AttributeProxy

AttributeProxy#

class tango.AttributeProxy(*args, green_mode=None)[source]#

AttributeProxy is the high level Tango object which provides the client with an easy-to-use interface to TANGO attributes.

To create an AttributeProxy, a complete attribute name must be set in the object constructor.

Example:

att = AttributeProxy(“tango/tangotest/1/long_scalar”)

Note: PyTango implementation of AttributeProxy is in part a python reimplementation of the AttributeProxy found on the C++ API.

delete_property(*args, **kwds)#

Delete the given properties for this attribute. :param value: Can be one of the following:

  1. str [in] - Single property data to be deleted.

  2. DbDatum [in] - Single property data to be deleted.

  3. DbData [in] - Several property data to be deleted.

  4. list[str | bytes | DbDatum] [in] - Several property data to be deleted.

  5. dict[str, object] [in] - Keys are property names to be deleted (values are ignored).

  6. dict[str, tango.DbDatum] [in] - Several DbDatum.name are property names to be deleted (keys are ignored).

Throws:

TypeError: Raised in case of value has the wrong type.

tango.NonDbDevice: Raised in case of a non-database device error.

tango.ConnectionFailed: Raised on connection failure with the database.

tango.CommunicationFailed: Raised on communication failure with the database.

tango.DevFailed: Raised on a device failure from the database device.

event_queue_size(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().event_queue_size(...)

For convenience, here is the documentation of tango.DeviceProxy.event_queue_size():

event_queue_size (self: tango._tango.DeviceProxy, event_id: typing.SupportsInt) -> int

Returns the number of stored events in the event reception buffer. After every call to DeviceProxy.get_events(), the event queue size is 0. During event subscription the client must have chosen the ‘pull model’ for this event. event_id is the event identifier returned by the DeviceProxy.subscribe_event() method.

Parameters:

event_id (int) – event identifier

Throws:

EventSystemFailed

Added in version 7.0.0.

get_config(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().get_attribute_config(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.get_attribute_config():

get_attribute_config(self, name: str, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> AttributeInfoEx get_attribute_config(self, names: list[str], green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> AttributeInfoList

Return the attribute configuration for a single or a list of attribute(s). To get all the attributes pass a sequence containing the constant tango.constants.AllAttr

Deprecated since version use: get_attribute_config_ex instead

Parameters:

  • name (str) – Attribute name.

  • names (list[str]) – Attribute names.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

An AttributeInfoEx or AttributeInfoList object containing the attribute(s) information.

Return type:

tango.AttributeInfoEx | tango.AttributeInfoList

Throws:

tango.ConnectionFailed: Raised in case of a connection failure.

tango.CommunicationFailed: Raised in case of a communication failure.

tango.DevFailed: Raised in case of a device failure.

TypeError: Raised in case of an incorrect type of input arguments.

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

get_device_proxy() → DeviceProxy[source]#

A method which returns the device associated to the attribute

get_events(event_id: int, callback: Callable = None, extract_as: ExtractAs = ExtractAs.Numpy) → None#

This method is a simple way to do:

self.get_device_proxy().get_events(...)

For convenience, here is the documentation of tango.DeviceProxy.get_events():

The method extracts all waiting events from the event reception buffer.

If callback is not None, it is executed for every event. During event subscription the client must have chosen the pull model for this event. The callback will receive a parameter of type EventData, AttrConfEventData or DataReadyEventData depending on the type of the event (event_type parameter of subscribe_event).

If callback is None, the method extracts all waiting events from the event reception buffer. The returned event_list is a vector of EventData, AttrConfEventData or DataReadyEventData pointers, just the same data the callback would have received.

Parameters:

  • event_id (int) – The event identifier returned by the DeviceProxy.subscribe_event() method.

  • callback (typing.Callable) – Any callable object or any object with a “push_event” method.

  • extract_as (ExtractAs) – (Description Needed)

Throws:

tango.EventSystemFailed: Raised in case of a failure in the event system.

TypeError: Raised in case of an incorrect type of input arguments.

ValueError: Raised in case of an invalid value.

