Python

The pre-built python package blickfeld_qb2 is available on pypi. It contains all the necessary sources to communicate securely to a Qb2 device.

Python is recommended for rapid prototyping, data evaluation and tools. Usually, the performance is sufficient for live point cloud streaming & processing as e.g. it directly provides efficient memory-mapped numpy arrays as output. In combination with Open3D, it can cover a lot of application scenarios.

In this guide the Blickfeld Qb2 Python library will be first installed and a secure connection to a Qb2 device will be established. Then the version of the currently installed firmware is fetched from the device. Finally we take a look at asynchronous usage of the Blickfeld library.

Installing Blickfeld Qb2 Library

The python package to control the Qb2 device is called blickfeld_qb2 and can be installed from the PyPI repository with the respective command shown below:

$ pip3 install -U blickfeld_qb2

After the installation the blickfeld_qb2 library can be imported in Python.

Python importing the blickfeld_qb2 library

import blickfeld_qb2

Connecting to Qb2

To interact with a Qb2 device, a gRPC connection must be established. Either the IP address or the hostname of the Qb2 device can be used to establish this connection.

The implementation uses the gRPC-Channel in combination with the TokenFactory from the blickfeld_qb2 library.

The resulting connection is encrypted by default. The TokenFactory additionally authenticates a configured account on the Qb2.

Simple example which encrypted & authenticated communication channel to Qb2

import blickfeld_qb2

# Create token factory using an application key
token_factory = blickfeld_qb2.TokenFactory(
    application_key_secret="application-key-for-qb2-xxxxxxxxx"
)

# Open a secure & authenticated connection to Qb2
with blickfeld_qb2.Channel(
        fqdn_or_ip="qb2-xxxxxxxxx",
        token=token_factory
    ) as channel:
    # Call methods of services here
    # ..
    pass

A matching application key for the account which should be authenticated has to be created before. If unsure please read the documentation about application keys in order to understand how to do this first.

If the application key is invalid, the first API call on the channel will result in a gRPC-error with code 5 / NOT_FOUND.

In the example, replace application-key-for-qb2-xxxxxxxxx with the generated application key.

The serial_number can be passed additionally to authenticate a particular device. If not supplied, the authentication only checks if it is a Blickfeld device.

Unauthenticated Connection

If user-management is not enabled on Qb2, an unauthenticated connection can be established. The resulting connection will still be encrypted. Here only the gRPC Channel from the blickfeld_qb2 library is used.

Simple example which creates a encrypted communication channel to Qb2

import blickfeld_qb2
# Open a secure connection to Qb2
with blickfeld_qb2.Channel(fqdn_or_ip="qb2-xxxxxxxxx") as channel:
    # Call methods of services
    # ..

Establishing unauthenticated connections is deprecated and will be disabled for upcoming Qb2 firmware releases.

Simple API Requests

The code snippet below provides an example of connecting to Qb2 with the hostname qb2-xxxxxxxxx and retrieving information about its firmware version. In this case, the Firmware service from blickfeld_qb2.system.services namespace is used over the gRPC channel making a request get_status() to receive a response containing information about the currently installed firmware.

import blickfeld_qb2

# Create token factory using an application key
token_factory = blickfeld_qb2.TokenFactory(
    application_key_secret="application-key-for-qb2-xxxxxxxxx"
)

# Open a secure & authenticated connection to Qb2
with blickfeld_qb2.Channel(
        fqdn_or_ip="qb2-xxxxxxxxx",
        token=token_factory
    ) as channel:
    # Request the firmware status
    response = blickfeld_qb2.system.services.Firmware(channel).get_status()
    # Extract the firmware version
    firmware_version = response.status.installed_firmware_info
    # Print the firmware version
    print(firmware_version)

    # NOTE: A channel can be reused for other requests. It is not required to create a channel per request.

Every response to the service call is an object that typically comprises several nested objects following the protocol schema defined in the corresponding sections of the documentation. In this scenario, the information about the installed firmware version can be obtained via the field response.status.installed_firmware_info.

The print output should resemble the example shown below:

FirmwareInfo(label='AEGIS v2.0.5', version=Version(major=2, patch=5, revision='3d7b970b', is_release=True), allowed_downgrade_version=Version(major=1, minor=12, patch=6))

Asynchronous Requests

If necessary, the Blickfeld library can also be used in a fully asynchronous mode. In the example shown below we introduce a coroutine get_firmware_version(), which opens an asynchronous gRPC channel and utilizes an asynchronous request async_get_status() to obtain the same information about the Qb2 firmware status.

import asyncio
import blickfeld_qb2

async def main():
    # Create token factory using an application key
    token_factory = blickfeld_qb2.TokenFactory(
        application_key_secret="application-key-for-qb2-xxxxxxxxx"
    )

    # Open a secure, authenticated & asynchronous connection to Qb2
    with blickfeld_qb2.Channel(
            fqdn_or_ip="qb2-xxxxxxxxx",
            token=token_factory
        ) as channel:
        # Request the firmware status asynchronously and await for the response
        response = await blickfeld_qb2.system.services.Firmware(channel).async_get_status()
        # Get the firmware version from the response
        firmware_version = response.status.installed_firmware_info
        # Print the firmware version
        print(firmware_version)

        # NOTE: A channel can be reused for other requests. It is not required to create a channel per request.

# run couroutine on the event loop
asyncio.run(main())

To make an asynchronous request over the asynchronous channel, the desired method name should be prefixed with async_.

This coroutine can be further executed on the event loop. The print output resemble the example shown below:

FirmwareInfo(label='JAKOB v1.5.0', version=Version(major=1, minor=5, revision='e87dfda0', is_release=True), allowed_downgrade_version=Version(minor=26))