Python: How can I use lxml objectify's iterchildren to get details of siblings which are in a different namespace

Question

This is my xml file.

get_subscribers_result.xml

<SOAP-ENV:Envelope xmlns:SOAP-ENV="http://schemas.xmlsoap.org/soap/envelope/">
    <SOAP-ENV:Header/>
    <SOAP-ENV:Body>
        <ns3:GetSubscriberResponse xmlns:ns3="http://example.com/123/ss/subscribermgmt/v1_0" xmlns:ns2="http://example.com/123/ss/base/v1_0" xmlns:ns4="http://example.com/123/ss/xyz/v1_0" >
            <ns3:subscriber>
                <ns2:created>2015-10-20T16:02:58.831Z</ns2:created>
                <ns2:createdBy>admin</ns2:createdBy>
                <ns2:lastModified>2015-10-20T16:02:58.824Z</ns2:lastModified>
                <ns2:lastModifiedBy>super</ns2:lastModifiedBy>
                <ns2:subscriberDetail>
                    <ns2:key>address</ns2:key>
                    <ns2:value>1st vivekanandar street</ns2:value>
                </ns2:subscriberDetail>
                <ns2:subscriberDetail>
                    <ns2:key>state</ns2:key>
                    <ns2:value>Abu Dhabi</ns2:value>
                </ns2:subscriberDetail>
            </ns3:subscriber>
        </ns3:GetSubscriberResponse>
    </SOAP-ENV:Body>
</SOAP-ENV:Envelope>

Now, I have seen an example from http://davekuhlman.org/Objectify_files/weather_test.py where iterchildren is used.

The code works if there are no namespaces. This code below would have worked for the xml i have , provided i did not have namespaces in my xml.

    obj2 = lxml.objectify.parse("get_subscribers_result.xml")
    root = obj2.getroot()

    tag = '{http://example.com/123/ss/base/v1_0}subscriberDetail'

    for subscriberDetail in enumerate(root.subscriber.iterchildren(tag=tag)):
                   print subscriberDetail.key
                   print subscriberDetail.value
                   print "*********************************"

If i run this, i get

AttributeError: no such child: {http://schemas.xmlsoap.org/soap/envelope/}subscriber

That's because subscriber doesn't belong to the default namespace and it is correct !

I tried

    for subscriberDetail in enumerate(root.{http://example.com/123/ss/subscribermgmt/v1_0}subscriber.iterchildren(tag=tag)):

Any ideas how to make this work when namespaces are present ?

Alex · Accepted Answer · 2015-11-27T17:09:44.917

1

You can use the "namespace aware" xpath function instead and explicitely specify the namespace:

from lxml import objectify    

obj2 =  lxml.objectify.parse('get_subscribers_result.xml')
root = obj2.getroot()

tag = '{http://example.com/123/ss/base/v1_0}subscriberDetail'

for subscriberDetail in (root.xpath('//ns2:subscriberDetail', namespaces={'ns2': 'http://example.com/123/ss/base/v1_0'})):
               print subscriberDetail.key
               print subscriberDetail.value               
               print "*********************************"

if you want to iterate over all the nodes including their children, you can do something like this:

for details in root.xpath('//SOAP-ENV:Envelope/descendant-or-self::*', namespaces={'SOAP-ENV':'http://schemas.xmlsoap.org/soap/envelope/','ns2': 'http://example.com/123/ss/base/v1_0','ns3':"http://example.com/123/ss/subscribermgmt/v1_0"}):
    for element in details:        
        cleaned_tag = element.tag.replace('{'+element.nsmap[element.prefix]+'}','')      
       if element.text:
           print("%s --> %s" % (element.prefix+':'+cleaned_tag,element.text))

edited Nov 27 '15 at 17:09

answered Nov 27 '15 at 15:25

Alex

21,273
10
61
73

Jaco, Would it be possible to use xpath to dump all the nodes including its children and all siblings by providing details of all the namespaces in the xml file ? – badri Nov 27 '15 at 16:23
1

I have updated my answer with an xpath expression that will iterate over all nodes – Alex Nov 27 '15 at 17:10
There is actually something interesting happening. If i have the same element more than once, your piece of code squares it up in the results. For 12 entries, i would have 144 therefore. Eg: For 2 elements, 1033 1022 has an output as - ns2:operationHistoryId --> 1033 ns2:operationHistoryId --> 1022 ns2:operationHistoryId --> 1033 ns2:operationHistoryId --> 1022 – badri Nov 27 '15 at 17:59
Jaco, i think i found why it was giving weird results. I was using lxml objectify to get the root, whereas i should use etree.parse if i wanted to use xpath ! Think i remember now from a previous documentation that etree and objectify cannot be mixed! – badri Nov 30 '15 at 10:26

Python: How can I use lxml objectify's iterchildren to get details of siblings which are in a different namespace

1 Answers1