With this asynchronous (faster) solution you can collect all urls from a site and write them to text files in the same directory as the program is in. These text files can then be used for further crawling or serve as a database, perhaps doing a diff at regular intervals.
The original code has been modified by me to make it work at all and do the things I want. You can increase the number of parallell tasks or whatever you like. Please feel free to publish your modifications here if you come up with some good solutions.
I have some ideas for distributed computing html-javascript-python
#!/usr/bin/env python3
# python 3.5 async web crawler.
# https://github.com/mehmetkose/python3.5-async-crawler
# Licensed under the MIT license:
# http://www.opensource.org/licenses/mit-license
# Copyright (c) 2016 Mehmet Kose mehmet@linux.com
# Copyright () 2017 Drill Bit but I am not so fuzzy about it.
'''
A bot that surfs a site and extracts all links from it. It does not at the moment move to other sites.
It puts in site pages in one text file, external sites in another text-file and left overs such as media files in a separate file.
After the search is complete it cleans the data and removes duplicates which is put in cleaned files. Than a back up is written and the content is rewritten to the original file for the next search. This way a database can increase over time.
You are only given web adresses with this program.
You can either put in content extraction into this program or write another program to read from the list this program generates and extract relevant content. It will really hammer the server so thread I mean async carefully. On a normal laptop, run at maximum four instances on different sites at the same time. It will use CPU. Not so much RAM though.
There are many interesting things one could do with multiple computers, a central web address database and distributed crawling. Any hand held device could help with a scraping collaboration in exchange for access to the total effort. Trade in crawling for access. If you want more info on the idea all I need is a place to cook ( Fear and Loathing in Las Vegas ).
It is sensitive to the root_url.
'''
import aiohttp
import asyncio
import async_timeout
from urllib.parse import urljoin, urldefrag
#calculate execution time. Hours...
import time
start = time.time()
antal_arbetare = 10 #original är 3
#PUT IN THE SITE TO CRAWL HERE
root_url = "http://www.bbc.com"
crawled_urls, url_hub = [], [root_url, "%s/sitemap.xml" % (root_url), "%s/robots.txt" % (root_url)]
#Choose a webbrowser to present to the server.
#headers = {'user-agent': 'Opera/9.80 (X11; Linux x86_64; U; en) Presto/2.2.15 Version/10.10'}
#headers = {'user-agent': 'Firefox/53 (X11; Linux x86_64; U; en) sinGUlaro/5.2.12 Ver/2.8'}
headers = {'user-agent': 'Chrome (Windows 7; B; br) Presto/2.2.15 Version/10.10'}
#headers = {'user-agent': 'Chromium112AD (Windows 10; CZ) Xenialtrouble-0.2.0'}
#headers = {'user-agent': 'Firefox12 (Win XP; build-1076-1) tubular bells'}
#headers = {ADD YOUR OWN}
async def get_body(url):
async with aiohttp.ClientSession() as session:
try:
with async_timeout.timeout(5):
async with session.get(url, headers=headers) as response:
if response.status == 200:
html = await response.text()
return {'error': '', 'html': html}
else:
return {'error': response.status, 'html': ''}
except Exception as err:
return {'error': err, 'html': ''}
async def handle_task(task_id, work_queue):
while not work_queue.empty():
queue_url = await work_queue.get()
if not queue_url in crawled_urls:
a = open('crawledsites.txt','a')
a.write(queue_url+'\n')
crawled_urls.append(queue_url)
body = await get_body(queue_url)
if not body['error']:
for new_url in get_urls(body['html']):
if root_url in new_url and not new_url in crawled_urls:
work_queue.put_nowait(new_url)
#Tests
else:
if root_url.split("//")[-1].split("/")[0].split('www.')[-1] not in new_url:
if 'file:///' in new_url:
wt = open('localfiles.txt','a')
wt.write(new_url+'\n')
wt.close()
else:
if 'adserver' not in new_url and 'facebook' not in new_url and 'tel:' not in new_url and 'file:///' not in new_url and 'googleapis' not in new_url and 'javascript' not in new_url and 'yimg.com' not in new_url and 'btrll.com' not in new_url and 'flickr.com' not in new_url and 'tv.nu' not in new_url and 'klart.se' not in new_url and 'twitter.com' not in new_url and 'linkedin.com' not in new_url and 'facebook.com' not in new_url and 'instagram.com' not in new_url and 'mailto:' not in new_url:
ct = open('externaurl.txt','a')
ct.write(new_url+'\n')
ct.close() #sluttest
else:
erroro = open('notapprovedlinks.txt','a')
erroro.write(queue_url+'\n')
erroro.close() #https://stackoverflow.com/questions/19457227/how-to-print-like-printf-in-python3
def remove_fragment(url):
pure_url, frag = urldefrag(url)
return pure_url
def get_urls(html):
new_urls = [url.split('"')[0] for url in str(html).replace("'",'"').split('href="')[1:]]
return [urljoin(root_url, remove_fragment(new_url)) for new_url in new_urls]
if __name__ == "__main__":
q = asyncio.Queue()
[q.put_nowait(url) for url in url_hub]
loop = asyncio.get_event_loop()
tasks = [handle_task(task_id, q) for task_id in range(antal_arbetare)]
loop.run_until_complete(asyncio.wait(tasks))
loop.close()
#print and log the time
end = time.time()
langd = end-start
logg= open('logg.txt','a')
logg.write(str(root_url)+' - '+str(langd)+' sek - '+str(len(crawled_urls))+'\n')
logg.close()
print(str(root_url)+' '+str(len(crawled_urls)))
#################################################################
# AFTERBIRTH - The miracle of birth #
#################################################################
lines_seen = set() # holds lines already seen
outfile = open('externa-sorterade.txt', "w")
for line in open('externaurl.txt', "r"):
if line not in lines_seen:# not a duplicate
lines_seen.add(line)
outfile.writelines(sorted(lines_seen))
outfile.close()
outfile = open(str(time.localtime()[0])+'_'+str(time.localtime()[2])+'_'+str(time.localtime()[4])+'_externbackupp.txt', 'w')
for line in open('externa-sorterade.txt', 'r'):
outfile.write(line)
outfile.close()
#Gör ny externfil med uppdaterade rader
outfile = open('externaurl.txt', "w")
for line in open('externa-sorterade.txt', "r"):
outfile.write(line)
outfile.close()
##########################################
lines_seen = set() # holds lines already seen
outfile = open('crawledsites-sorterad.txt', "w")
for line in open('crawledsites.txt', "r"):
if line not in lines_seen: # not a duplicate
lines_seen.add(line)
outfile.writelines(sorted(lines_seen))
outfile.close()
outfile = open(str(time.localtime()[0])+'_'+str(time.localtime()[2])+'_'+str(time.localtime()[4])+'_kravlarbackupp.txt', 'w')
for line in open('crawledsites-sorterad.txt', 'r'):
outfile.write(line)
outfile.close()
outfile = open('crawledsites.txt', "w")
for line in open('crawledsites-sorterad.txt', "r"):
outfile.write(line)
outfile.close()
