Errors with conditional etree lxml

I'm trying to delete everything between if between is number 66:

开发者_运维问答

I get the following error: TypeError: argument of type 'NoneType' is not iterable...if element.tag == 'answer' and '-66' in element.text:

What is wrong with that? Any help?

#!/usr/local/bin/python2.7
# -*- coding: UTF-8 -*- 

from lxml import etree

planhtmlclear_utf=u"""
<questionaire>
<question>
<questiontext>What's up?</questiontext>
<answer></answer>
</question>
<question>
<questiontext>Cool?</questiontext>
<answer>-66</answer>
</question>
</questionaire>

"""

html = etree.fromstring(planhtmlclear_utf)
questions = html.xpath('/questionaire/question')
for question in questions:
    for element in question.getchildren():
        if element.tag == 'answer' and '-66' in element.text:
            html.xpath('/questionaire')[0].remove(question)
print etree.tostring(html) 

---

element.text seems to be None on some iterations. The error is saying that it cant look through None for "-66", so check that element.text is not None first like this:

html = etree.fromstring(planhtmlclear_utf)
questions = html.xpath('/questionaire/question')
for question in questions:
    for element in question.getchildren():   
        if element.tag == 'answer' and element.text and '-66' in element.text:
            html.xpath('/questionaire')[0].remove(question)
print etree.tostring(html) 

The line its failing at in the xml is <answer></answer> where there is no text in between the tag.

---

Edit (for the second part of your issue about combining tags):

You can use BeautifulSoup like this:

from lxml import etree
import BeautifulSoup

planhtmlclear_utf=u"""
<questionaire>
<question>
<questiontext>What's up?</questiontext>
<answer></answer>
</question>
<question>
<questiontext>Cool?</questiontext>
<answer>-66</answer>
</question>
</questionaire>"""

html = etree.fromstring(planhtmlclear_utf)
questions = html.xpath('/questionaire/question')
for question in questions:
    for element in question.getchildren():   
        if element.tag == 'answer' and element.text and '-66' in element.text:
            html.xpath('/questionaire')[0].remove(question)

soup = BeautifulSoup.BeautifulStoneSoup(etree.tostring(html))
print soup.prettify()

Prints:

<questionaire>
 <question>
  <questiontext>
   What's up?
  </questiontext>
  <answer>
  </answer>
 </question>
</questionaire>

Here is a link where you can download the BeautifulSoup module.

---

Or, to do this a more compact way:

from lxml import etree
import BeautifulSoup    

# abbreviating to reduce answer length...
planhtmlclear_utf=u"<questionaire>.........</questionaire>"

html = etree.fromstring(planhtmlclear_utf)
[question.getparent().remove(question) for question in html.xpath('/questionaire/question[answer/text()="-66"]')]
print BeautifulSoup.BeautifulStoneSoup(etree.tostring(html)).prettify()

---

An alternative to checking if element.text is None, is to refine your XPath:

questions = html.xpath('/questionaire/question[answer/text()="-66"]')
for question in questions:
    question.getparent().remove(question)

The brackets [...] mean "such that". So

question                          # find all question elements
[                                 # such that 
  answer                          # it has an answer subelement
    /text()                       # whose text 
  =                               # equals
  "-66"                           # "-66"
]