SSE with different LabelToNode instance

[cygri]

This is a request for a way to use the SSE class with a ParseHandler that uses LabelToNode.createUseLabelEncoded() instead of the hardcoded LabelToNode.createScopeGlobal().
Perhaps:

• Add a new method public static SSE.setParseHandler(ParseHandler handler).
• Add a method public ParseHandlerPlain.setLabelToNode(LabelToNode bNodeLabels) to set the bNodeLabels field (or make emitBNode(...) non-final so that it can be overridden to use a different LabelToNode instance).

Does this sound reasonable? I'm happy to submit a PR with these changes, or with any other API changes that would enable my goal.

[afs]

SSE is old. It ought to take a FactoryRDF (usually as part of a ParserProfile) to make nodes. These can have a specific label mapping.

What's the use case? SSE is primarily for Jena's tests.

[cygri]

Same use case, making objects for use in unit tests.

[afs]

<_:ThisIsTheLabel> is parsed to _:ThisIsTheLabel

It may print as _:BThisIsTheLabel in NT but the label is "ThisIsTheLabel".