See also:

subscribe_event()

get_last_event_date(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().get_last_event_date(...)

For convenience, here is the documentation of tango.DeviceProxy.get_last_event_date():

get_last_event_date (self: tango._tango.DeviceProxy, event_id: typing.SupportsInt) -> tango._tango.TimeVal

Returns the arrival time of the last event stored in the event reception buffer. After every call to DeviceProxy:get_events(), the event reception buffer is empty. In this case an exception will be returned. During event subscription the client must have chosen the ‘pull model’ for this event. event_id is the event identifier returned by the DeviceProxy.subscribe_event() method.

Parameters:

event_id (int) – event identifier

Throws:

EventSystemFailed

Added in version 7.0.0.

get_poll_period(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().get_attribute_poll_period(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.get_attribute_poll_period():

get_attribute_poll_period(self: tango._tango.DeviceProxy, attr_name: str) -> int

Return the attribute polling period in milliseconds

Parameters:

attr_name (str) – attribute name

get_property(*args, **kwds)#

Get a (list) property(ies) for an attribute.

Parameters:

  • propname –

    Can be one of the following:

    1. str [in] - Single property data to be fetched.

    2. DbDatum [in] - Single property data to be fetched.

    3. DbData [in] - Several property data to be fetched.

    4. list[str | bytes] [in] - Several property data to be fetched.

    5. list[tango.DbDatum] [in] - Several property data to be fetched.

    6. dict[str, object] [in] - Keys are property names to be fetched (values are ignored).

    7. dict[str, tango.DbDatum] [in] - Several DbDatum.name are property names to be fetched (keys are ignored).

  • value (tango.DbData, optional) – Optional. For propname overloads with str and list[str] will be filed with the property values, if provided.

Returns:

A dict object, which keys are the property names the value associated with each key being a sequence of strings being the property value.

Throws:

TypeError: Raised in case of propname has the wrong type.

tango.NonDbDevice: Raised in case of a non-database device error.

tango.ConnectionFailed: Raised on connection failure with the database.

tango.CommunicationFailed: Raised on communication failure with the database.

tango.DevFailed: Raised on a device failure from the database device.`

Added in version 10.1.0:: overloads with dict as propname parameter

Changed in version 10.1.0:: raises if propname has an invalid type instead of returning None

get_transparency_reconnection(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().get_transparency_reconnection(...)

For convenience, here is the documentation of tango.DeviceProxy.get_transparency_reconnection():

get_transparency_reconnection (self: tango._tango.Connection) -> bool

Returns the device transparency reconnection flag.

history(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().attribute_history(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.attribute_history():

attribute_history(self: tango._tango.DeviceProxy, attr_name: str, depth: typing.SupportsInt, extract_as: tango._tango.ExtractAs = ExtractAs.Numpy) -> list[tango._tango.DeviceAttributeHistory]

Retrieve attribute history from the attribute polling buffer. See chapter on Advanced Feature for all details regarding polling

Parameters:

  • attr_name (str) – attribute name

  • depth (int) – the wanted history depth

  • extract_as (ExtractAs) – to which format extract data

Return type:

list[DeviceAttributeHistory]

Throws:

NonSupportedFeature, ConnectionFailed, CommunicationFailed, DevFailed from device

is_event_queue_empty(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().is_event_queue_empty(...)

For convenience, here is the documentation of tango.DeviceProxy.is_event_queue_empty():

is_event_queue_empty (self: tango._tango.DeviceProxy, event_id: typing.SupportsInt) -> bool

Returns true when the event reception buffer is empty. During event subscription the client must have chosen the ‘pull model’ for this event. event_id is the event identifier returned by the DeviceProxy.subscribe_event() method.

Parameters:

event_id (int) – event identifier

Throws:

EventSystemFailed

Added in version 7.0.0.

is_polled(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().is_attribute_polled(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.is_attribute_polled():

is_attribute_polled(self: tango._tango.DeviceProxy, attr_name: str) -> bool

True if the attribute is polled.

