diff --git a/ox/web/arsenal.py b/ox/web/arsenal.py
new file mode 100644
index 0000000..64112ac
--- /dev/null
+++ b/ox/web/arsenal.py
@@ -0,0 +1,74 @@
+# -*- coding: utf-8 -*-
+# vi:si:et:sw=4:sts=4:ts=4
+from datetime import datetime
+from urllib import urlencode
+import json
+import os
+import re
+
+from ox import find_re, strip_tags, decode_html
+from ox.cache import read_url
+from ox.net import open_url
+
+def get_data(id, language='en'):
+    if language == 'de':
+        url = 'http://films.arsenal-berlin.de/index.php/Detail/Object/Show/object_id/%d/lang/de_DE' % id
+    else:
+        url = 'http://films.arsenal-berlin.de/index.php/Detail/Object/Show/object_id/%d' % id
+    html = read_url(url, unicode=True)
+    if 'ID does not exist' in html:
+        return None
+    if 'Willkommen in der Datenbank des Arsenal' in html:
+        return None
+    data = {}
+    data[u'id'] = id
+    data[u'url'] = url
+    m = re.compile('<h1>(.*?)</h1>').findall(html)
+    if m:
+        data[u'title'] = m[0]
+    m = re.compile("<b>Director: </b><a href='.*?'>(.*?)</a>").findall(html)
+    if m:
+        data[u'director'] = m[0]
+
+    m = re.compile("caUI.initImageScroller\(\[\{url:'(.*?)'").findall(html)
+    if m:
+        data[u'image'] = m[0]
+
+    units = re.compile("<div class='unit'>(.*?)</div>", re.DOTALL).findall(html)
+    for x in map(re.compile('<b>(.*?)</b>: (.*)', re.DOTALL).findall, units):
+        if x:
+            #data[x[0][0].lower()] = strip_tags(x[0][1])
+            key = x[0][0].lower()
+            data[key] = x[0][1]
+            if key == "forum catalogue pdf":
+                data[key] = find_re(data[key], '"(http:.*?)"')
+            else:
+                data[key] = strip_tags(data[key])
+    if "running time (minutes)" in data:
+        data[u'runtime'] = float(data.pop("running time (minutes)").replace(',', '.')) * 60
+    for key in ('year', 'length in metres', 'forum participation year', 'number of reels'):
+        if key in data and data[key].isdigit():
+            data[key] = int(data[key])
+    return data
+
+def backup(filename):
+    if os.path.exists(filename):
+        with open(filename) as f:
+            data = json.load(f)
+    else:
+        data = {}
+    start = ids and max(map(int, data)) or 1
+    for i in range(start, 11872):
+        info = get_data(i)
+        if info:
+            data[i] = info
+            if len(data) % 10 == 0:
+                print 'save', filename, len(data)
+                with open(filename, 'w') as f:
+                    json.dump(data, f)
+        else:
+            print 'ignore', i
+    with open(filename, 'w') as f:
+        json.dump(data, f)
+    return data
+