Python: Write a new xml file based on a xml template

Question

I want to generate a new xml file (new.xml) based on a xml template (template.xml) using xml.etree.ElementTree. The idea is to change only the value of the <name> tag from 'all' to 'New' leaving the rest of the new.xml file looking exactly as the template.xml. I can change the value of the<name> but the new.xml does not look exactly the same as template.xml

Here is the template.xml:

<?xml version="1.0"?>
<example>
  <version>15.0</version>
  <lastchange/>
  <theme>black</theme>
  <group>
    <name>all</name>
    <description><![CDATA[All Users]]></description>
    <scope>system</scope>
    <gid>1998</gid>
  </group>
</example>

and here is the new.xml:

<example>
  <version>15.0</version>
  <lastchange />
  <theme>black</theme>
  <group>
    <name>New</name>
    <description>All Users</description>
    <scope>system</scope>
    <gid>1998</gid>
  </group>
</example>

As you can notice, in the new.xml the first line is missing and the value of the <description> tag does not have ![CDATA][] structure. This is the script I wrote and I am using:

import xml.etree.ElementTree as ET

def load_xml(name):
    ''' Takes an xml file as input. Outputs ElementTree and element'''
    tree = ET.parse(name)
    root = tree.getroot()
    return tree, root

if __name__ == "__main__":
     # Change and write the new xml
     tree, root = load_xml('template.xml')
     group = root.find('group')
     group.find('name').text = 'New'
     tree.write('new.xml')

Any help? Thank you

`ElementTree` in the core python doesn't seem to support CData section : [1](http://stackoverflow.com/questions/174890/how-to-output-cdata-using-elementtree), [2](http://stackoverflow.com/questions/9027081/lxml-etree-fromsting-and-tostring-are-not-returning-the-same-data). How about switch to `lxml`? — har07, May 09 '16 at 08:39
Is it a problem to get rid of xml declaration and CDATA sections? In the end, it's the same information set. — Vincent Biragnet, May 09 '16 at 08:44
Are you saying that the presence or absence of xml declaration and CDATA section does not change the result? i.e. template.xml and new.xml are interpreted as the same file? — diegus, May 09 '16 at 08:47
@har07 How would I write the same python code I wrote using lxml? thanks — diegus, May 09 '16 at 08:49
I just open the XMLs as pure text and search for the string and substitute sometimes. — curious_weather, May 09 '16 at 08:55
@krork I prefer to use the appropriate library because this is just a part of the much longer xml file I have. — diegus, May 09 '16 at 09:05

score 0 · Accepted Answer · answered May 09 '16 at 09:05

lxml provides compatible API, so you only need to specify strip_cdata=False parameter, and use the exact same codes everywhere else :

form lxml import etree as ET

def load_xml(name):
    ''' Takes an xml file as input. Outputs ElementTree and element'''
    # specify parser setting
    parser = ET.XMLParser(strip_cdata=False)
    # pass parser to do the actual parsing
    tree = ET.parse(name, parser)

    root = tree.getroot()
    return tree, root

if __name__ == "__main__":
     # Change and write the new xml
     tree, root = load_xml('template.xml')
     group = root.find('group')
     group.find('name').text = 'New'
     tree.write('new.xml')

Python: Write a new xml file based on a xml template

1 Answers1