Ticket8445 handle removal of defaults from config

[LowriJenkins]

Description of work

Add a check when adding macros to the blockserver configs to not create them if the gui passes them with the useDefault value set to true. And run ruff format and check.

To test

Which ticket does this PR fix?

Acceptance criteria

List the acceptance criteria for the PR

---

Code Review

• Is the code of an acceptable quality?
• Has the author taken into account the multi-threaded nature of the code?
• Have the changes been recorded appropriately in a PR for release notes?
• Has the manual system tests spreadsheet been updated?

Functional Tests

• Do changes function as described? Add comments below that describe the tests performed.

Final steps

• Reviewer has updated the submodule in the main EPICS repo? See Reviewing work for the subModules of EPICS in the Git workflow page for details.
• Reviewer has merged the associated PR for the release notes

[Tom-Willemsen]

Unfortunately, this branch completely crashes my local blockserver with:

Exception in thread Thread-11 (consume_write_queue):
[2026-02-12 17:33:18.604585] MINOR: ComponentSwitcher: component_switcher config file at C:/Instrument/Settings/config/NDW2922/configurations\ComponentSwitcher\component_switcher.json does not exist - assuming empty config
Traceback (most recent call last):
  File "C:\Instrument\Apps\Python3\Lib\threading.py", line 1043, in _bootstrap_inner
    self.run()
    ~~~~~~~~^^
  File "C:\Instrument\Apps\Python3\Lib\threading.py", line 994, in run
    self._target(*self._args, **self._kwargs)
    ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Instrument\Apps\EPICS\ISIS\inst_servers\master\\block_server.py", line 737, in consume_write_queue
    self.update_server_status(state)
    ~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^
  File "C:\Instrument\Apps\EPICS\ISIS\inst_servers\master\\block_server.py", line 678, in update_server_status
    raise RuntimeError
RuntimeError
Traceback (most recent call last):
  File "C:\Instrument\Apps\Python3\Lib\site-packages\pcaspy\driver.py", line 631, in getValue
    newValue = driver.read(self.info.reason)
  File "C:\Instrument\Apps\EPICS\ISIS\inst_servers\master\\block_server.py", line 288, in read
    raise RuntimeError
RuntimeError
Traceback (most recent call last):
  File "C:\Instrument\Apps\Python3\Lib\site-packages\pcaspy\cas.py", line 582, in read
    return _cas.PV_read(self, ctx, protoIn)
           ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^
RuntimeError: SWIG director method error. Exception Calling Python Code
RuntimeError: SWIG director method error. Exception Calling Python Code

The above exception was the direct cause of the following exception:

SystemError: <function SimplePV.interestDelete at 0x000001A39A011BC0> returned a result with an exception set

This crash does not occur on the master branch.

[Tom-Willemsen]

I'm somewhat sceptical about importing genie_python in the blockserver... seems like a recipe to get conflicting channel access libraries etc.

Suggested change

	from genie_python.genie import PVValue
	type PVBaseValue = bool | int | float | str
	type PVValue = PVBaseValue | list[PVBaseValue] | NDArray | None

[LowriJenkins]

would it be more sensible to handle this in a if TYPE_CHECKING: block rather than redefining the type?

[Tom-Willemsen]

My preference would be to avoid the dependency on genie_python alltogether - it's quite a big dependency to pull in for a 2-line type definition. But if you really prefer then this shouldn't block the merge if you use if TYPE_CHECKING

[LowriJenkins]

Ok, I guess I'm just unsure when avoiding a large dependency import is more important than avoiding code duplication, especially in this case as I believe the pyright error in the system tests shows that we do actaully need to alter our PVValue definition to allow the Byte type (The old error in the tests was caused by me passing the bytes as a string)

[Tom-Willemsen]

An alternative would be to factor out the definition of PVValue into ibex-non-ca-helpers, and depend on that common definition both here and in genie_python.

[LowriJenkins]

I might create a friday ticket to do this, as it would require changing this everywhere we import it?

[Tom-Willemsen]

Not sure it's actually that many places outside genie_python, but yeah. Up to you - I don't want to get too hung up on this review comment, the main blocker was the comment above about blockserver crashing, this is a nicety.

[LowriJenkins]

doing some more bugfixing for some errors introduced during pyright/ruff have led to another file needing pyright checking, but once thats done this should be ready to go