Question

I have the following XML that I am trying to recreate with the lxml.objectify package

<file>
  <customers>
    <customer>
        <phone>
            <type>home</type>
            <number>555-555-5555</number>
        </phone>
        <phone>
            <type>cell</type>
            <number>999-999-9999</number>
        </phone>
        <phone>
            <type>home</type>
            <number>111-111-1111</number>
        </phone>
    </customer>
   </customers>
</file>

I can’t figure out how to create the phone element multiple times. Basically, I have the following non-working code:

    # create phone element 1
    root.customers.customer.phone = ""
    root.customers.customer.phone.type = data_dict['PRIMARY PHONE1']
    root.customers.customer.phone.number = data_dict['PRIMARY PHONE TYPE 1']

    # create phone element 2
    root.customers.customer.phone = ""
    root.customers.customer.phone.type = data_dict['PRIMARY PHONE2']
    root.customers.customer.phone.number = data_dict['PRIMARY PHONE TYPE 2']

    # create phone element 3
    root.customers.customer.phone = ""
    root.customers.customer.phone.type = data_dict['PRIMARY PHONE3']
    root.customers.customer.phone.number = data_dict['PRIMARY PHONE TYPE 3']

Which of course only outputs one section of phone information in the resulting XML. Does anyone have any ideas?

Answer 1

You should create objectify.Element objects, and add them as children of root.customers.

For example, inserting two phone numbers can be done like this:

phone = objectify.Element('phone')
phone.type = data_dict['PRIMARY PHONE1']
phone.number = data_dict['PRIMARY PHONE TYPE 1']
root.customers.customer.append(phone)

phone = objectify.Element('phone')
phone.type = data_dict['PRIMARY PHONE2']
phone.number = data_dict['PRIMARY PHONE TYPE 2']
root.customers.customer.append(phone)

If you get unnecessary attributes on these elements when transforming the xml back to a string, use objectify.deannotate(root, xsi_nil=True, cleanup_namespaces=True).
See lxml’s objectify documentation for exact parameters of objectify.deannotate.

(If you’re using an older version of lxml, which doesn’t include the cleanup_namespaces keyword argument, do this instead:

from lxml import etree
# ...
objectify.deannotate(root, xsi_nil=True)
etree.cleanup_namespaces(root)

)