Upgrade to XSLT Version 2 or 3

[snsb-seifert]

Currently the XSLT is set to Version 1 which makes splitting stings cumbersome. Would it matter to switch to Version=2 or Version=3? From version 2 on the handy function tokenize() could be used for this.

[DavidFichtmueller]

Now that the processing of the XSLT is done with the Saxon processor in the Jupyter Notebook files, we can switch it to XSLT 3. Maybe we should keep the version 1 in the repository though, for people who need to process it on their machines and can not install Saxon or a similar good processor, even if this offers a limited set of functionality.