Parameters:

attr_name (str) – attribute name

name() → str[source]#

Returns the attribute name

ping(*, green_mode: GreenMode = None, wait: bool = None, timeout: int = None) → int#

This method is a simple way to do:

self.get_device_proxy().ping(...)

For convenience, here is the documentation of tango.DeviceProxy.ping():

A method which sends a ping to the device

Parameters:

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

time elapsed in microseconds

Return type:

int

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

poll(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().poll_attribute(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.poll_attribute():

poll_attribute(self: tango._tango.DeviceProxy, attr_name: str, period_ms: typing.SupportsInt) -> None

Add an attribute to the list of polled attributes.

Parameters:

  • attr_name (str) – attribute name

  • period_ms (int)) – polling period in milliseconds

put_property(*args, **kwds)#

Insert or update a list of properties for this attribute.

Parameters:

value –

Can be one of the following:

  1. str - Single property data to be inserted.

  2. DbDatum - Single property data to be inserted.

  3. DbData - Several property data to be inserted.

  4. list[str | bytes | DbDatum] - Several property data to be inserted.

  5. dict[str, DbDatum] -

    DbDatum is property to be inserted (keys are ignored).

  6. dict[str, list[str]] - Keys are property names,

    and value has data to be inserted.

  7. dict[str, object] - Keys are property names, and str(obj) is property value.

Throws:

TypeError: Raised in case of value has the wrong type.

tango.NonDbDevice: Raised in case of a non-database device error.

tango.ConnectionFailed: Raised on connection failure with the database.

tango.CommunicationFailed: Raised on communication failure with the database.

tango.DevFailed: Raised on a device failure from the database device.`

read(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().read_attribute(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.read_attribute():

Read a single attribute.

Parameters:

  • attr_name (str) – The name of the attribute to read.

  • extract_as (ExtractAs) – Defaults to numpy.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

DeviceAttribute object with read attribute value.

Return type:

DeviceAttribute

Throws:

ConnectionFailed: Raised in case of a connection failure

CommunicationFailed: Raised in case of a communication failure.

DevFailed: Raised in case of a device failure

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

Changed in version 7.1.4: For DevEncoded attributes, before it was returning a DeviceAttribute.value as a tuple (format<str>, data<str>) no matter what was the extract_as value was. Since 7.1.4, it returns a (format<str>, data<buffer>) unless extract_as is String, in which case it returns (format<str>, data<str>).

Changed in version 8.0.0: For DevEncoded attributes, now returns a DeviceAttribute.value as a tuple (format<str>, data<bytes>) unless extract_as is String, in which case it returns (format<str>, data<str>). Careful, if using python >= 3 data<str> is decoded using default python utf-8 encoding. This means that PyTango assumes tango DS was written encapsulating string into utf-8 which is the default python encoding.

Added in version 8.1.0: green_mode parameter. wait parameter. timeout parameter.

Changed in version 9.4.0: For spectrum and image attributes with an empty sequence, no longer returns DeviceAttribute.value and DeviceAttribute.w_value as None. Instead, DevString and DevEnum types get an empty tuple, while other types get an empty numpy.ndarray. Using extract_as can change the sequence type, but it still won’t be None.

read_asynch(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().read_attribute_asynch(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.read_attribute_asynch():

read_attribute_asynch(self, attr_name: str, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> int read_attribute_asynch(self, attr_name: str, cb: typing.Callable, extract_as: ExtractAs=Numpy, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> None

Read asynchronously the specified attributes.

Important

by default, TANGO is initialized with the polling model. If you want to use the push model (the one with the callback parameter), you need to change the global TANGO model to PUSH_CALLBACK. You can do this with the tango.ApiUtil.set_asynch_cb_sub_model()

Parameters:

  • attr_name (str) – an attribute to read

  • cb (typing.Optional[typing.Callable]) – push model: as soon as attributes read, core calls cb with read results. This callback object should be an instance of a user class with an attr_read() method. It can also be any callable object.

