Create a new device class for tcavs #261

phys-cgarnier · 2025-04-18T21:09:19Z

TCAV Device class:
Changed generator.py to include extract_tcav function. Also implemented advanced filtering in extract devices. You can pass an unpacked dictionary that is organized by {field_name: expected value}
For example :
additional_filter_constraints = {"Engineering Name": "TRANS_DEFL"}
Added relevant fields to metadata.py

…sing two pvs for fb state

…t on mcclogin

…ther tcavs are different

eloise-nebula

Hi Chris, looks consistent with our other device classes. I found a bunch of places where pydantic validation could replace your manual validation. This device is unique because its configuration depends on its mode_config, but I think that's handled well here. Once we figure out the validation stuff.

lcls_tools/common/devices/tcav.py

eloise-nebula · 2025-04-24T21:28:13Z

lcls_tools/common/devices/tcav.py

+                    options = [options_to_check]
+                for option in options:
+                    if option not in self.controls_information.mode_config_options:
+                        print(


I think an error here would be more appropriate. As it stands, the check_options decorator doesn't do anything to communicate to the code that something went wrong.

eloise-nebula · 2025-04-24T21:28:42Z

lcls_tools/common/devices/tcav.py

+
+        def decorated(self, *args, **kwargs):
+            if self.mode_config == "Disable":
+                print("Unable to perform action, TCAV is in Disabled state")


Another place where an error might be better than a print statement.

eloise-nebula · 2025-04-24T21:48:19Z

lcls_tools/common/devices/tcav.py

+
+    @amp_set.setter
+    @check_state
+    def amp_set(self, amplitude):


Pydantic lets you validate function arguments with function annotations like so:

def amp_set(self, amplitude: float):

The if statement shouldn't be necessary.

I was wrong about this, the annotations require you to wrap your functions with another decorator. Your current pattern of checking in place is fine.

eloise-nebula · 2025-04-24T21:49:02Z

lcls_tools/common/devices/tcav.py

+    @amp_fbenb.setter
+    @check_state
+    def amp_fbenb(self, state: Union[str, int]):
+        if not isinstance(state, str) or not isinstance(state, int):


Pydantic should already be validating this.

Sorry, the expected type is actually an enum_str, I couldn't get the validation to work how I thought I could.
Unless I made the properties pydantic objects

eloise-nebula · 2025-04-24T21:49:25Z

lcls_tools/common/devices/tcav.py

+    @phase_fbenb.setter
+    @check_state
+    def phase_fbenb(self, state: Union[str, int]):
+        if not isinstance(state, str) or not isinstance(state, int):


Another place where pydantic covers function validation.

eloise-nebula · 2025-04-24T21:49:38Z

lcls_tools/common/devices/tcav.py

+    @amp_fbst.setter
+    @check_state
+    def amp_fbst(self, state: Union[str, int]):
+        if not isinstance(state, str) or not isinstance(state, int):


Same as above.

eloise-nebula · 2025-04-24T21:49:48Z

lcls_tools/common/devices/tcav.py

+    @phase_fbst.setter
+    @check_state
+    def phase_fbst(self, state: Union[str, int]):
+        if not isinstance(state, str) or not isinstance(state, int):


Same as above.

lcls_tools/common/devices/tcav.py

kabanaty · 2025-04-28T18:08:06Z

lcls_tools/common/devices/tcav.py

+    controls_information: SerializeAsAny[TCAVControlInformation]
+    metadata: SerializeAsAny[TCAVMetadata]
+
+    def __init__(self, *args, **kwargs):


I don't think you need an __init__ function when using Pydantic

kabanaty

Looks good in general. Some cleanup to be done with implementing Pydantic validation in some of the TCAV properties.

eloise-nebula

Hi Chris, thanks for the thoughtful update. I like the new pattern for checking the amplitude and phase feedback options a lot better.

I was wrong when I said the validation should just require annotations like this def foo(bar: int). They need a decorator too, and it looks like we only do that in one place in our code:

lcls-tools/lcls_tools/common/devices/magnet.py

Lines 38 to 40 in fbaaa7f

    
           @field_validator("*", mode="before") 
        
           def validate_pv_fields(cls, v: str) -> PV: 
        
               return PV(v)

If you want to do validation still, I would suggest you just throw ValueError in the appropriate places. Either that or remove the silent failures.

