https://svn.boost.org/trac10/ticket/277 <pre class="wiki">During some experiments with the BGL-GraphViz interface, I observed a significant asymmetry between a given DOT graph input with boost::read_graphviz() and output with boost::write_graphviz() (without changing any attributes). Given the input DOT file &lt;metro.dot&gt; (which is, of course, much longer with the full set of information), the elementary BGL program &lt;metro.cxx&gt; transforms this into the output DOT file &lt;metro_boost.dot&gt; (the three files are attached to this submission). It seems that somewhere on the way, the graph and node (and edge) attributes get clobbered. Note that even the graph name "metro" gets replaced by the name of the (last) subgraph "ligne_1", and /all/ node attributes get mixed up. This seems to be somewhat similar to the recent edge-related problem posted by Hideaki Hiraki to the boost-users list. I have not (yet) tried to mess with the graphviz parser code, so I am asking here for help. Thanks. Greetings, Andreas &lt;andreas dot scherer at rb-its dot de&gt; </pre> en-us /htdocs/site/boost.png https://svn.boost.org/trac10/ticket/277 Trac 1.4.3 Douglas Gregor Thu, 09 Sep 2004 14:11:04 GMT https://svn.boost.org/trac10/ticket/277#comment:2 https://svn.boost.org/trac10/ticket/277#comment:2 <ul> <li><strong>status</strong> <span class="trac-field-old">assigned</span> → <span class="trac-field-new">closed</span> </li> </ul> <pre class="wiki">Logged In: YES user_id=249098 The CVS version of the BGL-GraphViz interface now contains a new GraphViz parser (based on Spirit) that can retain all attributes when parsing a graph, along with a writer that can reproduce those attributes. </pre> Ticket nobody Wed, 23 Mar 2005 21:04:26 GMT