Parse xml file with mwxml package

wiki_stats.py

@@ -3,6 +3,7 @@ import re
 import pandas as pd
 from lxml import etree
 import matplotlib.pyplot as plt
+import mwxml  # pip install mwxml
 
 OUTPUT_DIR = 'csv'
 PROJECTS_TEAM = [  # Oderered by first entry
@@ -68,25 +69,21 @@ PROJECTS_TEAM = [  # Oderered by first entry
 ]
 
 
-
 def main(xml_name):
-    ns = {'mediawiki': 'http://www.mediawiki.org/xml/export-0.3/'}
-    root = etree.parse(xml_name)
-    pages = root.xpath('//mediawiki:page', namespaces=ns)
-
     revisions = []
-    for page in pages:
+    with open(xml_name, 'rb') as f:
-        title = page.xpath('./mediawiki:title', namespaces=ns)[0].text
+        pages = mwxml.Dump.from_file(f)
-        for revision in page.xpath('./mediawiki:revision', namespaces=ns):
+        for page in pages:
-            timestamp = revision.xpath('./mediawiki:timestamp',
+            title = page.title
-                                       namespaces=ns)[0].text
+            for revision in page:
-            contributor = revision.getchildren()[2].getchildren()[0].text
+                timestamp = str(revision.timestamp)
-
+                contributor = revision.user.text
-            revisions.append({
+
-                'page': title,
+                revisions.append({
-                'user': contributor,
+                    'page': title,
-                'date': str(timestamp)
+                    'user': contributor,
-            })
+                    'date': timestamp
+                })
 
     df = pd.DataFrame(revisions)
     df['date'] = pd.to_datetime(df['date'])