value
return self.value
count = Counter()
p.map(lambda elem: scraper(elem, count.inc()), list1)That way you handle the incrementing before you hand things over to different processes.If you actually want
scraper to keep a global count (...you shouldn't), then you'd need to use some way for the processes to talk to each other, like a queue or this Value thing.
RE: Problem With Simple Multiprocessing Script - digitalmatic7 - Apr-13-2018
I looked over some __init__ tutorials, but I still don't really understand what that does. How would I setup Counter class within the scraper function?
I was playing around with multiprocessing manager (it seems to offer the functionality I need).. but I can't get it to work! Any idea where I'm going wrong?
from multiprocessing import Pool, Manager
def test(current_item, manager):
counter = manager.value(+1)
print(counter)
print(current_item)
if __name__ == '__main__':
list1 = ["item1",
"item2",
"item3",
"item4",
"item5",
"item6",
"item7",
"item8",
"item9",
"item10",
"item11",
"item12"]
manager = Manager()
p = Pool(4) # worker count
p.map(test, list1) # (function, iterable)
p.terminate()
p.join()