Device Proxy JSON Wrapper
DeviceProxyJson is a convenience wrapper that transparently encodes/decodes
JSON strings on their way to/from a tango.DeviceProxy.
Usage
The intent is that operation is transparent and things work as per a normal DeviceProxy,
except you’re now free of the drudgery of typing json.loads and json.dumps all the time.
Instantiation
You can instantiate via an existing DeviceProxy object, or just pass a device name and
one will be created for you:
# Using a DeviceProxy
cnic_proxy = tango.DeviceProxy("low-cbf/cnic/1")
cnic = DeviceProxyJson(cnic_proxy)
# Or, straight from the name
cnic = DeviceProxyJson("low-cbf/cnic/1")
Attributes & Commands using JSON are automatically detected if their description includes “JSON”. Optional parameters can add Attributes & Commands that are not picked up by the automatic detection logic:
cnic = DeviceProxyJson("low-cbf/cnic/1", json_commands=["CallMethod"], json_attributes=["vd__configuration"])
You can also exclude Attributes & Commands that are incorrectly detected:
# Perhaps MyCommand's doc_in says "Use CSV, never JSON!"
# and notJsonAttr's description is "This is a colon-separated list, NOT JSON!"
# These would normally be automatically detected as using JSON by DeviceProxyJson!
DeviceProxyJson("my/device/name", not_json_commands=["MyCommand"], not_json_attributes=["notJsonAttr"])
These parameters can be combined as you see fit. If you try to include & exclude the same thing, exclusion wins.
Reading a JSON Attribute
cnic = DeviceProxyJson("low-cbf/cnic/1", json_attributes=["vd__configuration"])
# We can loop over a list of dicts
for cfg in cnic.vd__configuration: # look Ma, no json.loads !
print("Beam:", cfg["beam"], "Station:", cfg["station"], cfg["substation"])
Beam: 1 Station: 345 1
Beam: 2 Station: 346 1
Beam: 3 Station: 347 1
Beam: 4 Station: 348 1
Beam: 5 Station: 349 1
Sending a JSON Command
If the Command is expecting a dict, you can use either keyword arguments, or pass a dict directly, your choice.
Note that DeviceProxyJson has no knowledge of the underlying Tango Device’s
JSON schema, so you’re on your own in terms of what data structure or keywords to supply.
cnic = DeviceProxyJson(cnic_proxy, json_commands=["CallMethod"])
# Keyword Arguments
cnic.CallMethod(method="stop_receive") # Command explicitly configured as JSON above
# Any data type that can be processed by json.dumps can be used directly
request = {"version": "0.1.9", "source": "nexus"}
cnic.SelectPersonality(request) # Command auto-detected, as its doc_in contains "JSON"
API
- class ska_low_cbf_integration.tango.DeviceProxyJson(target, json_commands=None, not_json_commands=None, json_attributes=None, not_json_attributes=None)
Transparently translate JSON to/from a Tango DeviceProxy.
- __getattr__(attr_name)
Get a Python attribute from our target (note: may be a Tango Command).
- __init__(target, json_commands=None, not_json_commands=None, json_attributes=None, not_json_attributes=None)
Create a transparent JSON wrapper.
JSON conversion is automatically applied to Tango Attributes or Commands that contain the word “JSON” (case-insensitive) in their description (doc_in aka in_type_desc for Commands). Parameters are available to override detection behaviour.
Attributes detected/configured as JSON will have their return value decoded.
Commands detected/configured as JSON will have their input argument encoded.
- Parameters:
target (
Union[DeviceProxy,str]) – Tango DeviceProxy or device namejson_commands (
Union[Iterable,str,None]) – Commands that take JSON input (overrides auto-detection)not_json_commands (
Union[Iterable,str,None]) – Commands that do not take JSON input (overrides auto-detection)json_attributes (
Union[Iterable,str,None]) – Attributes that return JSON strings (overrides auto-detection)not_json_attributes (
Union[Iterable,str,None]) – Attributes that do not return JSON strings (overrides auto-detection)
- __setattr__(key, value)
Delegate to target, once we are initialised.