Python reference implementation of the Laboratory Experiment COntrol (LECO) protocol.
The reviewed branch contains reviewed code, which does not yet contain all necessary modules and classes. Development happens in the main branch.
Note: LECO is still under development, such that the code and API might change. The LECO protocol branch pyleco-state contains the assumptions used in this project, which are not yet accepted into the LECO main branch. These things might change, if LECO defines them differently.
For a tutorial on how to get started, see GETTING_STARTED.md.
- Install Python,
- Execute
pip install pyleco
, - Import
pyleco
in your python scripts.
PyLECO is an implementation of LECO, for the full protocol specifications please visit https://github.com/pymeasure/leco-protocol. LECO offers a protocol for data exchange, for example for laboratory experimental control.
There exist two different communication protocols in LECO.
- The control protocol allows to exchange messages between any two Components in a LECO network, which is useful for controlling devices.
- The data protocol is a broadcasting protocol to send information to all those, who want to receive it, which is useful for regular measurement data or for log entries.
A LECO network needs at least one Coordinator (server), which routes the messages among the connected Components.
Each Component has a name unique in the network.
This name consists in the name of the Coordinator they are connected to and their own name.
For example N1.component1
is the full name of component1
connected to the Coordinator of the Namespace N1
.
The Coordinator istelf is always called COORDINATOR
.
The default messaging content of the control protocol are remote procedure calls (RPC) according to jsonrpc.
With these RPCs you execute a method on the remote Component.
For example you have an Actor, which is for example a Component controlling a measurement instrument.
In order to set the output of that measurement instrument, you want to call the set_output
method of that instrument.
For that purpose, you send a message which encodes exactly that (via jsonrpc): the method to call and the parameters of that method.
For a minimum setup, you need:
- a Coordinator (just execute
coordinator
in your terminal or run thecoordinator.py
file with your Python interpreter) - one Component
For example, you can use a Communicator
instance to send/receive messages via LECO protocol.
The following example requests the list of Components connected to the Coordinator:
from pyleco.utils.communicator import Communicator
c = Communicator(name="TestCommunicator")
connected_components = c.ask_rpc(method="send_local_components")
print(connected_components)
Let's say you have an instrument with a pymeasure driver Driver
, which you want to control.
You need to start (in different threads):
- a Coordinator (as shown above).
- an
Actor
instance listening to commands and controlling the instrument:actor = Actor(name="inst_actor", cls=Driver)
. For an example see thepymeasure_actor.py
in the examples folder. - a
TransparentDirector
:director=TransparentDirector(actor="inst_actor")
. Theactor
parameter has to match the Actor'sname
parameter. For an example of a measurement script seemeasurement_script.py
in the examples folder.
If you want to set some property of the instrument (e.g. instrument.voltage = 5
), you can just use the director
transparently: director.device.voltage = 5
.
In the background, the TransparentDirector, which does not have a device
, sends a message to the Actor to set that parameter.
The Actor in turn sets that parameter of the instrument driver, which in turn will send some command to the device to take an appropriate action (e.g. setting the voltage to 5 V).
Currently you cannot call methods in a similar, transparent way, without manual intervention.
You can add RemoteCall
descriptor (in transparent_director module) to the director
for each method call you want to use.
Afterwards you can use these methods transparently similar to the property shown above.
See the docstrings of the individual classes for more information and for examples.
- The
core
subpackage contains elements necessary for implementing LECO, especially theMessage
class, which helps to create and interpret LECO messages. - The
utils
subpackage contains modules useful for creating LECO Components.- The
Communicator
can send and receive messages, but neither blocks (just for a short time waiting for an answer) nor requires an extra thread. It is useful for usage in scripts. - The
MessageHandler
handles incoming messages in a continuous loop (blocking until stopped). It is useful for creating standalone scripts, like tasks for the Starter. - The
Listener
offers the same interface as the Communicator, but listens in an extra thread for incoming messages. It is useful if you want to react to incoming messages (via data or control protocol) and if you want to send messages of your own accord, for example for GUI applications.
- The
- The
coordinators
subpackage contains the differenc Coordinators.Coordinator
is the Coordinator for the control protocol (exchanging messages).proxy_server
is the Coordinator for the data protocol (broadcasting).
- The
actors
subpackage contains Actor classes to control devices. - The
management
subpackage contains Components useful for experiment management- The
Starter
can execute tasks in separate threads. A task could be an Actor controlling some Device. - The
DataLogger
listens to published data (data protocol) and collects them.
- The
- The
directors
subpackage contains Directors, which facilitate controlling actors or management utilities.- For example the
CoordinatorDirector
has a method for getting Coordinators and Components connected to a Coordinator. - The
TransparentDirector
reads / writes all messages to the remote actor, such that you use the director'sdevice
as if it were the instrument itself.
- For example the
The pyleco-extras package contains additional modules.
Among them are GUIs controlling the DataLogger
and the Starter
.