What is this?

The [xml.etree][1] library is a great way to manipulate XML in Python, but in most domains it is clearer to use objects that model your domain entities instead of manipulating the XML representation directly. The motivating example was [Jenkins][2] XML job configurations, but the same problem is likely to exist for any XML-based modeling.

This project attempts to use Python [data descriptors][3] to bind an element tree to Python objects.

[1]: http://lxml.de/tutorial.html [2]: http://jenkins-ci.org/ [3]: http://docs.python.org/2/howto/descriptor.html

Example

Assume that you have an API using XML representations of people with first and last names:

xml = “<person><first>John</first><last>Doe</last></person”

lxmlbind makes it easy to define a Person object that maps to this structure:

from lxmlbind.api import Base, Property

class Person(Base):

first = Property(“first”) last = Property(“last”)

Instances of Person wrap an lxml.etree and support manipulation of this structure through instance attributes:

# create the person from XML person = Person.from_xml(xml)

# get the first name attribute assert person.first == “John”

# set the last name attribute person.last = “Smith”

# print the resulting XML print person.to_xml()

# delete the first name entirely del person.first