ixml

================================================= iXML - A simple iterative event-driven XML parser

.. image:: https://secure.travis-ci.org/YAmikep/ixml.png :target: https://travis-ci.org/YAmikep/ixml

.. image:: https://coveralls.io/repos/YAmikep/ixml/badge.png :target: https://coveralls.io/r/YAmikep/ixml

.. image:: https://pypip.in/v/ixml/badge.png :target: https://crate.io/packages/ixml/

.. image:: https://pypip.in/d/ixml/badge.png :target: https://crate.io/packages/ixml/

iXML is an iterative event-driven XML parser with a standard Python iterator interface.

Docs

.. http://ixml.readthedocs.org/en/latest/

Install

From PyPI (stable)::

pip install ixml

From Github (unstable)::

pip install git+git://github.com/YAmikep/ixml.git#egg=ixml

iXML currently requires the lxml <http://lxml.de/>_ library because there is no fallback backend based on the standard library yet.

Main API

• ixml.parse(data): iterator returning parsing events.

• ixml.items(data, path, builder_klass=DictObjectBuilder): iterator returning Python objects found under a specified path.

Notes:

• data must be a file like object.

• The Python objects yielded by ixml.items are constructed from the parsing events by an ObjectBuilder (DictObjectBuilder by default). Please make your own if you wish as long as it implements the ObjectBuilder interface (see ixml.builders.interface).

• Top-level ixml module tries to automatically find and import a suitable parsing backend. You can also explicitly import a required backend from ixml.backends.

Usage and examples

All examples will be using this XML document:

.. code:: python

>>> XML = '''<?xml version="1.0" encoding="utf-8"?>
<cities>
    <city name="Paris">
        <country>France</country>
        <language>French</language>
        <attractions>
            <monument>Eiffel Tower</monument>
            <monument>Triumphal Arch</monument>
            <museum>Louvre Museum</museum>
            <museum>Quai Branly Museum</museum>
        </attractions>          
    </city>
    <city name="Dallas">
        <country>USA</country>
        <language>English</language>
        <attractions>
            <monument>Bank America Plaza</monument>
            <monument>Dallas Theatre Center</monument>
            <museum>Dallas Museum of Art</museum>
            <museum>Old Red Museum</museum>
        </attractions>          
    </city> 
</cities>'''

• ixml.parse

Using the parse function, you can react on individual events:

.. code:: python

>>> import ixml
>>> from StringIO import StringIO

# The parse function takes a file like object
>>> data = StringIO(XML)

# Extract only the languages and the countries
>>> languages, countries = set(), set()
>>> for path, event, value in ixml.parse(data):
...     if path == 'cities.city.language':
...         languages.add(value)
...     elif path == 'cities.city.country':
...         countries.add(value)
>>> print languages, countries
set(['French', 'English']) set(['USA', 'France'])

Below are all the parsing events from parse:

.. code:: python

    ('cities', u'start', None)
    ('cities.city', u'start', None)
    ('cities.city.@name', 'data', 'Paris')
    ('cities.city.country', 'data', 'France')
    ('cities.city.language', 'data', 'French')
    ('cities.city.attractions', u'start', None)
    ('cities.city.attractions.monument', 'data', 'Eiffel Tower')
    ('cities.city.attractions.monument', 'data', 'Triumphal Arch')
    ('cities.city.attractions.museum', 'data', 'Louvre Museum')
    ('cities.city.attractions.museum', 'data', 'Quai Branly Museum')
    ('cities.city.attractions', u'end', None)
    ('cities.city', u'end', None)
    ('cities.city', u'start', None)
    ('cities.city.@name', 'data', 'Dallas')
    ('cities.city.country', 'data', 'USA')
    ('cities.city.language', 'data', 'English')
    ('cities.city.attractions', u'start', None)
    ('cities.city.attractions.monument', 'data', 'Bank America Plaza')
    ('cities.city.attractions.monument', 'data', 'Dallas Theatre Center')
    ('cities.city.attractions.museum', 'data', 'Dallas Museum of Art')
    ('cities.city.attractions.museum', 'data', 'Old Red Museum')
    ('cities.city.attractions', u'end', None)
    ('cities.city', u'end', None)
    ('cities', u'end', None)

• ixml.items

Another usage is having iXML yields native Python objects for a specific path with items:

.. code:: python

>>> import ixml
>>> from StringIO import StringIO

# The items function takes a file like object
>>> data = StringIO(XML)

>>> for city in ixml.items(data, 'cities.city'):
...     do_something_with(city)

Below are the two "city" Python objects created. They are constructed as a dict by default. You can change this behavior by providing another builder class to the items function.

.. code:: python

{   
    'country': 'France', 
    '@name': 'Paris', 
    'language': 'French', 
    'attractions': {
        'museum': ['Louvre Museum', 'Quai Branly Museum'],
        'monument': ['Eiffel Tower', 'Triumphal Arch']
    }
}
{
    'country': 'USA',
    '@name': 'Dallas',
    'language': 'English',
    'attractions': {
        'museum': ['Dallas Museum of Art', 'Old Red Museum'], 
        'monument': ['Bank America Plaza', 'Dallas Theatre Center']
    }
}

Parsing events

Parsing events contain the XML tree context (path), an event and a value::

(path, event, value)

1. The tree context (or path)

It is a simplified path format that:

• uses dots to define different levels
• uses namespace prefixes in the tag name
• ignores default namespaces (handled automatically behind the scene)
• uses @ for attributes

Examples:

• rss.channel.item
• rss.channel.item.@myAttr
• rss.channel.ns1:item.title

1. The events

• "start" and "end" for containers:

.. code:: python

<rss>   # => ('rss', 'start', None)
    <...>
</rss>  # => ('rss', 'end', None)

• "data" for leaves and attributes:

.. code:: python

<rss>   
    <title myAttr="Test">Some text</title>  # => ('rss.title', 'data', 'Some text'), ('rss.title.@myAttr', 'data', 'Test')
</rss>

3. The value

If there is a value, it will always be a string, None otherwise. There is currently no automatic conversion feature (to int, etc).

Backends

iXML can provide several implementation of the parsing by using backends located in ixml/backends:

• lxmliterparse: wrapper around the well known iterparse LXML <http://lxml.de/parsing.html#iterparse-and-iterwalk>_ function.

You can import a specific backend and use it in the same way as the top level library:

.. code:: python

>>> import ixml.backends.lxmliterparse as ixml
>>> for path, event, value in ixml.parse(...):
...     # ... 

Importing the top level library as import ixml tries to import all backends in order.

iXML currently requires the lxml <http://lxml.de/>_ library because there is no fallback backend based on the standard library yet.

ObjectBuilder

The items function uses an ObjectBuilder to build an object while parsing the data.

The events are passed into the event function of the builder that accepts three parameters: path, event type and value. The object being built is available at any time from the value attribute.

You can make your own builder as long as it implements the ObjectBuilder interface (see ixml/builders/interface).

Contribute

Clone and install testing dependencies::

$ python setup.py develop 
$ pip install -r requirements_tests.txt

Ensure tests pass::

$ ./scripts/runtests.sh

Or using tox::

$ tox