XmlModels2¶

XmlModels allows you to define Models similar in nature to Django models that are backed by XML endpoints rather than a database. Using a familiar declarative definition, the fields map to values in the XML document by means of XPath expressions. With support for querying external REST APIs using a django-esque approach, we have strived to make writing and using xml backed models as close to django database models as we can, within the limitations of the available API calls.

User Documentation¶

Mapping XML to Models¶

XML is mapped to xml_models.Model via Fields. Each field requires an xpath expression that determines which node or attribute to get the data from. Each field also has an optional default value for when no value can be retrieved from the XML.

Basic Fields¶

TThe available field mappings are

CharField – returns string data
IntField – returns integers
DateField – returns a date from using the supplied date_format mask or the default ISO8601 format
FloatField – returns a floating point number
BoolField – returns a boolean
OneToOneField – returns a xml_model.Model subclass
CollectionField – returns a collection of either one of the above types, or an xml_model.Model subclass

Most of these fields are fairly self explanatory. The CollectionField and OneToOneField is where it gets interesting. This is what allows you to map instances or collections of nested entities, such as:-

<Person id="112">
  <firstName>Chris</firstName>
  <lastName>Tarttelin</lastName>
  <occupation>Code Geek</occupation>
  <website>http://www.pyruby.com</website>
  <contact-info>
    <contact type="telephone">
      <info>(555) 555-5555</info>
      <description>Cell phone, but no calls during work hours</description>
    </contact>
    <contact type="email">
      <info>me@here.net</info>
      <description>Where possible, contact me by email</description>
    </contact>
    <contact type="telephone">
      <info>1-800-555-5555</info>
      <description>Toll free work number for during office hours.</description>
    </contact>
  </contact-info>
</Person>

This can be mapped using a Person and a ContactInfo model:-

class Person(Model):
  id = IntField(xpath="/Person/@id")
  firstName = CharField(xpath="/Person/firstName")
  lastName = CharField(xpath="/Person/lastName")
  contacts = CollectionField(ContactInfo, order_by="contact_type", xpath="/Person/contact-info/contact")

class ContactInfo(Model):
  contact_type = CharField(xpath="/contact/@type")
  info = CharField(xpath="/contact/info")
  description = CharField(xpath="/contact/description", default="No description supplied")

This leads to the usage of a person as :-

>>> person.contacts[0].info
me@here.com

Collections¶

When querying collections or lists, it is assumed that a collection of zero or more results are returned wrapped in an enclosing collection tag.

As some REST APIs may return lists wrapped in one or more layers of metadata, Models may also define a collection_node attribute. this allows the XML processor to find the relevant node.

Note

collection_node is the tag name only and not an xpath expression.

For example, given the following XML

<reponse status="200">
  <metadata count="2">
  <collection>
    <model ... />
    <model ... />
  </collection>
  </metadata>
</response>

We would need to define a Model with a collection_node like so

class SomeModel(Model):
  fieldA = CharField(xpath="/some/node")

  collection_node = 'collection'

Querying Models¶

Querying is based on Django Model querying. Each Model has a class attribute objects which is the entry point for querying. However the querying ability is limited in scope and only supports very basic filtering.

To do anything interesting you will need to define finders on your models. There are no assumptions made about the nature of the REST API e.g. it is not even assumed that an id attribute can be queried.

Finders¶

An external REST api will present a limited number of options for querying data. Because the different options do not have to follow any specific convention, the model must define what finders are available and what parameters they accept. This still attempts to follow a Django-esque approach

class Person(xml_models.Model:
    ...
    finders = { (firstName, lastName): "http://person/firstName/%s/lastName/%s",
                (id,): "http://person/%s"}

The above defines two query options. The following code exercises these options

>>> people = Person.objects.filter(firstName='Chris', lastName='Tarttelin')
>>> people.count()
1
>>> person = Person.objects.get(id=123)
>>> person.firstName
Chris

Self-signed HTTPS Endpoints¶

Self-signed, or endpoints signed by a non-standard CA, are supported by setting xml_models.VERIFY as per the requests documentation for [SSL certificate verification](http://docs.python-requests.org/en/latest/user/advanced/#ssl-cert-verification).

For example, to not verify a self-signed certificate you can use the following:

API Documentation¶

XmlModels is based on Django database models in terms of functionality and interface.

Manager¶

class xml_models.xml_models.ModelManager(model, finders)¶

Handles what can be queried for, and acts as the entry point for querying.

The API and usage is intended to be familiar to Django users however it does not support the complete Django ModelManager API.

count()¶

Get a count

Returns:	int

filter(**kw)¶

Filter models by key-value pairs.

Example:

Model.objects.filter(attr1=value1,attr2=value2)

How the actual HTTP request is handled is determined by Finders.

Parameters:	kw – key value pairs of field name and value
Returns:	lazy query

filter_custom(url)¶

Set a URL to be called when querying

Parameters:	url – full URL
Returns:	lazy query

get(**kw)¶

Get a single object.

This can be called directly with key-value pairs or after setting some filters

Parameters:	kw –
Returns:

Model¶

class xml_models.xml_models.Model(xml=None, dom=None)¶

A model is a representation of the XML source, consisting of a number of Fields. It can be constructed with either an xml string, or an etree.Element.

Example:

class Person(xml_models.Model):
    namespace="urn:my.default.namespace"
    name = xml_models.CharField(xpath"/Person/@Name", default="John")
    nicknames = xml_models.CollectionField(CharField, xpath="/Person/Nicknames/Name")
    addresses = xml_models.CollectionField(Address, xpath="/Person/Addresses/Address")
    date_of_birth = xml_models.DateField(xpath="/Person/@DateOfBirth", date_format="%d-%m-%Y")

If you define Finders on your model you will also be able to retreive models from an API endpoint using a familiar Django-esque object manager style of access with chainable filtering etc.

to_tree()¶

etree.Element representation of Model

Return type:	`lxml.etree.Element`

to_xml()¶

XML representation of Model

Return type:	string

validate_on_load()¶

Perform validation when the model is instantiated.

Override on your model to perform validation when the XML data is first passed in.

Note

You will need to raise appropriate exceptions as no checking of the return value occurs