X-Git-Url: https://code.communitydata.science/covid19.git/blobdiff_plain/4fe5deb013a899e1410fbdbb957149c5d6f65440..b26a2b5a861494079f7ecaf2c8d2862d1f01b589:/wikipedia/scripts/wikiproject_scraper.py diff --git a/wikipedia/scripts/wikiproject_scraper.py b/wikipedia/scripts/wikiproject_scraper.py index 6e01051..b03b04c 100755 --- a/wikipedia/scripts/wikiproject_scraper.py +++ b/wikipedia/scripts/wikiproject_scraper.py @@ -25,12 +25,13 @@ import logging import re import math from bs4 import BeautifulSoup +import digobs 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('-L', '--logging_level', help='Logging level. Options are debug, info, warning, error, critical. Default: info.', default='info', type=digobs.get_loglevel), parser.add_argument('-W', '--logging_destination', help='Logging destination file. (default: standard error)', type=str), args = parser.parse_args() @@ -41,24 +42,11 @@ 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) + logging.basicConfig(filename=args.logging_destination, filemode='a', level=args.logging_level) else: - logging.basicConfig(level=loglevel) + logging.basicConfig(level=args.logging_level) 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()