]> code.communitydata.science - covid19.git/blob - wikipedia_views/scripts/wikiproject_scraper.py
migrating to new directory structure
[covid19.git] / wikipedia_views / scripts / wikiproject_scraper.py
1 #!/usr/bin/env python3
2
3 ###############################################################################
4 #
5 # This script scrapes the Covid-19 Wikiproject
6
7 # It (1) hits the fcgi to find out how many rounds. Then (2) hit the fcgi 
8 # that many rounds, cooking that information down to just a list of article names and
9 # then (3) saves it out.
10 #
11 # At time of writing:
12 # the fCGI returns only 1000 max, no matter what you put in the limit. page 1 looks like this....
13 # https://tools.wmflabs.org/enwp10/cgi-bin/list2.fcgi?run=yes&projecta=COVID-19&namespace=&pagename=&quality=&importance=&score=&limit=1000&offset=1&sorta=Importance&sortb=Quality
14 #
15 # and page 2 looks like this
16 # https://tools.wmflabs.org/enwp10/cgi-bin/list2.fcgi?namespace=&run=yes&projecta=COVID-19&score=&sorta=Importance&importance=&limit=1000&pagename=&quality=&sortb=Quality&&offset=1001
17 #
18 ###############################################################################
19
20 import argparse
21 import requests
22 import datetime
23 import logging
24 import re
25 import math
26 from bs4 import BeautifulSoup
27
28 def parse_args():
29
30     parser = argparse.ArgumentParser(description='Get a list of pages tracked by the COVID-19 Wikiproject.')
31     parser.add_argument('-o', '--output_folder', help='Where to save output', default="../resources/", type=str)
32     parser.add_argument('-L', '--logging_level', help='Logging level. Options are debug, info, warning, error, critical. Default: info.', default='info'), 
33     parser.add_argument('-W', '--logging_destination', help='Logging destination.', default='../logs/') 
34     args = parser.parse_args()
35
36     return(args)
37
38
39 def main():
40
41     args = parse_args()
42
43     outputPath = args.output_folder
44
45     #handle -W
46     today = datetime.datetime.today().strftime('%Y%m%d')
47     dest = args.logging_destination
48     logHome = f"{dest}scraping{today}"
49
50     #handle -L
51     loglevel = args.logging_level
52     if loglevel == 'debug':
53         logging.basicConfig(filename=logHome, filemode='a', level=logging.DEBUG)
54     elif loglevel == 'info':
55         logging.basicConfig(filename=logHome, filemode='a', level=logging.INFO)
56     elif loglevel == 'warning':
57         logging.basicConfig(filename=logHome, filemode='a', level=logging.WARNING)
58     elif loglevel == 'error':
59         logging.basicConfig(filename=logHome, filemode='a', level=logging.ERROR)
60     elif loglevel == 'critical':
61         logging.basicConfig(filename=logHome, filemode='a', level=logging.CRITICAL)
62     else:
63         print("Choose a valid log level: debug, info, warning, error, or critical")
64         exit
65
66
67     outputFile = f"{outputPath}articles.txt"
68     logging.debug(f"Starting scrape at {datetime.datetime.now()} and destructively outputting article list to {outputFile}.")
69     #1 How many hits to the fcgi?
70
71     #make a session
72
73     session = requests.Session()
74
75     originalURL = "https://tools.wmflabs.org/enwp10/cgi-bin/list2.fcgi?run=yes&projecta=COVID-19&namespace=&pagename=&quality=&importance=&score=&limit=1000&offset=1&sorta=Importance&sortb=Quality"
76     headURL = "https://tools.wmflabs.org/enwp10/cgi-bin/list2.fcgi?run=yes&projecta=COVID-19&namespace=&pagename=&quality=&importance=&score=&limit=1000&offset=" 
77     tailURL = "&sorta=Importance&sortb=Quality" #head + offset + tail = original when offset = 1
78
79     # find out how many results we have
80     response = session.get(originalURL)
81
82     soup = BeautifulSoup(response.text, features="html.parser")
83     nodes = soup.find_all('div', class_="navbox")
84     rx = re.compile("Total results:\D*(\d+)") 
85     m = rx.search(nodes[0].get_text())
86     #print(nodes[0].get_text())
87     numResults = int(m.group(1))
88
89     logging.debug(f"fcgi returned {numResults}")
90     rounds = math.ceil(numResults/1000) 
91
92     #2 Fetch and parse down to just the article names
93     articleNames = []
94
95     for i in range(1, rounds+1):
96         offset = (i - 1)*1000 + 1 #offset is 1, then 1001, then 2001 
97         url = f"{headURL}{offset}{tailURL}"
98         response = session.get(url)
99         soup = BeautifulSoup(response.text, features="html.parser") #make fresh soup
100         article_rows = soup.find_all('tr', class_="list-odd") #just the odds first
101         for row in article_rows:
102             a = row.find('a')
103             articleNames.append(a.get_text())
104         article_rows = soup.find_all('tr', class_="list-even") #now the events
105         for row in article_rows:
106             a = row.find('a')
107             articleNames.append(a.get_text())
108
109     #3 Saves the list to a file
110
111     with open(outputFile, 'w') as f:
112         f.write("Article\n")
113         f.write('\n'.join(articleNames)+'\n')
114     logging.debug(f"Finished scrape and made a new article file at {datetime.datetime.now()}")
115
116
117 if __name__ == "__main__":
118
119     main()
120

Community Data Science Collective || Want to submit a patch?