reflecting the scopes and folding in xref generation

[tulinkry]

fixes #1627

[tarzanek]

I think you just miss a fix for unit tests ... otherwise it looks OK

[tulinkry]

I adapted the tests by turning off scopes and I added one tests where scopes are enabled

[vladak]

From maintenance perspective I don't like the addition of verbatim HTML code too much. Ideally it should be parsed to be get the scopes.

[tulinkry]

For test purpose this seems ok to me now, taken into account that I only followed the other test case in this package. Of course a kind of less strict test would be welcomed.

[vladak]

As far as I understand this and AnalyzerGuru.java#writeXref() are used only for testing. Otherwise during normal indexing the xref is created via this path:

IndexDatabase#addFile()
  FileAnalyzer fa = AnalyzerGuru.getAnalyzer() // fa can be e.g. CAnalyzer
  AnalyzerGuru#populateDocument()
    fa.analyze()
    AbstractSourceCodeAnalyzer#analyze()
      super.analyze()
        PlainAnalyzer#analyze()
          writeXref()
            xref = newXref() // xref can be e.g. CXref
            JFlexXref.write()

where the scopes/folding properties are in effect thanks to the inheritance and these lines in IndexDatabase#addFile():

662          fa.setScopesEnabled(RuntimeEnvironment.getInstance().isScopesEnabled());
663          fa.setFoldingEnabled(RuntimeEnvironment.getInstance().isFoldingEnabled());

So having these properties set into the writeXref() seems like short-circuiting however there is no other way given the method is static.

The only nit I have is that the call to reInit() in JFlexXref above contains:

        scopes = new Scopes();

which is not necessary when scopes are disabled. Granted, this method does not do anything however it's a bit of wasted memory.

[vladak]

is this fixing any misbehavior ?