Skip to main content

Ophyd Registry

Python Tests

A registry to keep track of, and retrieve, Ophyd objects.

The Ophyd registry provides a way to keep track of the devices (including components, motors, signals, etc.) that have been defined across a project. In order for the registry to know of a device, that device must first be registered, though there are ways to do this automatically.

This allows for a simple way to keep track of the Ophyd devices that were created in your project.

import ophyd
from ophydregistry import OphydRegistry

# Register the devices when they're created
registry = OphydRegistry()
registry.register(ophyd.sim.motor)

# Then elsewhere in your project, use them...
registry['motor'].set(15.3)

Installation

The easiest way to install the ophyd registry is with pip from PyPI:

python -m pip install ophyd-registry

To create a developer installation:

git clone https://github.com/spc-group/ophyd-registry.git
python -m pip install "ophyd_registry[dev]"

Usage

Registering Devices

There are three ways to have an instrument registry know about a device.

  1. Implicitly capture all Ophyd objects
  2. Register a device class
  3. Register individual objects

By default, a new instrument registry will alert itself to all future devices:

from ophydregistry import Registry
registry = Registry()

the_device = MyDevice("255id:Dev:", name="my_device")

assert registry.find("my_device") is the_device

This greedy behavior can be suppressed with the auto_register parameter. It can also be turned off after initialization by setting the auto_register attribute to False::

registry = Registry(auto_register=False)

# Make a bunch of devices
...

# Turn if off for this one
registry.auto_register = False
device = MyDevice(...)
registry.auto_register = True

# Register some other devices maybe
...

If auto_register is false, then a device class can be decorated to allow the registry to find it:

from ophyd import Device
from ophydregistry import Registry

registry = Registry(auto_register=False)

@registry.register
class MyDevice(Device):
    ...

the_device = MyDevice("255id:Dev:", name="my_device")

assert registry.find("my_device") is the_device

Lastly, individual instantions of a device can be explicitly registered.

from ophyd import Device
from ophydregistry import Registry

registry = Registry(auto_register=False)

class MyDevice(Device):
    ...

the_device = MyDevice("255id:Dev:", name="my_device")
registry.register(the_device)

assert registry.find("my_device") is the_device

Looking Up Registered Devices/Components

Registered objects can be found by name, label, or both. The easist way is to treat the registry like a dictionary: registry['motor1']. This will look for an individual device first by name then by label. It will raise an exception if the number of devices is not 1.

For more sophisticated queries, the Registry.find() method will return a single result, while Registry.findall() returns more than one. By default, findall() will raise an exception if no objects match the criteria, but this can be overridden with the allow_none keyword argument.

The registry uses the built-in concept of device labels in Ophyd. The registry's find() and findall() methods allows devices to be looked up by label or device name. For example, assuming four devices exist with the label "ion_chambers", then these devices can be retrieved using the registry:

ion_chambers = registry.findall(label="ion_chambers")
assert len(ion_chambers) == 4

Devices can also be found by name:

ion_chambers = registry.find(name="I0")
assert len(ion_chambers) == 4

A set of the root devices, those without a parent, can be retrieved at Registry.root_devices.

Looking Up Sub-Components by Dot-Notation

For simple devices, the full name of the sub-component should be enough to retrieve the device. For example, to find the signal preset_time on the device named "vortex_me4", the following may work fine:

preset_time = haven.registry.find("vortex_me4_preset_time")

However, if the component is lazy and has not been accessed prior to being registered, then it will not be available in the registry. Sub-components can instead be accessed by dot notation. Unlike the full device name, dot-notation names only resolve when the component is requested from the registry, at which point the lazy components can be accessed.

For example, area detectors use many lazy components. If sim_det is an area detector with a camera component sim_det.cam, then the name of the gain channel is "sim_det_cam_gain", however this is a lazy component so is not available. Instead, retrieving the device by haven.registry.find("sim_det.cam.gain") will first find the area detector ("sim_det"), then access the cam attribute, and then cam's gain attribute. This has the side-effect of instantiating the lazy components.

