rgerhards
commented
8 years ago That's it for now. Read comments in above commit for more details. To save a thought: we may think about using the parser ids as nodes, and do not have nodes shown in the dot plot for the actual pdag nodes. Thus the tree would represent the potential flow of parsers, which should probably convey the idea very well. Not enough time left right now to try this out.
For extended parsers, we should add "edge node" -- that enables us to convey more information and have e.g. "repeat" show it's subdag. They should be drawn in a different box style (dotted?).