Introduce find_unlinked_contexts_in_context() helper

[Enselic]

Glad to hear you liked my PoC for lazy-loading syntaxes! No worries about having a busy life :)

Here comes another PR. This one is pretty simple. But it will make the upcoming work and diff for #374 easier to do and read.

The code can be further simplified and refactored, but spare time is limited, and at this point I'd prefer to prioritize getting lazy-loading production ready.

First commit splits out a helper function (ignore white space when diffing to get a nice diff), and the second commit runs cargo fmt on the new helper function.

[Enselic]

I would like to give an overview of the journey ahead of us. The plan is to split up the prototype quality code in the PoC into 6 incremental steps. One PR per step, with the code turned into production quality code. The steps are:

• Make ContextId implement Hash and simplify the code #377
• Introduce find_unlinked_contexts_in_context() helper #378
• Rename contexts in SyntaxReference to context_ids #379
• Move contexts from SyntaxSet to individual SyntaxReferences
• Lazy-load the context_ids and contexts of each SyntaxReference
• Stop compressing the "toplevel" SyntaxSet when serializing it

And then, we're done ✅