Question

Bit of a beginner question here:

Say I have a block of xml:

<root>
 <district>
  <house><room><door/><room></house>
 </district>
 <district>
  <street>
   <house>and so on</house>
  </street>

etc.

With ElementTree I can do:

houses=doc.findall(".//house")

to select all the house nodes, regardless of their parent. What I want to do now is turn each of the house nodes into a separate tree object.

Part of the reason for doing this is that I then want to do another find:

door=houseXml.findall(".//door")

I can do something like:

for _house in houses:
    houseXml=_house.getiterator

but this doesn’t seem to do what I want.

Where am I going wrong?

Answer 1

You can call findall on the elements returned by the first findall:

>>> doc = """<root>
...  <district>
...   <house><room><door/></room></house>
...  </district>
...  <district>
...   <street>
...    <house>and so on</house>
...   </street>
...  </district>
... </root>"""
>>>
>>> from xml.etree import cElementTree as ET
>>>
>>> r = ET.XML(doc)
>>>
>>> for house in r.findall('.//house'):
...   print house, house.findall('.//door')
...
<Element 'house' at 0xb7f3ea70> [<Element 'door' at 0xb7f3eab8>]
<Element 'house' at 0xb7f3eb00> []
>>>