Recovery from bizarrely nested content
This commit is contained in:
parent
299dfeb6cc
commit
33fb5bdd86
@ -158,14 +158,21 @@ def content(xentry, name, detail, bozo):
|
|||||||
for div in body.childNodes:
|
for div in body.childNodes:
|
||||||
if div.nodeType != Node.ELEMENT_NODE: continue
|
if div.nodeType != Node.ELEMENT_NODE: continue
|
||||||
if div.nodeName != 'div': continue
|
if div.nodeName != 'div': continue
|
||||||
div.normalize()
|
try:
|
||||||
if len(div.childNodes) == 1 and \
|
div.normalize()
|
||||||
div.firstChild.nodeType == Node.TEXT_NODE:
|
if len(div.childNodes) == 1 and \
|
||||||
data = div.firstChild
|
div.firstChild.nodeType == Node.TEXT_NODE:
|
||||||
else:
|
data = div.firstChild
|
||||||
data = div
|
else:
|
||||||
xcontent.setAttribute('type', 'xhtml')
|
data = div
|
||||||
break
|
xcontent.setAttribute('type', 'xhtml')
|
||||||
|
break
|
||||||
|
except:
|
||||||
|
# in extremely nested cases, the Python runtime decides
|
||||||
|
# that normalize() must be in an infinite loop; mark
|
||||||
|
# the content as escaped html and proceed on...
|
||||||
|
xcontent.setAttribute('type', 'html')
|
||||||
|
data = xdoc.createTextNode(detail.value.decode('utf-8'))
|
||||||
|
|
||||||
if data: xcontent.appendChild(data)
|
if data: xcontent.appendChild(data)
|
||||||
|
|
||||||
|
34
tests/data/reconstitute/stack_overflow.xml
Normal file
34
tests/data/reconstitute/stack_overflow.xml
Normal file
@ -0,0 +1,34 @@
|
|||||||
|
<!--
|
||||||
|
Description: content with extremely nested markup
|
||||||
|
Expect: content[0].type == 'text/html'
|
||||||
|
-->
|
||||||
|
|
||||||
|
<feed xmlns="http://www.w3.org/2005/Atom">
|
||||||
|
<entry>
|
||||||
|
<content type="html">
|
||||||
|
<![CDATA[
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
<span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span><span>
|
||||||
|
Stack overflow
|
||||||
|
<span><span><span><span><span><span><span><span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span></span>
|
||||||
|
</span></span></span>]]></content>
|
||||||
|
</entry>
|
||||||
|
</feed>
|
Loading…
Reference in New Issue
Block a user