feat: add cubi-tk sodar update-samplesheet

[Nicolai-vKuegelgen]

New function to allow flexible updates of ISA samplesheets. Also introduce a base class to implement SodarAPI features, as the currently used sodar_cli package is heavily outdated.
Features:

• can read ped files and/or command line defined sample specific (meta)data
• mapping of all ped/sampledata fields to ISA columns can be speciefied. Dynamic/derived columns are also possible
• default settings for the ISA germline-DNA-sequencing sheet or the new Modellvorhaben project
• global metadata can be defined for all samples at once
• both adding new entries and updating entries is possible (--overwrite required for updating non-empty values)
• Values are mapped to the ISA table based on column name matching, parsing/validation of the ISA graph structure is NOT done by this function (but errors will be returned by the Sodar API anyway)

TBD:

• tsv or similar input for collected sampledata in another file format than ped

Note: this is a potential replacement/side-grade of cubi-tk sodar add-ped

[github-actions]

Please format your code with black: make black.

[Nicolai-vKuegelgen]

Missing so far:

• tests for the new SodarAPI class
• smoke test for the while update-samplesheet workflow (execute)

[tedil]

I'm not too familiar with ISA, so most of my comments are more from a coding perspective and some general remarks.
I really like all the sanity checks and validations that you do here.

[tedil]

LGTM, the only thing I'd like to see as a little bonus is an example in the documentation/manual or so.

[tedil]

I suspect this might behave unexpectedly depending on shell/quoting?

[tedil]

Ah, nvm, arg_list is forwarded as is to argparse.

[Nicolai-vKuegelgen]

ah sry, didn't fully read your comment, I can add something to the documentation though