NOTE: This project reached end of support and is no longer maintained. Parts of the project are continued in https://github.com/IndustryFusion/DigitalTwin.
================================= This library provides Python bindings to the Open IoT Service Platform (OISP).
Note In order to experiment and test more easily, it is recommended to install OISP locally. You can follow these instructions to get going easily.
If you only want to use the SDK, just run:
pip install oisp
If you want to develop the SDK, install using the git repository instead.
- Clone the oisp-sdk-python repository from GitHub
- Install the library:
cd oisp-sdk-python
make installAssuming you have OISP running locally, open your browser and visit localhost. You can create a new user using the sign up button in the top right corner.
It is possible to create a user using the REST API as well, but you will still have to check you e-mail to activate the user. The methods for retrieving users are defined in the Client class.
Connection to the host is managed by the Client class. You need the URL for the API to create an instance.
import oisp
client = oisp.Client(api_root="http://localhost/v1/api/")If you are connecting over proxies, you can specify those using the proxies parameter, see method documentation for __init__ for details.
OISP offer couple of different authentication mechanism for different purposes. In order to manage accounts and devices you need to authenticate as a user. We will have a look at alternative strategies later.
client.auth("username", "password")Once you log in using the auth method, you receive a token that is valid for one hour and will automatically be included in the requests made by the client object. You will get an AuthenticationError the first time you try to make a request after the token expires and need to reauth using the same method.
A User represents a person interacting with the system, wheres an Account is an organizational unit. An account can be managed by multiple users with different roles, and a user can manage multiple accounts.
Accounts are used to manage devices, component types etc.
You can create a new account simply by specifying its name. Please be aware that account names are not necessarily unique. After creating a new account, you will need to reauth to refresh your token.
account = client.create_account("test_account")
client.auth("username", "password")In order to get a list of accounts that are managed by the user logged in, use the following method>
accounts = client.get_accounts()A devices represents an IoT Endpoint. As devices are managed by accounts, a device is created by the following method:
new_device = account.create_device("device_id", "device_name")Device id has to be globally unique. We recommend using the MAC adress for the device. Device name does not have to be unique.
Functionality regarding a specific device is implemented in Device class (for example sending data, activation, deleting a device), functionality regarding all devices that belong to an account (data search, device creation etc) are implemented in the Account class.
Before sending data, a device needs to be activated. Activation returns a device token, which can be used for simple operations like sending data. This is the preferred way for requests that do not need priviliges, so save the returned device token. A device token can be updated by activating the device again.
device_token = device.activate()You can retrieve a previously activated device directly from the client using the device token and device id. Note that the following method does not require you to auth before.
device = client.get_device(device_token, device_id)A device can provide multiple datasets, which are organized in 'components'. Every component has a type and a name. For this example we will use a predefined component type but you can also create your own.
Component types catalog belongs to an account and can be managed using the get_component_types_catalog, create_component_type, get_component_type and other similar methods provided by the Account class. See respective method descriptions for details.
We add a component to a device by specifying the name and the the type. This returns a json dictionary representing the newly created component and contains the component id, which is used to add data to the component.
response = device.add_component("test_temperature_component", "temperature.v1.0")
cid = response["cid"]You can access the existing components of a device with device.components.
In order to add data, use the add_sample method of a device (this does not submit the data to the service yet):
device.add_sample(cid, value_1)
device.add_sample(cid, value_2)You can use the on and loc parameters if you want to include the time and location in which the data was sampled. If ommited on will be set to current time and loc will be left blank.
Once you want to submit and save your data, call the following method:
device.submit_data()Seperation of adding samples and submisson allows you to add datapoints as you collect the input from sensors, but save on requests by submitting multiple values at once.
You need to build a query to search for data that belongs to an account. The structure of the query is described in the API documentation ( here if OISP is running locally ) and you can use a json style dictionary for your query.
However, it is more convenient to use the DataQuery class. A query created without any parameters will return all data submitted so far. For details on query parameters see the documentation on this class.
query = oisp.DataQuery()
response = account.search_data(query)response is of type QueryResponse and response.samples array contains Sample objects which store data, location, time and device information.
# See all data values from the response
data_values = [sample.value for sample in response.samples]