Community
Participate
Working Groups
Build Identifier: 20100618-0524 The .dot to Zest parser seems sensitive to the input ordering of nodes and edges; according to the DOT specification there is no requirement that nodes precede edges in the input file. The two attached .dot files produce differing graphs, although semantically equivalent, show one with apparent node duplication. This is not a purely theoretical issue, as the problem .dot file was produced by the Grand Ant visualization facility, when called as an Ant task (see http://www.ggtools.net/grand/#usage ). Reproducible: Always Steps to Reproduce: View these .dot files in the Zest view: they should produce identical results.
Created attachment 189331 [details] Shows disconnected nodes
Created attachment 189332 [details] Shows nodes correctly connected.
I wanted to attach the dot4zest attribute to this bug, but couldn't see how to: apologies to anyone spammed.
Created attachment 189334 [details] Original build.xml file. For information.
Thanks for your report. Fixed in Zest 2.0 master and nightly p2 repository (https://hudson.eclipse.org/hudson/job/gef-zest-nightly/lastSuccessfulBuild/artifact/targetPlatform/), see TestGraphInstanceDotImport#nodesAfterEdges()