missing edgecolormap in BFS traversal

[CarloLucibello]

while it is present in DFS.

Maybe there is a way, to maximize efficiency, to choose if edge discovery has to be considered or not

[jpfairbanks]

Do you have more details?

[CarloLucibello]

on this?

Maybe there is a way, to maximize efficiency, to choose if edge discovery has to be considered or not

no! just whishfull thinking :)

[sbromberger]

I don't think this is correct. We have calls to open_vertex, discover_vertex!, and examine_neighbor in the BFS code. Am I misunderstanding your issue?

[CarloLucibello]

@sbromberger you are right, actually I misundertstood the issue. Actually what is missing in BFS with respect to DFS is an edgecolormap, whose elements should be passed to examine_neighbor. At the moment examine_neighbor in BFS always takes a -1 dummy argument.

This could be fixed easily, I'll do it along the other changes in #332

@jpfairbanks For traversals not needing an edgecolormap maybe one could define a dummy dict-like type for edgecolormap that doesn't allocate memory and doesn't perform any computation...

[sbromberger]

can we close this with the merge of #332?

[CarloLucibello]

Yep