  • extract_as (ExtractAs) – Defaults to ExtractAs.Numpy

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

an asynchronous call identifier which is needed to get attribute value if poll model, None if push model

Return type:

int | None

Throws:

ConnectionFailed

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

Important

Multiple asynchronous calls are not guaranteed to be executed by the device server in the same order they are invoked by the client. E.g., a call to the method write_attribute_asynch("a", 1) followed immediately with a call to read_attribute_asynch("a") could result in the device reading the attribute a before writing to it.

Added in version 7.0.0.

read_reply(*args, green_mode: GreenMode = None, wait: bool = None, timeout: int = None, **kwargs) → DeviceAttribute#

This method is a simple way to do:

self.get_device_proxy().read_attribute_reply(...)

For convenience, here is the documentation of tango.DeviceProxy.read_attribute_reply():

read_attribute_reply(self, id: int, extract_as: ExtractAs=ExtractAs.Numpy, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> DeviceAttribute read_attribute_reply(self, id: int, poll_timeout: int, extract_as: ExtractAs=ExtractAs.Numpy, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> DeviceAttribute

Get the answer of an asynchronous read_attribute call, if it has arrived (polling model).

If the reply is ready, but the attribute raised an exception while reading, an exception will be raised by this function (DevFailed, with reason API_AttrValueNotSet).

Changed in version 10.0.0: To eliminate confusion between different timeout parameters, the core (cppTango) timeout (previously the optional second positional argument) has been renamed to “poll_timeout”. Conversely, the pyTango executor timeout remains as the keyword argument “timeout”. These parameters have distinct meanings and units:

  • The cppTango “poll_timeout” is measured in milliseconds and blocks the call until a reply is received. If the reply is not received within the specified poll_timeout duration, an exception is thrown. Setting poll_timeout to 0 causes the call to wait indefinitely until a reply is received.

  • The pyTango “timeout” is measured in seconds and is applicable only in asynchronous GreenModes (Asyncio, Futures, Gevent), and only when “wait” is set to True. The specific behavior when a reply is not received within the specified timeout period varies depending on the GreenMode.

Parameters:

  • id (int) – the asynchronous call identifier

  • poll_timeout (int, optional) –

    cppTango core timeout in ms:

    • If the reply has not yet arrived, the call will wait for the time specified (in ms).

    • If after timeout, the reply is still not there, an exception is thrown.

    • If timeout set to 0, the call waits until the reply arrives.

    • If the argument is not provided, then there is no timeout check, and an exception is raised immediately if the reply is not ready.

  • extract_as (ExtractAs) – Defaults to numpy.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

Could be:

  • If the reply is arrived and if it is a valid reply, it is returned to the caller in a DeviceAttribute.

  • If the reply is an exception, it is re-thrown by this call.

  • If the reply is not yet arrived, the call will wait (blocking the process) for the time specified in timeout. If after timeout milliseconds, the reply is still not there, an exception is thrown. If timeout is set to 0, the call waits until the reply arrived.

Return type:

tango.DeviceAttribute

Throws:

tango.AsynCall, tango.AsynReplyNotArrived, tango.ConnectionFailed, tango.CommunicationFailed, tango.DevFailed

Added in version 7.0.0.

set_config(value: AttributeInfo | AttributeInfoEx | Sequence[AttributeInfo] | Sequence[AttributeInfoEx], *, green_mode: GreenMode = None, wait: bool = None, timeout: int = None) → None#

This method is a simple way to do:

self.get_device_proxy().set_attribute_config(...)

For convenience, here is the documentation of tango.DeviceProxy.set_attribute_config():

set_attribute_config(self, attr_info: AttributeInfo | Sequence[AttributeInfo], green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> None set_attribute_config(self, attr_info_ex: AttributeInfoEx | Sequence[AttributeInfoEx], green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> None

Change the attribute configuration/extended attribute configuration for the specified attribute(s)

Parameters:

  • attr_info (tango.AttributeInfo | typing.Sequence[AttributeInfo]]) – Attribute information. This parameter is used when providing basic attribute(s) information.

