Parse XML with Python Etree and Return Specified Tag Regardless of Namespace

Question

I am working with some XML data that, in some locations in each file, redefines the namespace. I'm trying to pull all tags of a specific type from the document regardless of the namespace that's active at the point where the tag resides in the XML.

I'm using findall('.//{namespace}Tag') to find the elements I'm looking for. But never knowing what the {namespace} will be at any given point in the file, makes it hit or miss whether I'll get all the requested Tags returned or not.

Is there a way to return all the Tag elements regardless of the {namespace} they fall under? Something along the lines of findall('.//{wildcard}Tag')?

Glabrescent

Given this question hasn't gotten an answer in some time now, here some suggestions. If you have already solved your problem, great! But be sure to also post it here so we don't get a Fermat thread on our hands. If not, a code and XML example of the problem would be handy. – Chip

Answer 1

The xpath function of lxml supports local-name()!

Here is a Python 3 example:

import io
from lxml import etree
xmlstring = '''<root
xmlns:m="http://www.w3.org/html4/"
xmlns:n="http://www.w3.org/html5/">
<m:table>
  <m:tr>
    <m:name>Sometext</m:name>
  </m:tr>
</m:table>
<n:table>
  <n:name>Othertext</n:name>
</n:table>
</root>'''
root = etree.parse(io.StringIO(xmlstring))
names = root.xpath("//*[local-name() = 'name']")
for name in names:
    print(name.text)

Your question might have been aswered before at: lxml etree xmlparser namespace problem

Parkins

what does this output. did you run it? Not sure if this actually works. – Parallelism

I get this error: Traceback (most recent call last): File xml_test.py", line 15, in <module> root = etree.parse(io.StringIO(xmlstring)) TypeError: initial_value must be unicode or None, not str – Parallelism

The actual output is "Sometext\nOthertext\n" – Parkins