XML parse into array in python

Question

I have XML like this:

<?xml version="1.0" ?>
<iq id="123" to="test" type="result">
    <query xmlns="jabber:iq:roster">
        <item jid="foo" subscription="both"/>
        <item jid="bar" subscription="both"/>
    </query>
</iq>

And I would like to parse jid from item into array. I thought something like this would work

import xml.etree.ElementTree as ET

myarr = []

xml = '<?xml version="1.0" ?><iq id="123" to="test" type="result"><query xmlns="jabber:iq:roster"><item jid="foo" subscription="both"/><item jid="bar" subscription="both"/></query></iq>'

root = ET.fromstring(xml)

for item in root.findall('query'):
    t = item.get('jid')
    myarr.append(t)
    print (t)

score 1 · Accepted Answer · edited May 23 '17 at 12:04

1

You need to handle namespaces. One option would to paste the namespace into the xpath expression:

for item in root.findall('.//{%(ns)s}query/{%(ns)s}item' % {'ns': 'jabber:iq:roster'}):
    t = item.attrib.get('jid')
    myarr.append(t)
    print (t)

Prints:

foo
bar

See also:

Parsing XML with namespace in Python via 'ElementTree'

edited May 23 '17 at 12:04

Community

1
1

answered Oct 08 '14 at 22:02

alecxe

462,703
120
1,088
1,195

While it's less general, the iterator could also be simplified to: `root.findall('.//{jabber:iq:roster}item')` – Jonathan Eunice Oct 08 '14 at 22:19
Thanks, I had no experience with namespaces in XML before. Now, it's quite clear – Tomas Bruckner Oct 09 '14 at 11:26

Jonathan Eunice · Answer 2 · 2014-10-31T02:34:24.727

I endorse @alecxe's approach, which I will label "handle the namespaces." That is the most general and correct approach. Unfortunately, namespaces are often ugly, wordy, and they needlessly complexity XPath expressions.

For the many simple cases where namespaces are an artifact of the XML world's desire for über-precision and not truly necessary to identify the nodes in a document, a simpler "eliminate the namespaces" alternative allows more concise searches. The key routine is:

def strip_namespaces(tree):
    """
    Strip the namespaces from an ElementTree in order to make
    processing easier. Adapted from @nonagon's answer
    at http://stackoverflow.com/a/25920989/240490
    """
    for el in tree.iter():
        if '}' in el.tag:
            el.tag = el.tag.split('}', 1)[1]  # strip namespaces
        for k, v in el.attrib.items():
            if '}' in k:
                newkey = k.split('}', 1)[1]
                el.attrib[newkey] = v
            del el.attrib[k]
    return tree

Then the program continues much as before, but without worrying about those pesky namespaces:

root = ET.fromstring(xml)
strip_namespaces(root)

for item in root.findall('.//item'):
    t = item.attrib.get('jid')
    myarr.append(t)
    print (t)

This is not effective if you are trying to modify the ElementTree and re-emit XML, but if you're just trying to deconstruct and grab data from the tree, it works well.

XML parse into array in python

2 Answers2