def unknown_starttag(self, tag, attrs):
# Called for each unhandled tag, where attrs is a list of (attr, value) tuples
# e.g. for <pre class="screen">, tag="pre", attrs=[("class", "screen")]
# The attr name will be translated to lower case, and quotes in the
# value have been removed and character and entity references
# have been replaced. Starting Python 2.6 all entity references from
# htmlentitydefs are replaced in the attribute values
s = "".join([' %s="%s"' % (key, escape_html(value)) for key, value in attrs])
if self.xhtml_mode and (tag in self.void_elements):
self.pieces.append("<%s%s />" % (tag, s))
else:
self.pieces.append("<%s%s>" % (tag, s))
def get_entry_content(entry):
"""
Select the best content from an entry
"""
candidates = entry.get('content', [])
if candidates:
log.debug('content found for entry %s' % entry.link)
if 'summary_detail' in entry:
log.debug('summary found for entry %s' % entry.link)
candidates.append(entry.summary_detail)
for c in candidates:
if 'html' in c.type: # Match text/html, application/xhtml+xml
return c.type, c.value
else:
# If the content is declared to be (or is determined to be) text/plain,
# it will not be sanitized by Feedparser. This is to avoid data loss.
return c.type, escape_html(c.value)
log.debug('no content found for entry %s' % entry.link)
return '', ''