Removing Devices

The OphydRegistry class behaves similarly to a python dictionary.

To remove all devices from the registry, use the clear() method:

registry.clear()

To remove disconnected devices from the registry, use the pop_disconnected() method with an optional timeout:

# Wait 5 seconds to give devices a chance to connect
disconnected_devices = registry.pop_disconnected(timeout=5)

To remove individual objects, use either the del keyword, or the pop() method. These approaches work with either the OphydObject instance itself, or the instance's name:

# Just delete the item and move on
# (by name)
del registry["motor1"]
# (by reference)
motor = registry['motor1']
del registry[motor]

# Remove the item and use it
# (return a simulated motor if "motor1" is not in the registry)
motor = registry.pop("motor1", ophyd.sim.motor)
motor.set(5).wait()

Keeping References

It may be useful to not actually keep a strong reference to the OphydObjects. This means that if all other references to the object are removed, the device may be dropped from the registry.

By default, the registry keeps direct references to the objects that get registered, but if initalized with keep_references=False the Registry will not keep these references. Instead, it is up to you to keep references to the registered objects.

# Create two registers with both referencing behaviors
ref_registry = Registry(keep_references=True)
noref_registry = Registry(keep_references=False)
motor = EpicsMotor(...)

# Check if we can get the motor (should be no problem)
ref_registry[motor.name]  # <- succeeds
noref_registry[motor.name]  # <- succeeds

# Delete the object and try again
del motor
gc.collect()  # <- make sure it's *really* gone

# Check again if we can get the motor (now it gets fun)
ref_registry[motor.name]  # <- succeeds
noref_registry[motor.name]  # <- raises ComponentNotFound

Integrating with Typhos

Typhos includes a PyDM plugin that can directly interact with ophyd devices. It requires ophyd objects to be registered in order to find them. ophyd_registry can automatically register devices with Typhos by simply passing the use_typhos argument when creating the registry:

    
from ophydregistry import Registry
registry = Registry(use_typos=True)

or setting the use_typhos attribute on an existing registry:

    
from ophydregistry import Registry
registry = Registry()
registry.use_typhos = True

If using the typhos registry, calling the clear() method on the ophyd registry will also clear the Typhos registry.

Release files for ophyd-registry 2.0.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for ophyd-registry 2.0.0
File Size Uploaded
ophyd_registry-2.0.0.tar.gz 13.0 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for ophyd-registry 2.0.0
File Interpreter ABI Platform
ophyd_registry-2.0.0-py3-none-any.whl Python 3 none any Details

Total release size: 27.2 kB

Release files / ophyd_registry-2.0.0.tar.gz

Download URL ophyd_registry-2.0.0.tar.gz
Size 13.0 kB
Tags Source
SHA-256 checksum
How to use checksums
554e0a58ae30dd8c631ed543f7155b45ec8f973f03c3d3cf0af9c416c8e63ec0
BLAKE2b-256 checksum
How to use checksums
b14b1ef9bf8f1c746eee977369e950630b74c4ac098912e17097d256f5141dcd
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.12.8

Release files / ophyd_registry-2.0.0-py3-none-any.whl

Download URL ophyd_registry-2.0.0-py3-none-any.whl
Size 14.2 kB
Tags Python 3
SHA-256 checksum
How to use checksums
fcb718620833eefe87b96d7716aac69531c6286455e487873eb9ffdf088197a8
BLAKE2b-256 checksum
How to use checksums
757ff2dc377aff9e87071c311dae55702696fbeb960c82da9e473ce72cd52d7a
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.12.8

Release history Release notifications | RSS feed

This release

2.0.0 This release

2 release files

1.5.0

2 release files

1.4.0

2 release files

1.3.3

2 release files

1.3.2

2 release files

1.3.1

2 release files

1.3.0

2 release files

1.2.1

2 release files

1.2.0

2 release files

1.1.0

2 release files

1.0.0

2 release files

0.8

2 release files

0.7

2 release files

0.6

2 release files

0.5

2 release files

0.4

2 release files

0.3

2 release files

0.2

2 release files

0.1

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page