This module parses and 'normalizes' alerts sent using the Common Alerting Protocol. It currently can handle XML in CAP1.1 and CAP1.2 format, and supports ATOM feeds.

This module originally returned an 'Event' object, but now simply returns a dictionary. It currently does not lookup geocode items or any other items, but please make a request if you want sample code to do this ([email protected]). Alert validation via signatures is also not supported, yet.

You can install it from source with:

$ python setup.py install

You can also install it directly with pip:

$ pip install capparselib

Basic usage includes (from the source directory):

>>> from capparselib.parsers import CAPParser
>>> f = r'test/data/weather.cap'
>>> src = open(f, 'r').read()
>>> alert_list = CAPParser(src).as_dict()

The CAPParser class returns a list of alerts, which are each a dictionary of items. The library provides its own default mapping of the alert fields. For instance, fields with names 'headline' (CAP1.2) and 'title' (CAP1.1) are both renamed to 'cap_headline'. Using the above basic usage example, you can then access fields as needed:

>>> alert = alert_list[0]
>>> alert['cap_sender']
    '[email protected]'
>>> alert.keys()
    ['cap_scope', 'cap_sender', 'cap_note', 'cap_status',
    'cap_id', 'cap_message_type', 'cap_sent', 'cap_info']
>>> alert['cap_info'].keys()
    ['cap_area', 'cap_sender', 'cap_expires', 'cap_severity',
    'cap_event', 'cap_certainty', 'cap_urgency', 'cap_event_code',
    'cap_effective', 'cap_description', 'cap_parameter', 'cap_headline',
    'cap_instruction', 'cap_category']
>>> alert['cap_info']['cap_severity']
    'Severe'

Instead of using the default mapping, you can use your own or those provided extra in the file cap_mappings.py (currently just one extra). Here is how:

>>> from capparselib.parsers import CAPParser
>>> import capparselib.cap_mappings 
>>> f = r'test/data/weather.cap'
>>> src = open(f, 'r').read()
>>> alert_list = CAPParser(src, mappings=cap_mappings.CAP_1_2_MAPPINGS).as_dict()

There are quite a number of tests included with various payloads.

You can run tests quite easily with:

$ pip install -r src/requirements-test.txt 
$ python setup.py test

This will install necessary dependencies, and run the included tests.

Deployment to PyPi should happen automatically when the package version numbers gets incremented and a new tag is added to a commit.

To test a deploy, you can do the following locally (notes for myself):

$ pip install twine
$ inv build package upload-test

Then can test installation with:

$ pip install -i https://test.pypi.org/simple/ capparselib