3 ###############################################################################
5 # This script assumes the presence of the COVID-19 repo.
7 # It (1) reads in the article list and then (2) calls the Wikimedia API to
8 # fetch view information for each article. Output is to (3) JSON and TSV.
10 ###############################################################################
19 from requests import Request
20 from csv import DictWriter
26 parser = argparse.ArgumentParser(description='Call the views API to collect Wikipedia revision data.')
27 parser.add_argument('-o', '--output_folder', help='Where to save output', default="wikipedia_views/data", type=str)
28 parser.add_argument('-i', '--article_file', help='File listing article names', default="wikipedia_views/resources/enwp_wikiproject_covid19_articles.txt", type=str)
29 parser.add_argument('-d', '--query_date', help='Date if not yesterday, in YYYYMMDD format.', type=str)
30 parser.add_argument('-L', '--logging_level', help='Logging level. Options are debug, info, warning, error, critical. Default: info.', default='info', type=str),
31 parser.add_argument('-W', '--logging_destination', help='Logging destination file. (default: standard error)', type=str),
32 args = parser.parse_args()
38 output_path = args.output_folder
39 article_filename = args.article_file
42 query_date = args.query_date
44 yesterday = datetime.datetime.today() - datetime.timedelta(days=1)
45 query_date = yesterday.strftime("%Y%m%d")
47 query_data = query_date + "00" #requires specifying hours
50 loglevel_mapping = { 'debug' : logging.DEBUG,
51 'info' : logging.INFO,
52 'warning' : logging.WARNING,
53 'error' : logging.ERROR,
54 'critical' : logging.CRITICAL }
56 if args.logging_level in loglevel_mapping:
57 loglevel = loglevel_mapping[args.logging_level]
59 print("Choose a valid log level: debug, info, warning, error, or critical")
63 if args.logging_destination:
64 logging.basicConfig(filename=args.logging_destination, filemode='a', level=loglevel)
66 logging.basicConfig(level=loglevel)
68 export_git_hash = subprocess.check_output(['git', 'rev-parse', 'HEAD']).decode().strip()
69 export_git_short_hash = subprocess.check_output(['git', 'rev-parse', '--short', 'HEAD']).decode().strip()
70 export_time = str(datetime.datetime.now())
72 logging.info(f"Starting run at {export_time}")
73 logging.info(f"Last commit: {export_git_hash}")
75 json_output_filename = os.path.join(output_path, f"digobs_covid19-wikipedia-enwiki_revisions-{query_date}.json")
76 tsv_output_filename = os.path.join(output_path, f"digobs_covid19-wikipedia-enwiki_revisions-{query_data}.tsv")
78 api_session = api.Session("https://en.wikipedia.org/w/api.php")
80 # list of properties from the API we want to gather (basically all of
81 # them supported by mediawik-utilities)
83 rv_props = {'revid' : 'ids',
84 'timestamp' : 'timestamp',
89 'contentmodel' : 'contentmodel',
91 'comment' : 'comment',
92 'content' : 'content' }
94 exclude_from_tsv = ['tags', 'comment', 'content']
96 # load the list of articles
97 with open(article_filename, 'r') as infile:
98 article_list = [art.strip() for art in list(infile)]
100 def get_revisions_for_page(title):
101 return api_session.revisions.query(properties=rv_props.values(),
105 tsv_fields = ['title', 'pageid', 'namespace']
106 tsv_fields = tsv_fields + list(rv_props.keys())
108 # drop fields that we identified for exclusion
109 tsv_fields = [e for e in tsv_fields if e not in exclude_from_tsv]
111 # add special export fields
112 tsv_fields = tsv_fields + ['url', 'export_timestamp', 'export_commit']
114 export_info = { 'git_commit' : export_git_hash,
115 'timestamp' : export_time }
117 with open(json_output_filename, 'w') as json_output, \
118 open(tsv_output_filename, 'w') as tsv_output:
120 tsv_writer = DictWriter(tsv_output, fieldnames=tsv_fields, delimiter="\t")
121 tsv_writer.writeheader()
123 for article in article_list:
124 logging.info(f"pulling revisiosn for: {article}")
125 for rev in get_revisions_for_page(article):
126 logging.debug(f"processing raw revision: {rev}")
128 # add export metadata
129 rev['exported'] = export_info
131 # save the json version of the code
132 print(json.dumps(rev), file=json_output)
134 # handle missing data
135 if "sha1" not in rev:
138 # add page title information
139 rev['title'] = rev['page']['title']
140 rev['pageid'] = rev['page']['pageid']
141 rev['namespace'] = rev['page']['ns']
144 rev['url'] = Request('GET', 'https://en.wikipedia.org/w/index.php',
145 params={'title' : rev['title'].replace(" ", "_"),
146 'oldid' : rev['revid']}).prepare().url
148 rev['export_timestamp'] = export_time
149 rev['export_commit'] = export_git_short_hash
151 tsv_writer.writerow({k: rev[k] for k in tsv_fields})
154 if __name__ == "__main__":