  • attr_info_ex (AttributeInfoEx | typing.Sequence[AttributeInfoEx]]) – Extended attribute information. This parameter is used when providing extended attribute information. It should not be used simultaneously with ‘attr_info’.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

None

Throws:

tango.ConnectionFailed: Raised in case of a connection failure.

tango.CommunicationFailed: Raised in case of a communication failure.

tango.DevFailed: Raised in case of a device failure.

TypeError: Raised in case of an incorrect type of input arguments.

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

set_transparency_reconnection(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().set_transparency_reconnection(...)

For convenience, here is the documentation of tango.DeviceProxy.set_transparency_reconnection():

set_transparency_reconnection (self: tango._tango.Connection, yesno: bool) -> None

Set the device transparency reconnection flag

Parameters:

yesno (bool) – True to set transparency reconnection or False otherwise

state(*, green_mode: GreenMode = None, wait: bool = None, timeout: int = None) → DevState#

This method is a simple way to do:

self.get_device_proxy().state(...)

For convenience, here is the documentation of tango.DeviceProxy.state():

A method which returns the state of the device.

Parameters:

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

A DevState constant.

Return type:

DevState

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

status(*, green_mode: GreenMode = None, wait: bool = None, timeout: int = None) → str#

This method is a simple way to do:

self.get_device_proxy().status(...)

For convenience, here is the documentation of tango.DeviceProxy.status():

A method which returns the status of the device as a string.

Parameters:

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

string describing the device status

Return type:

str

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

stop_poll(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().stop_poll_attribute(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.stop_poll_attribute():

stop_poll_attribute(self: tango._tango.DeviceProxy, attr_name: str) -> None

Remove an attribute from the list of polled attributes.

Parameters:

attr_name (str) – attribute name

subscribe_event(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().subscribe_event(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.subscribe_event():

subscribe_event (self, attr_name: str, event_type: EventType, cb: typing.Callable, sub_mode: EventSubMode=EventSubMode.SyncRead, extract_as: ExtractAs=ExtractAs.Numpy, *, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> int

subscribe_event (self, attr_name: str, event_type: EventType, queuesize: int, sub_mode: EventSubMode=EventSubMode.SyncRead, extract_as: ExtractAs=ExtractAs.Numpy, *, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> int

subscribe_event (self, event_type: EventType, cb: typing.Callable, sub_mode: EventSubMode=EventSubMode.SyncRead, *, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> int

Note

This function is heavily overloaded, and includes three additional signatures that have been deprecated (see warning below).

The client call to subscribe for event reception. There are two main categories:

  • Subscribe to events for a specific attribute, like change events, providing either:

    • a callback function for immediate notification (push callback model), or

    • a queue length, allowing events to be processed later (pull callback model).

  • Subscribe to device-level events (not linked to a specific attribute), like interface-changed events. These require a callback function (push callback model).

More details of the push and pull callback models are provided in the cppTango Events API docs.

A common example, subscribing to change events on the “voltage” attribute with a callback, is shown below.

from tango import DeviceProxy, EventData, EventType

def on_change_event(evt: EventData):
    if evt.err:
        e = evt.errors[0]
        print(f"Event error: [{e.reason}] {e.desc}")
    else:
        print(f"Event received: [{evt.attr_value.quality}] {evt.attr_value.value}")

# subscribe
dp = DeviceProxy("my/power_supply/1")
eid = dp.subscribe_event("voltage", EventType.CHANGE_EVENT, on_change_event)
...
# when done with subscription
dp.unsubscribe_event(eid)
Parameters:

  • attr_name (str) – The device attribute name which will be sent as an event, e.g., “current”.

  • event_type (EventType) –

    The event reason, which must be one of the enumerated values in EventType. This includes:

    • EventType.CHANGE_EVENT

    • EventType.ALARM_EVENT

    • EventType.PERIODIC_EVENT

    • EventType.ARCHIVE_EVENT

    • EventType.USER_EVENT

    • EventType.DATA_READY_EVENT

    • EventType.ATTR_CONF_EVENT

    • EventType.INTERFACE_CHANGE_EVENT

  • cb (typing.Callable) – Any callable object or an object with a callable push_event method (i.e., use the push callback model). The callable has the signature def cb(event) (or async def cb(event) for asyncio green mode DeviceProxy objects). The event parameter’s data type depends on the type of event subscribed to. In most cases it is tango.EventData. Special cases are EventType.DATA_READY, EventType.ATTR_CONF_EVENT, and EventType.INTERFACE_CHANGE_EVENT - see Event arrived structures.

