Use unique indices to identify datasets for appending

[bkloeckl]

Change the append dialog, get_compatibles and append_data functions to include original and unique indices used for identifying the data.

[cbrnr]

Thanks @bkloeckl! The current changes lack the ability to sort datasets (via drag and drop) in the Source and Destination lists though, which I think we should not remove.

I would also keep it simpler, instead of using a table, I'd stick to a list, but just add the index to the label. This could also be done for the sidebar then. I'm not sure how and if such a label can be formatted nicely, e.g. it would be nice to have

• numbers right-aligned
• numbers use some lighter color (not black or white)

Something like this:

If this requires a table (which might very well be the case), I'm sure that the style can be adapted so that it doesn't contain a header row or any borders.