A framework for using remote lab instruments as local resources, built on Pyro5.
Developed by Sequoia Ploeg (for CamachoLab, Brigham Young University).
This project aims to allow all laboratory instruments to be accessed as local objects from a remote machine. Instruments that don't natively support such access, such as those required to be connected by a USB cable (or similar), are wrapped with a Pyro5 interface. However, this library may contain other instruments that are already internet-capable and don't rely on Pyro5. That's alright; we're just trying to create a minimal-dependency, one-stop-shop for laboratory instruments!
Note: while the software says "OS Independent", some of the servers are OS-specific. For example, ThorLabs DLL's only work on Windows. However, you could use PyroLab to connect to those devices from any operating system.
PyroLab is pip installable:
pip install pyrolab
PyroLab declares several "extras", depending on which instruments you need to support:
pip install pyrolab[tsl550, oscope]
The full list of supported extras is:
- tsl550
- oscope
- arduino
- monitor
You can also clone the repository, navigate to the toplevel, and install in editable mode (make sure you have pip >= 21.1):
pip install -e .
There's a web application that can monitor your PyroLab nameserver and provide an easy-to-access status board. It's a Flask app that can be served using a production grade server. Install it using pip:
pip install pyromonitor
To run:
pyromonitor up
Locally available instruments just import drivers without using any of the other features of PyroLab.
from pyrolab.drivers.lasers.tsl550 import TSL550
laser = TSL550("COM4")
laser.on()
laser.power_dBm(12)
laser.open_shutter()
laser.sweep_set_mode(continuous=True, twoway=True, trigger=False, const_freq_step=False)
First, make sure all configurations on the nameserver computer, instrument server computer, and client are correct (with the proper keys, if configured, etc.).
Run a nameserver:
from pyrolab.api import start_ns_loop
start_ns_loop()Provide a service:
from pyrolab.api import Daemon, locate_ns
from pyrolab.drivers.sample import SampleService
daemon = Daemon()
ns = locate_ns(host="localhost")
uri = daemon.register(SampleService)
ns.register("test.SampleService", uri)
try:
daemon.requestLoop()
finally:
ns.remove("test.SampleService")Connect using a remote client:
from pyrolab.api import locate_ns, Proxy
ns = locate_ns(host="localhost")
uri = ns.lookup("test.SampleService")
with Proxy(uri) as service:
resp = service.echo("Hello, server!")
print(type(resp), resp)PyroLab stores information about instruments and servers when it closes. This means that once PyroLab has been configured once, each time it is restarted, it will remember and reload the previous configuration. Hence, once a server is set up, unless the available instruments, nameserver, or other configurations change, PyroLab will automatically work when started, every time!
For an example of how a new PyroLab instrument server should be configured the
first time it's run, see examples/library-catalog/config.yaml.
-
Another instrument library? What about all the others?
In our experience, many of the other libraries are buggy or have difficulty with network connections. So, our approach was to rely on a well developed and time-tested framework (Pyro) instead of worrying about developing and supporting our own custom set of servers. -
Is this a standalone software that automatically supports all the advertised instruments?
No; many of these instruments depend on other software already being installed. In particular, ThorLabs equipment depends on ThorLabs software already being installed on the computer connected to the physical hardware (but not on the remote computer!). As much as possible, though, we try to make the drivers standalone capable.
Since the passing of data is, by definition, between hosts and over IP, PyroLab avoids the use of complex Python objects for return values that will be transmitted to remote machines. Since serialization is complicated, and security is even harder, we resort to using only basic Python types when interfacing with hardware (i.e., Python lists, ints, tuples, and not NumPy arrays, matplotlib plot objects, custom objects, etc.).
Make sure you have committed a changelog file under docs/changelog titled
<major>.<minor>.<patch>-changelog.md before bumping version. Also, the git
directory should be clean (no uncommitted changes).
To bump version prior to a release, run one of the following commands:
bumpversion major
bumpversion minor
bumpversion patch
This will automatically create a git tag in the repository with the corrresponding version number and commit the modified files (where version numbers were updated). Pushing the tags (a manual process) to the remote will automatically create a new release. Releases are automatically published to PyPI and GitHub when git tags matching the "v*" pattern are created (e.g. "v0.3.2"), as bumpversion does.
After bumping version, you can view the tags on the local machine by running
git tag. To push the tags to the remote server and trigger the release
workflow, you can run git push origin <tagname>.
For code quality, please run isort and black before committing (note that the latest release of isort may not work through VSCode's integrated terminal, and it's safest to run it separately through another terminal).
Much of the API documentation is autogenerated from the source code. Gitignores are in place to prevent you from committing autogenerated pages.
To build the docs, navigate to the docs/ directory and run:
pip install -r requirements.txt
make html
You can test the build and view the bundled artifacts using build. It's recommended you build locally before pushing to PyPI. In particular, double check the included files and make sure only the required files are there by modifying MANIFEST.in as necessary.