  • queuesize (int) –

    The size of the event reception buffer (i.e., use the pull callback model). The event reception buffer is implemented as a round-robin buffer. This way the client can set up different ways to receive events:

    • Event reception buffer size = 1 : The client is interested only in the value of the last event received. All other events that have been received since the last reading are discarded.

    • Event reception buffer size > 1 : The client has chosen to keep an event history of a given size. When more events arrive since the last reading, older events will be discarded.

    • Event reception buffer size = tango.constants.ALL_EVENTS : The client buffers all received events. The buffer size is unlimited and only restricted by the available memory for the client.

  • sub_mode (EventSubMode) – The event subscription mode.

  • extract_as (ExtractAs) – In which format to return the attribute values. Default: ExtractAs.NumPy

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

An event id which has to be specified when unsubscribing from this event.

Return type:

int

Throws:

tango.EventSystemFailed: Raised in case of a failure in the event system.

tango.DevFailed: Raised in case of general communication failures.

TypeError: Raised in case of an incorrect type of input arguments.

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

Deprecated since version PyTango: 10.1.0

The following signatures, which use the filters and/or stateless parameters, are deprecated. The version for removal has not been decided, but the earliest is version 11.0.0.

  • subscribe_event(self, attr_name, event_type, cb, filters=[], stateless=False, extract_as=ExtractAs.Numpy, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> int

  • subscribe_event(self, attr_name, event_type, queuesize, filters=[], stateless=False, extract_as=ExtractAs.Numpy, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> int

  • subscribe_event(self, event_type, cb, stateless=False, *, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> int

filters (sequence<str>)

The filters apply to the original Notifd-based event system, not the current ZeroMQ-based event system (added in Tango 8). Notifd support is scheduled for removal in Tango 11.

A variable length list of name-value pairs which define additional filters for events. Provide an empty list, if this feature is not needed (typically, for all Tango servers from version 8).

Filters cannot be used with the sub_mode parameter.

stateless (bool)

Instead of setting stateless=True use sub_mode=EventSubMode.Stateless.

When this flag is set to false, an exception will be thrown if the event subscription encounters a problem. With the stateless flag set to true, the event subscription will not raise an exception, even if the corresponding device is not running, or the attribute doesn’t exist. A keep-alive thread will attempt to subscribe for the specified event every 10 seconds, executing a callback with the corresponding exception at every retry.

The stateless flag cannot be used with the sub_mode parameter.

Added in version 10.1.0: Three new signatures using the sub_mode parameter.

Changed in version 10.1.0: All parameters can now be passed as keyword arguments.

unsubscribe_event(event_id: int, *, green_mode: GreenMode = None, wait: bool = None, timeout: int = None) → None#

This method is a simple way to do:

self.get_device_proxy().unsubscribe_event(...)

For convenience, here is the documentation of tango.DeviceProxy.unsubscribe_event():

Unsubscribes a client from receiving the event specified by event_id.

Parameters:

  • event_id (int) – The event identifier returned by DeviceProxy::subscribe_event(). Unlike in TangoC++, this implementation checks that the event_id has been subscribed to in this DeviceProxy.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

None

Throws:

tango.EventSystemFailed: Raised in case of a failure in the event system.

KeyError: Raised if the specified event_id is not found or not subscribed in this DeviceProxy.

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

write(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().write_attribute(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.write_attribute():

write_attribute(self, attr, value, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> None

Write a single attribute.

Parameters:

  • attr (str | tango.AttributeInfoEx) – The name or AttributeInfoEx structure of the attribute to write.

  • value (typing.Any) – The value to write to the attribute.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Throws:

tango.ConnectionFailed, tango.CommunicationFailed, tango.DeviceUnlocked, tango.DevFailed: from device

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

Added in version 8.1.0: green_mode parameter. wait parameter. timeout parameter.

