http://documentation.metawrap.com/tests/dot/test3.png (generated automagically)
http://documentation.metawrap.com/tests/dot/test.png (the original target plan)
It lurches into being.
I’m thinking that I may have to write this with MwW3ParserLexerState being the node instead of the current plan for MwW3ParserLexerStateGroup. The issue at hand is that a MwW3ParserLexerStateGroup can consist of multiple MwW3ParserLexerState’s, but its each MwW3ParserLexerState that represents a point state within the grammar – and its these that will give me my lovely dot notation description in the digraph nodes.
http://documentation.metawrap.com/tests/dot/test4.png is a tidy version of test3 which is a state group based walk of the parse table.
http://documentation.metawrap.com/tests/dot/test5.png is a strange mutant graph of a state based walk of the parse tables.. it actually represents something very interesting but mostly useless unless you are into quantum computing 🙂
http://documentation.metawrap.com/tests/dot/test6.png is state based but with filtering duplicate edges. Compared to test5, the world lines no longer represent the set of all possible paths, but simply the transitions that can be used to build each path.