What's the best practices for indexing XML Content with dynamic XML Elements (SOLR 6.1) ?

[Stan Lee]

We currently have a Microsoft SQL table with a XML datatype. We use DIH to
import the XML Content as is, that is not using the XPathEntityProcessor.
If the elements of the XML content is known, XPathEntity make sense. Could
someone kindly suggest the right way of handling such scenario, without
impacting search performance?
Which tokenizer should we be using?


Thanks.