# HG changeset patch
# User Paul Boddie <paul@boddie.org.uk>
# Date 1396383457 -7200
# Node ID f8d989e2f62ff2c58f80e3788a918671262ce74b
# Parent  3605eb85d56f6b34ad18816ea105f035747aefd3
Introduced a separate function for Atom/RSS feed data processing.
Added rejection of "file:" URLs.

diff -r 3605eb85d56f -r f8d989e2f62f MoinShare.py
--- a/MoinShare.py	Mon Mar 31 23:56:56 2014 +0200
+++ b/MoinShare.py	Tue Apr 01 22:17:37 2014 +0200
@@ -182,7 +182,7 @@
 class FeedContentTypeError(FeedError):
     pass
 
-# Feed retrieval from URLs.
+# Update retrieval from URLs.
 
 def getUpdates(request, feed_url, max_entries, show_content):
 
@@ -197,7 +197,10 @@
     returned.
     """
 
-    feed_updates = []
+    # Prevent local file access.
+
+    if feed_url.startswith("file:"):
+        raise FeedMissingError
 
     # Obtain the resource, using a cached version if appropriate.
 
@@ -208,11 +211,34 @@
 
     # Interpret the cached feed.
 
-    feed = StringIO(data)
-    _url, content_type, _encoding, _metadata = getCachedResourceMetadata(feed)
+    f = StringIO(data)
+    try:
+        _url, content_type, _encoding, _metadata = getCachedResourceMetadata(f)
+
+        if content_type in ("application/atom+xml", "application/rss+xml", "application/xml"):
+            return getUpdatesFromFeed(f, max_entries, show_content)
+        else:
+            raise FeedContentTypeError
+
+    finally:
+        f.close()
+
+# Update retrieval from feeds.
 
-    if content_type not in ("application/atom+xml", "application/rss+xml", "application/xml"):
-        raise FeedContentTypeError
+def getUpdatesFromFeed(feed, max_entries, show_content):
+
+    """
+    Retrieve from 'feed' up to the given number 'max_entries' of update entries.
+    The 'show_content' parameter can indicate that a "summary" is to be obtained
+    for each update, that the "content" of each update is to be obtained
+    (falling back to a summary if no content is provided), or no content
+    (indicated by a false value) is to be obtained.
+
+    A tuple of the form ((feed_type, channel_title, channel_link), updates) is
+    returned.
+    """
+
+    feed_updates = []
 
     try:
         # Parse each node from the feed.