Dataloader problematic character in feild used for file name

[Hrovatin]

I tried to make a dataloader with multiple species across files. Thus I put "/" in the filed "Dataset-wide main organism annotation (from NCBItaxon)" as this was a required filed in the dataloader creation, but I wanted to edit it manually at the end. However, no loader was created due to PermissionError: [Errno 13] Permission denied: '/_x_2021_10x3v3_chen_001.py' This happens because I have put "/" into one of the fields that are used for filenames. Could you add a check that all characters form the fields you use for filename are valid?

[davidsebfischer]

Why did you want to use "/" there or is this question about this mistaken entry not messing up phase 1a, ie more lenient error catching? I do not think it is the right field value here, it should be left empty I think and the later filled in phase 1b as a multi sample loader. More lenient error catching we can implement, good idea! You identified the problem correctly that "/" is problematic because this string is used to build file names.

[Hrovatin]

The organism in sfaira command line tool is necessary at this point so I needed to put something for the start and I just put "/" at random, hoping to change it later manually in yaml. Yes, the more informative error message would have been useful.