]> code.communitydata.science - covid19.git/blobdiff - wikipedia/scripts/wikiproject_scraper.py
renamed the wikipedia_views module to wikipedia
[covid19.git] / wikipedia / scripts / wikiproject_scraper.py
diff --git a/wikipedia/scripts/wikiproject_scraper.py b/wikipedia/scripts/wikiproject_scraper.py
new file mode 100755 (executable)
index 0000000..6e01051
--- /dev/null
@@ -0,0 +1,117 @@
+#!/usr/bin/env python3
+
+###############################################################################
+#
+# This script scrapes the Covid-19 Wikiproject
+# 
+# It (1) hits the fcgi to find out how many rounds. Then (2) hit the fcgi 
+# that many rounds, cooking that information down to just a list of article names and
+# then (3) saves it out.
+#
+# At time of writing:
+# the fCGI returns only 1000 max, no matter what you put in the limit. page 1 looks like this....
+# 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
+#
+# and page 2 looks like this
+# 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
+#
+###############################################################################
+
+import argparse
+import subprocess
+import requests
+import datetime
+import logging
+import re
+import math
+from bs4 import BeautifulSoup
+
+def parse_args():
+
+    parser = argparse.ArgumentParser(description='Get a list of pages tracked by the COVID-19 Wikiproject.')
+    parser.add_argument('-o', '--output_file', help='Where to save output', default="wikipedia/resources/enwp_wikiproject_covid19_articles.txt", type=str)
+    parser.add_argument('-L', '--logging_level', help='Logging level. Options are debug, info, warning, error, critical. Default: info.', default='info'), 
+    parser.add_argument('-W', '--logging_destination', help='Logging destination file. (default: standard error)', type=str), 
+    args = parser.parse_args()
+
+    return(args)
+
+def main():
+
+    args = parse_args()
+    outputFile = args.output_file
+
+    #handle -L
+    loglevel_mapping = { 'debug' : logging.DEBUG,
+                         'info' : logging.INFO,
+                         'warning' : logging.WARNING,
+                         'error' : logging.ERROR,
+                         'critical' : logging.CRITICAL }
+
+    if args.logging_level in loglevel_mapping:
+        loglevel = loglevel_mapping[args.logging_level]
+    else:
+        print("Choose a valid log level: debug, info, warning, error, or critical") 
+        exit
+        
+    #handle -W
+    if args.logging_destination:
+        logging.basicConfig(filename=args.logging_destination, filemode='a', level=loglevel)
+    else:
+        logging.basicConfig(level=loglevel)
+
+    export_git_hash = subprocess.check_output(['git', 'rev-parse', 'HEAD']).decode().strip()
+    export_git_short_hash = subprocess.check_output(['git', 'rev-parse', '--short', 'HEAD']).decode().strip()
+    export_time = str(datetime.datetime.now())
+
+    logging.info(f"Starting at {export_time} and destructively outputting article list to {outputFile}.")
+    logging.info(f"Last commit: {export_git_hash}")
+
+    #1 How many hits to the fcgi?
+    session = requests.Session()
+
+    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"
+    headURL = "https://tools.wmflabs.org/enwp10/cgi-bin/list2.fcgi?run=yes&projecta=COVID-19&namespace=&pagename=&quality=&importance=&score=&limit=1000&offset=" 
+    tailURL = "&sorta=Importance&sortb=Quality" #head + offset + tail = original when offset = 1
+
+    # find out how many results we have
+    response = session.get(originalURL)
+
+    soup = BeautifulSoup(response.text, features="html.parser")
+    nodes = soup.find_all('div', class_="navbox")
+    rx = re.compile("Total results:\D*(\d+)") 
+    m = rx.search(nodes[0].get_text())
+    #print(nodes[0].get_text())
+    numResults = int(m.group(1))
+
+    logging.debug(f"fcgi returned {numResults}")
+    rounds = math.ceil(numResults/1000) 
+
+    #2 Fetch and parse down to just the article names
+    articleNames = []
+
+    for i in range(1, rounds+1):
+        offset = (i - 1)*1000 + 1 #offset is 1, then 1001, then 2001 
+        url = f"{headURL}{offset}{tailURL}"
+        response = session.get(url)
+        soup = BeautifulSoup(response.text, features="html.parser") #make fresh soup
+        article_rows = soup.find_all('tr', class_="list-odd") #just the odds first
+        for row in article_rows:
+            a = row.find('a')
+            articleNames.append(a.get_text())
+        article_rows = soup.find_all('tr', class_="list-even") #now the events
+        for row in article_rows:
+            a = row.find('a')
+            articleNames.append(a.get_text())
+
+    #3 Saves the list to a file
+
+    with open(outputFile, 'w') as f:
+        f.write('\n'.join(articleNames)+'\n')
+    logging.debug(f"Finished scrape and made a new article file at {datetime.datetime.now()}")
+
+
+if __name__ == "__main__":
+
+    main()
+

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