Changed in version 10.1.0: attr_name parameter was renamed to attr

write_asynch(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().write_attribute_asynch(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.write_attribute_asynch():

Write asynchronously the specified attribute.

Important

by default, TANGO is initialized with the polling model. If you want to use the push model (the one with the callback parameter), you need to change the global TANGO model to PUSH_CALLBACK. You can do this with the tango.ApiUtil.set_asynch_cb_sub_model()

Parameters:

  • attr (str | AttributeInfoEx) – an attribute name to write or AttributeInfoEx object (see Note below)

  • value (typing.Any) – value to write

  • cb (typing.Optional[typing.Callable]) – push model: as soon as attribute written, core calls cb with write results. This callback object should be an instance of a user class with an attr_written() method. It can also be any callable object.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

an asynchronous call identifier which is needed to get the server reply if poll model, None if push model

Return type:

int | None

Throws:

tango.ConnectionFailed

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

Important

Multiple asynchronous calls are not guaranteed to be executed by the device server in the same order they are invoked by the client. E.g., a call to the method write_attribute_asynch("a", 1) followed immediately with a call to read_attribute_asynch("a") could result in the device reading the attribute a before writing to it.

Note

There are two possibilities for the attr parameter: if you give attribute name, then PyTango will do additional synchronous(!) IO to fetch AttributeInfoEx object from device server, since we must know to which c++ data type cast python value. If you would like to avoid this IO you can give AttributeInfoEx instead of attribute name.

Changed in version 10.1.0:

  • attr_name parameter was renamed to attr

  • added support for AttributeInfoEx for attr parameter

write_read(*args, **kwds)#

This method is a simple way to do:

self.get_device_proxy().write_read_attribute(self.name(), ...)

For convenience, here is the documentation of tango.DeviceProxy.write_read_attribute():

Write then read a single attribute in a single network call. By default (serialisation by device), the execution of this call in the server can’t be interrupted by other clients.

Parameters:

  • attr (str | AttributeInfoEx) – The name or AttributeInfoEx structure of the attribute to write.

  • value (typing.Any) – The value to write to the attribute.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

A DeviceAttribute object with readout value.

Return type:

tango.DeviceAttribute

Throws:

tango.ConnectionFailed: Raised in case of a connection failure.

tango.CommunicationFailed: Raised in case of a communication failure.

tango.DeviceUnlocked: Raised in case of an unlocked device.

tango.WrongData: Raised in case of a wrong data format.

tango.DevFailed: Raised in case of a device failure.

TypeError: Raised in case of an incorrect type of input arguments.

Throws:

TimeoutError: (green_mode == Futures) If the future didn’t finish executing before the given timeout

gevent.Timeout: (green_mode == Gevent) If the async result didn’t finish executing before the given timeout.

Note

There are two possibilities for the attr parameter: if you give attribute name, then PyTango will do additional IO to fetch AttributeInfoEx object from device server, since we must know to which c++ data type cast python value. If you would like to avoid this IO you can give AttributeInfoEx instead of attribute name.

Added in version 7.0.0.

Added in version 8.1.0: green_mode parameter. wait parameter. timeout parameter.

Changed in version 10.1.0:

  • attr_name parameter was renamed to attr

  • added support for AttributeInfoEx for attr parameter

write_reply(*args, green_mode: GreenMode = None, wait: bool = None, timeout: int = None, **kwargs) → None#

This method is a simple way to do:

self.get_device_proxy().write_attribute_reply(...)

For convenience, here is the documentation of tango.DeviceProxy.write_attribute_reply():

write_attribute_reply(self, id: int, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> None write_attribute_reply(self, id: int, poll_timeout: int, green_mode: GreenMode=None, wait: bool=True, timeout: float=None) -> None

Check if the answer of an asynchronous write_attributes is arrived (polling model). If the reply is arrived and if it is a valid reply, the call returned. If the reply is an exception, it is re-thrown by this call. An exception is also thrown in case of the reply is not yet arrived.

