Dataset create raises `Dataset exists` for similar names

[jsam]

There is a bug on dataset create command for datasets with long names and high similarity score.

Example:

renku dataset create "my super awesome dataset research project with super long name 1" 

results in

Use the name "my_super_awesome_dataset" to refer to this dataset.
OK

Now we create another one:

renku dataset create "my super awesome dataset research project with super long name 2"

results in

Error: Dataset exists: "my_super_awesome_dataset".

Expected behavior is that both of those datasets should be created.

[mohammad-alisafaee]

We cannot avoid this fully because we create short_name from name. Even if we remove the length limitation we cannot do something meaningful regarding special character: the following two commands create the same short_name and the second one will fail:

renku dataset create 'some name'
renku dataset create '#some name#'

One solution will be to treat name as short_name (i.e. like before) in renku dataset create and pass a human-friendly name with --name or --title. @rokroskar what do you think?

[rokroskar]

@mohammad-sdsc yes this (I think) was sort of along the lines of what I was advocating for before at some point. It has the added benefit of solving the issue that @emmjab has brought up, which is that the user creates a dataset with some name but then the name they need to use to refer to that dataset is somehow different.