Closed GoogleCodeExporter closed 8 years ago
I think the way to go is to make a separate grapher class that knows how to
graph a
struct array with a particular format. This grapher could produce GraphViz
graphs
for a few different purposes:
- The profiler output (includes collaboration information)
- The ObjectGrapher (object references)
- Statelist diagrams (nonexistent, yet)
Each element of a struct array would correspond to a graph node.
A node naming function would take one struct element and generate a node name
(by looking at a particular field, or whatever).
An index into the struct array would specify an edge. A certain field of would
specify
where to look for edges from one struct element to others.
An edge naming function would take two struct elements and their indices, and
comu
up with a name (by looking at a particular field, or whatever)
It will probably be clearer to put edge names in rows of a "record" node rather
than
out in edge space. To do this, it would be necessary to process all edges
before
writing a dot file.
Original comment by Benjamin.Heasly
on 16 Feb 2010 at 11:10
Although this is not critical functionality, it should be a key part of the
user experience-
-the main point of Snow Dots.
Original comment by Benjamin.Heasly
on 16 Feb 2010 at 11:11
This kind of thing is addressed in the utilities/graphing folder of Tower of
Psych, so it's no longer a Snow Dots issue.
Original comment by Benjamin.Heasly
on 23 Jun 2010 at 11:29
Original issue reported on code.google.com by
Benjamin.Heasly
on 29 Jan 2010 at 5:34