Parsing XML file with lxml in Python

Question

I need to parse an xml file, lest say called example.xml, that looks like the following:

<?xml version="1.0" encoding="ISO-8859-1"?>
<nf:rpc-reply xmlns:nf="urn:ietf:params:xml:ns:netconf:base:1.0" xmlns="http://www.cisco.com/nxos:1.0:if_manager">
 <nf:data>
  <show>
   <interface>
    <__XML__OPT_Cmd_show_interface___readonly__>
     <__readonly__>
      <TABLE_interface>
       <ROW_interface>
        <interface>Ethernet1/1</interface>
        <state>down</state>
        <state_rsn_desc>Link not connected</state_rsn_desc>
        <admin_state>up</admin_state>

I need to get "interface", and "state" elements as such: ['Ethernet1/1', 'down'] Here is my solution that doesnt work:

from lxml import etree

parser = etree.XMLParser()
tree = etree.parse('example.xml', parser)

print tree.xpath('//*/*/*/*/*/*/*/*/interface/text()')
print tree.xpath('//*/*/*/*/*/*/*/*/state/text()')

alecxe · Accepted Answer · 2014-07-02T10:06:23.340

4

You need to handle namespaces here:

from lxml import etree

parser = etree.XMLParser()
tree = etree.parse('example.xml', parser)
ns = {'ns': 'http://www.cisco.com/nxos:1.0:if_manager'}

interface = tree.find('//ns:ROW_interface', namespaces=ns)
print [interface.find('.//ns:interface', namespaces=ns).text,
       interface.find('.//ns:state', namespaces=ns).text]

Prints:

['Ethernet1/1', 'down']

Using collections.namedtuple():

interface_node = tree.find('//ns:ROW_interface', ns)

Interface = namedtuple('Interface', ['interface', 'state'])
interface = Interface(interface=interface_node.find('.//ns:interface', ns).text,
                      state=interface_node.find('.//ns:state', ns).text)

print interface

Prints:

Interface(interface='Ethernet1/1', state='down')

edited Jul 02 '14 at 10:06

answered Jul 02 '14 at 02:11

alecxe

462,703
120
1,088
1,195

that works, is it possible to add the extracted data directly to a named tuple instead of a list? like so interafces = ('interfaces'['interface', 'state']) – dragon Jul 02 '14 at 09:42
@dragon great, you can also try to make use of [`lxml.objectify`](http://lxml.de/objectify.html).. – alecxe Jul 02 '14 at 10:10
one last bit, can we also parse a string in the same way rather than a file? – dragon Jul 02 '14 at 11:25
@dragon sure, use `fromstring` instead of `parse`. – alecxe Jul 02 '14 at 11:26

Parsing XML file with lxml in Python

1 Answers1