Changed in version 10.0.0: To eliminate confusion between different timeout parameters, the core (cppTango) timeout (previously the optional second positional argument) has been renamed to “poll_timeout”. Conversely, the pyTango executor timeout remains as the keyword argument “timeout”. These parameters have distinct meanings and units:

  • The cppTango “poll_timeout” is measured in milliseconds and blocks the call until a reply is received. If the reply is not received within the specified poll_timeout duration, an exception is thrown. Setting poll_timeout to 0 causes the call to wait indefinitely until a reply is received.

  • The pyTango “timeout” is measured in seconds and is applicable only in asynchronous GreenModes (Asyncio, Futures, Gevent), and only when “wait” is set to True. The specific behavior when a reply is not received within the specified timeout period varies depending on the GreenMode.

Parameters:

  • id (int) – the asynchronous call identifier

  • poll_timeout (int, optional) –

    cppTango core timeout in ms:

    • If the reply has not yet arrived, the call will wait for the time specified (in ms).

    • If after timeout, the reply is still not there, an exception is thrown.

    • If timeout set to 0, the call waits until the reply arrives.

    • If the argument is not provided, then there is no timeout check, and an exception is raised immediately if the reply is not ready.

  • extract_as (ExtractAs) – Defaults to numpy.

  • green_mode (GreenMode, optional) – Defaults to the current tango GreenMode. Refer to get_green_mode() and set_green_mode() for more details.

  • wait (bool, optional) – Specifies whether to wait for the result. If green_mode is Synchronous, this parameter is ignored as the operation always waits for the result. This parameter is also ignored when green_mode is Synchronous.

  • timeout (float, optional) – The number of seconds to wait for the result. If set to None, there is no limit on the wait time. This parameter is ignored when green_mode is Synchronous or when wait is False.

Returns:

None

Return type:

None

Throws:

tango.AsynCall, tango.AsynReplyNotArrived, tango.ConnectionFailed, tango.CommunicationFailed, tango.DevFailed

Added in version 7.0.0.

tango.get_attribute_proxy(self, full_attr_name: str, green_mode: GreenMode = None, wait: bool = True, timeout: float = None) → AttributeProxy[source]#
tango.get_attribute_proxy(self, device_proxy: DeviceProxy, attr_name: str, green_mode: GreenMode = None, wait: bool = True, timeout: float = None) → AttributeProxy

Returns a new AttributeProxy. There is no difference between using this function and the direct AttributeProxy constructor if you use the default kwargs.

The added value of this function becomes evident when you choose a green_mode to be Futures or Gevent. The AttributeProxy constructor internally makes some network calls which makes it slow. By using one of the green modes as green_mode you are allowing other python code to be executed in a cooperative way.

Parameters:

  • full_attr_name (str) – the full name of the attribute

  • device_proxy (DeviceProxy) – the DeviceProxy

  • attr_name (str) – attribute name for the given device proxy

  • green_mode (GreenMode) – determines the mode of execution of the device (including the way it is created). Defaults to the current global green_mode (check get_green_mode() and set_green_mode())

  • wait (bool) – whether or not to wait for result. If green_mode Ignored when green_mode is Synchronous (always waits).

  • timeout (float) – The number of seconds to wait for the result. If None, then there is no limit on the wait time. Ignored when green_mode is Synchronous or wait is False.

Returns:

if green_mode is Synchronous or wait is True:

AttributeProxy

elif green_mode is Futures:

concurrent.futures.Future

elif green_mode is Gevent:

gevent.event.AsynchResult

elif green_mode is Asyncio:

asyncio.Future

Throws:
DevFailed if green_mode is Synchronous or wait is True

and there is an error creating the proxy.

concurrent.futures.TimeoutError if green_mode is Futures,

wait is False, timeout is not None and the time to create the device has expired.

gevent.timeout.Timeout if green_mode is Gevent, wait is False,

timeout is not None and the time to create the device has expired.

asyncio.TimeoutError if green_mode is Asyncio,

wait is False, timeout is not None and the time to create the device has expired.

Added in version 8.1.0.

Contents