amigocloud

python-amigocloud

Python client for the AmigoCloud <https://www.amigocloud.com>__ REST API.

Installation

Install via pip:

::

pip install amigocloud

Dependencies

• |requests|_: Handles the HTTP requests to the AmigoCloud REST API.
• |gevent|_: Handles the websocket connections.
• |socketIO_client|_: Handles the AmigoCloud websocket connection.
• |six|_: A library to assit with python2 to python3 compatibility.

This dependencies will be automatically installed.

Usage

Authentication

This library uses API token to authenticate you. To generate or access your API tokens, go to `API tokens <https://www.amigocloud.com/accounts/tokens>`__.

.. code:: python

    from amigocloud import AmigoCloud
    amigocloud = AmigoCloud(token='R:dlNDEiOWciP3y26kG2cHklYpr2HIPK40HD32r1')

You could also use a project token. Remember that project tokens can only be used to query endpoints relative to the project it belongs to.
If the project URL doesn't match its project, `AmigoCloudError` will be thrown.

.. code:: python

    from amigocloud import AmigoCloud
    amigocloud = AmigoCloud(token='C:Ndl3xGWeasYt9rqyuVsByf5HPMAGyte10y1Mub',
                            project_url='users/123/projects/1234')


You can use a READ token if you only want to do requests that won't alter data. Otherwise, you'll need to use more permissive tokens.

Requests
~~~~~~~~

Once you're logged in you can start making requests to the server. You
can use full urls or relative API urls:

.. code:: python

    # All three will do the same request:
    amigocloud.get('me')
    amigocloud.get('/me')
    amigocloud.get('https://www.amigocloud.com/api/v1/me')

For convenience, when using project tokens, urls are relative to the project's url (unless it starts with `/`):

.. code:: python

    # All three will do the same request:
    amigocloud.get('datasets')
    amigocloud.get('/users/123/projects/1234/datasets')
    amigocloud.get('https://www.amigocloud.com/api/v1/users/123/projects/1234/datasets')

Creating a new AmigoCloud project from Python is as simple as:

.. code:: python

    data = {'name': 'New Project', 'description': 'Created from Python'}
    amigocloud.post('me/projects', data)

All responses are parsed as JSON and return a Python object (usually a
``dict``). This data can be later used in another request:

.. code:: python

    me = amigocloud.get('me')
    visible_projects = amigocloud.get(me['visible_projects'])

    print 'My projects:'
    for project in visible_projects['results']:
        print '*', project['name']

You can get the raw response if you want by using the ``raw`` parameter:

.. code:: python

    me = amigocloud.get('me')
    images = amigocloud.get(me['images'])

    with open('thumbnail.png', 'wb') as thumbnail:
        image_data = amigocloud.get(images['thumbnail'], raw=True)
        thumbnail.write(image_data)


Cursor Requests

Many requests return a paginated list. For example: projects, datasets, base layers, and sql queries. They can be identified when the request returns a dictionary with four items.

.. code:: python

from amigocloud import AmigoCloud
amigocloud = AmigoCloud(token='yourapitoken')

project_list = amigocloud.get('/me/projects')
pprint ( project_list )

will return a dictionary like this (modified for brevity):

.. code:: javascript

{
    u'count': 319,
    u'next': u'https://app.amigocloud.com/api/v1/me/projects?limit=20&offset=20&token=yourapitoken',
    u'previous': None,
    u'results': [] 
} 

From the results, you can see that this endpoint can be iterated through. To make it easier to iterate through these lists, you can use the get_cursor function. The cursor iterates over the results and if it reaches the limit of the response it will automatically make a request to get the next values. So you can get all data and iterate over it, without worrying about the pagination.

.. code:: python

projects = amigocloud.get_cursor('/me/projects')
for project in projects:
    print('Project:', project['name'])

If you want to iterate one request at a time it can be requested as:

.. code:: python

# using a project token to authenticate

datasets = amigocloud.get_cursor('datasets')

dataset1 = datasets.next()
print('Dataset1:', dataset1['name'])

# Boolean to ask if there is a next value.
# otherwise a StopIteration exception is raised.
if datasets.has_next:
    dataset2 = datasets.next()
    print('Dataset2:', dataset2['name'])

Also, you can request some extra values, that are included in the response.

.. code:: python

dataset_rows = amigocloud.get_cursor(
    'https://www.amigocloud.com/api/v1/projects/1234/sql',
    {'query': 'select * from dataset_1'})

print('Response extra values:', dataset_rows.get('columns'))

for row in dataset_rows:
    print('Row:', row)

Cursors can be used for Projects, Datasets, BaseLayers, SQL queries, etc. It also supports non-iterable responses. For this cases it returns only one result.

.. code:: python

cursor = amigocloud.get_cursor('me')

for me in cursor:
    print('Me:', me)

Websocket connection

The websocket connection is started when the AmigoCloud object is
instantiated, and it is closed when the object is destroyed. You always
need to use a user token for websockets.

Make sure to read `our help page about our websocket events <http://help.amigocloud.com/hc/en-us/articles/204246154>`__ before continue reading.

To start listening to websocket events related to your user (multicast
events), do (you must be logged in to start listening to your events):

.. code:: python

    amigocloud.listen_user_events()

Once you're listening to your events, you can start adding callbacks to
them. A callback is a function that will be called everytime the event
is received. These functions should have only one parameter, that would be a python dict.

.. code:: python

    def project_created(data):
        print 'User id=%(user_id)s created project id=%(project_id)s' % data
    amigocloud.add_callback('project:creation_succeeded', project_created)

Realtime events are broadcast events related to realtime dataset. To start listening to them, do:

.. code:: python

    amigocloud.listen_dataset_events(owner_id, project_id, dataset_id)

Then add a callback for them:

.. code:: python

    def realtime(data):
        print 'Realtime dataset id=%(dataset_id)s' % data
        for obj in data['data']:
            print "Object '%(object_id)s' is now at (%(latitude)s, %(longitude)s)" % obj
    amigocloud.add_callback('realtime', realtime)

Finally, start running the websocket client:

.. code:: python

    ac.start_listening()

This method receives an optional parameter ``seconds``. If ``seconds``
is ``None`` (default value), the client will listen forever. You might
want to run this method in a new thread.

Exceptions
~~~~~~~~~~

An ``AmigoCloudError`` exception will be raised if anything fails during
the request:

.. code:: python

    try:
        amigocloud.post('me/projects')
    except AmigoCloudError as err:
        print 'Something failed!'
        print 'Status code was', err.response.status_code
        print 'Message from server was', err.text

.. |requests| replace:: ``requests``
.. _requests: https://requests.readthedocs.io/en/master/
.. |gevent| replace:: ``gevent``
.. _gevent: https://github.com/gevent/gevent
.. |socketIO_client| replace:: ``socketIO_client``
.. _socketIO_client: https://github.com/invisibleroads/socketIO-client
.. |six| replace:: ``six``
.. _six: https://github.com/benjaminp/six