Thank you!

eloise-nebula · 2025-05-01T21:20:39Z

lcls_tools/common/devices/tcav.py

+            )
+
+        self._mode_config_options.update(
+            {option: i for i, option in enumerate(mode_config_options["enum_strs"])}


I like this pattern better than the old one, nice.

eloise-nebula · 2025-05-01T21:21:06Z

lcls_tools/common/devices/tcav.py

+                return setter_method(self, enum_str)
+            return decorated
+        return decorator
+    """


This can be removed I think.

eloise-nebula · 2025-05-01T21:21:52Z

lcls_tools/common/devices/tcav.py

+
+    @amp_set.setter
+    @check_state
+    def amp_set(self, amplitude):


I was wrong about this, the annotations require you to wrap your functions with another decorator. Your current pattern of checking in place is fine.

eloise-nebula · 2025-05-01T21:24:39Z

lcls_tools/common/devices/tcav.py

+        self.set_amplitude_feedback_options()
+        self.setup_phase_feedback_option()
+
+    def set_mode_config_option(self):


Maybe this should be called get_mode_config_option?

eloise-nebula · 2025-05-01T21:26:48Z

lcls_tools/common/devices/tcav.py

+            }
+        )
+
+    def setup_phase_feedback_option(self):


This should be consistent with the above set_amplitude_feedback_option too I think.

…ponding row in lcls_elements.csv

phys-cgarnier added 19 commits March 24, 2025 15:41

updated python files to account for tcav

4eb4a59

added stcav2 to lcls_elements.csv, some fields may be in correct

932bc05

Merge branch 'main' into tcav

60b57f5

fixed some import errors

60189ec

committing change to lcls_elements.csv

73b9596

added tcavs to yaml and updated generate code.

e55cbc9

added comment about problematic use case

4fb9447

flake8

b9a9295

updating PVs for generate.py

9c3c9f5

initial commit of skeleton code for tcav devices

4b23307

changed PVs in skeleton code

24d7482

fixed a typo in extract tcav

cefa0d1

Merge branch 'main' into tcav

6be4e83

merged main into branch, updated metadata fields for tcav

35bf110

ran generation script to pick up tcavs

65bd63d

added some properties for getting and setting PVs, realized I am mis…

9957648

…sing two pvs for fb state

built out tcav class, changed some pvs in generate, need to run scrip…

6750b19

…t on mcclogin

added mode_config pv decorator, need to build out the setter more

18449b3

linter

e78ce3f

phys-cgarnier linked an issue Apr 18, 2025 that may be closed by this pull request

Create a new device class for TCAVs #254

Closed

phys-cgarnier added 6 commits April 18, 2025 14:21

linter....

0cb374c

used pre-commit this time..

8efd654

updated yamls to have new PVs

441c57b

added a way to set modecfg value. added docstrings

f9c9ff3

set up write to only write tcav entries on DIAG0 area

57246b8

removed tcavs that are not stcav2, this is because the controls for o…

d2993ce

…ther tcavs are different

phys-cgarnier requested review from nneveu and kabanaty April 22, 2025 20:41

bug in check_options tcav function, should be resolved

9c7f53b

phys-cgarnier changed the title ~~254 create a new device class for tcavs~~ Create a new device class for tcavs Apr 22, 2025

phys-cgarnier requested a review from eloise-nebula April 22, 2025 22:47

changed rf frequencey from a str to a float

13eb26a

eloise-nebula reviewed Apr 24, 2025

View reviewed changes

kabanaty reviewed Apr 28, 2025

View reviewed changes

phys-cgarnier added 3 commits April 29, 2025 09:51

Merge branch 'main' into tcav

e21dc28

updated tcav.py to reflect requested changes in PR review

910043e

removed tcav collection from import since it will never be used

8a84326

eloise-nebula approved these changes May 1, 2025

View reviewed changes

changed enum strs properties for device, device name, and its corres…

d53c42a

…ponding row in lcls_elements.csv

phys-cgarnier merged commit 8c73f90 into main May 5, 2025
12 checks passed

phys-cgarnier deleted the 254-create-a-new-device-class-for-tcavs branch July 1, 2025 18:09

	@field_validator("*", mode="before")
	def validate_pv_fields(cls, v: str) -> PV:
	return PV(v)

Create a new device class for tcavs #261

Create a new device class for tcavs #261

Uh oh!

Conversation

phys-cgarnier commented Apr 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

eloise-nebula left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kabanaty left a comment

Choose a reason for hiding this comment

Uh oh!

eloise-nebula left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

phys-cgarnier commented Apr 18, 2025 •

edited

Loading