Feature/keithley 2000 scan

[marcelhohn]

New features:

• added class for Keithley 2000-SCAN card
• added scanning card to Keithely_6500 class
• Keithley 2000-SCAN can be included into other DMM classes if necessary
• Keithley DMM6500: Check if correct terminal is activated
• Keithley DMM6500: Allow for simpler access of parameters (e.g. dmm.resistance.get() instead of dmm.measure.res.get(). The latter still works to ensure "downwards" compatibilty.)

[FarBo]

@MarceloSemO

Thank you for the contribution. I will start the review ASAP.

[FarBo]

I have kind of general comments from skimming in the files:

1. Please provide a little bit of docstring for classes and functions (ideally, all arguments in the classes and functions could have some descriptions)
2. All defined parameters in both .py files should have validators specifying valid data for that parameter and min/max if required (you can find validators in qcodes.utils.validators)
3. Please provide typing for all arguments

[marcelhohn]

Thanks for the feedback, I added docstrings and typings. Two problems/questions are remaining:

1. Should parameter which are only gettable also have a validator? I am asking because all settable parameters in my drivers already have validators, but you still asked me to add validators to the parameters. As far as I understood, the purpose of a validator is to prevent to send values which could cause unpredictable behavior of the device. In case of non-settable parameters this cannot happen anyways. Please correct me if I am mistaken and if I should also add validators to non-settable parameters.

2. The test now fails, because the autodoc fails to import the module Keithley_2000_Scan. I don't know how to solve this, since the only import of the Keithley_2000_Scan module happens in the Keithley_6500 module, and I did not change the import statement there. What I changed is that I imported the Keithley_6500 module in the Keithley_2000_Scan module, since I needed this to provide the desired type hint for the variable "dmm" in the __init__ function of the Keithley_2000_Scan_Channel class. My guess is that the circular import is what causes the test to failm but I don't know how to fix, so I would appreciate your help.

[jenshnielsen]

@MarceloSemO I pushed a small fix for the circular import. Basically you need to only import the Parent instrument into the scan module when you are typechecking

[jenshnielsen]

I agree that validators are not needed for get only parameters. I left a few suggestions inline

[FarBo]

@MarceloSemO

Regarding the first question, there is no need for validators for only-gettable parameters.

Thanks for the changes. I am reviewing it in a continuous basis.

[marcelhohn]

@jenshnielsen and @FarBo, thank you very much for your comments! I have implemented the changes you suggested and performed a rebase with respect to the latest commit on the master branch. I think the pull request should now be ready to be merged.

[marcelhohn]

I have now re-run the notebook and implemented the required changes. I also performed a rebase, now the PR should be ready to be merged.