`cubi-tk sodar add-ped` expects the `Library Name` Column to exist

[Nicolai-vKuegelgen]

Describe the bug
The Sodar / cubi-tk isa-tab template have been modified not use the Library Name column anymore, since this is not actually ISA compliant.
However, the cubi-tk sodar add-ped function will treat all Extract Name column the same and only write the expected content of the "last material column" into a column named Library Name.

To Reproduce
Steps to reproduce the behavior:

1. Create a new DNA germlin (or cancer?) samplesheet with Sodar GUI or cubi-tk
2. Use the template in a Sodar project
3. Upload new sample data using cubi-tk sodar add-ped
4. Both Extract Name columns have the same content

Expected behavior
The 2nd Extract Name Column should be treated the same as Library Name

Additional context
I think recent changes to cubi-tk snappy pull-samplehseet already contain the code that can fix this.
We should also consider re-factoring all the ISA graph parsing code that is used/overwritten by different functions.