Don't emit separate open and close tags for empty void elements using htmltmpl

planet/shell/tmpl.py

@@ -1,7 +1,10 @@
 from xml.sax.saxutils import escape
-import sgmllib, time, os, sys, new, urlparse
+import sgmllib, time, os, sys, new, urlparse, re
 from planet import config, feedparser, htmltmpl
 
+voids=feedparser._BaseHTMLProcessor.elements_no_end_tag
+empty=re.compile(r"<((%s)[^>]*)></\2>" % '|'.join(voids))
+
 class stripHtml(sgmllib.SGMLParser):
     "remove all tags from the data"
     def __init__(self, data):
@@ -130,9 +133,12 @@ def tmpl_mapper(source, rules):
         node = source
         for path in rule[2:]:
             if isinstance(path, str) and path in node:
-                if path == 'value' and node.get('type','')=='text/plain':
-                    node['value'] = escape(node['value'])
-                    node['type'] = 'text/html'
+                if path == 'value':
+                    if node.get('type','')=='text/plain':
+                        node['value'] = escape(node['value'])
+                        node['type'] = 'text/html'
+                    elif node.get('type','')=='application/xhtml+xml':
+                        node['value'] = empty.sub(r"<\1 />", node['value'])
                 node = node[path]
             elif isinstance(path, int):
                 node = node[path]

tests/data/filter/tmpl/content_xhtml2.xml

@@ -0,0 +1,13 @@
+<!--
+Description:  xhtml content
+Expect:       Items[0]['content'] == '<img src="x.jpg" />'
+-->
+
+<feed xmlns="http://www.w3.org/2005/Atom">
+  <entry>
+    <content type="xhtml">
+      <div xmlns="http://www.w3.org/1999/xhtml"><img src="x.jpg"/></div>
+    </content>
+  </entry>
+</feed>
+