Re: pfuTraverse

New Message Reply Date view Thread view Subject view Author view

John Rohlf (jrohlf++at++tubes)
Mon, 19 Jun 95 14:11:01 PDT

         
        I believe there was a bug in pfuTraverser. The new Inventor
loader (sgigate:/usr/people/ftp/pub/Performer/src/pfiv1.6.tar.Z or
pfiv_src1.6.tar.Z for source only)
contains a more robust pfuTraverse although it may be called
pfdTraverse.

>
>
> Hi,
>
> I traverse a Performer scene graph using pfuTraverse. Here are some
> code fragments.
>
> (1) setup traversal:
>
> ...
>
> pfuTraverser traverser;
>
> // init traverser struct
> pfuInitTraverser (&traverser);
>
> // assign callbacks
> traverser.postFunc = postCallback;
> traverser.preFunc = preCallback;
>
> ...
>
> // start traversal
> pfuTraverse ((pfNode*) pfRoot, &traverser);
>
>
> (2) preCallback()
>
> static long
> preCallback (pfuTraverser *traverser)
> {
> ...
>
> // debugging messages
> cerr << "preCallback(): " << pfGetTypeName ((pfObject *) traverser->node)
> << pfGetType (traverser->node);
>
> ...
> }
>
> (3) postCallback
>
> static long
> postCallback (pfuTraverser *traverser)
> {
> ...
>
> // debugging messages
> cerr << "postCallback() " << pfGetTypeName ((pfObject *) traverser->node)
> << pfGetType (traverser->node);
>
> ...
>
> }
>
>
> The data I use (read as Inventor file) include several pfGroup's with
> pfGeode's.
>
> Everything works fine in the preCallback(), in the postCallback() the
> traverser struct passed as a parameter seems to be wrong. This happens
> as soon as the PostCallback() is called for a pfGroup. The
> traverser->node member contains the same pointer as in the previous
> call of the postCallback(), when it was called for the pfGeode, which
> is contained in the pfGroup.
>
> A pfuTravPrintNodes() output for a simple data example looks like this:
>
>
> Node bird.iv: type: pfSCS = d0064
>
> Node (null): type: pfGroup = 50064
>
> Node (null): type: pfGeode = 30064
> num GeoSets: 1
> GeoSet 0: OVERALL - 2 prims, cbind=OVERALL, nbind=PER_VERT, tbind=OFF flatshade=1
> GeoState: 0x2
>
> Node (null): type: pfGeode = 30064
> num GeoSets: 1
> GeoSet 0: OVERALL - 1 prims, cbind=OVERALL, nbind=PER_VERT, tbind=OFF flatshade=1
> GeoState: 0x2
>
> Node (null): type: pfGeode = 30064
> num GeoSets: 1
> GeoSet 0: OVERALL - 1 prims, cbind=OVERALL, nbind=PER_VERT, tbind=OFF flatshade=1
> GeoState: 0x2
>
> Node (null): type: pfGroup = 50064
>
> Node (null): type: pfGeode = 30064
> num GeoSets: 1
> GeoSet 0: OVERALL - 3 prims, cbind=OVERALL, nbind=PER_VERT, tbind=OFF flatshade=1
> GeoState: 0x2
>
> Node (null): type: pfGeode = 30064
> num GeoSets: 1
> GeoSet 0: OVERALL - 3 prims, cbind=OVERALL, nbind=PER_VERT, tbind=OFF flatshade=1
> GeoState: 0x2
>
> Node (null): type: pfGroup = 50064
>
> Node (null): type: pfGeode = 30064
> num GeoSets: 1
> GeoSet 0: OVERALL - 3 prims, cbind=OVERALL, nbind=PER_VERT, tbind=OFF flatshade=1
> GeoState: 0x2
>
> Node (null): type: pfGeode = 30064
> num GeoSets: 1
> GeoSet 0: OVERALL - 3 prims, cbind=OVERALL, nbind=PER_VERT, tbind=OFF flatshade=1
> GeoState: 0x2
>
>
>
> The output coming from my debugging messages looks looks like this:
>
> preCallback(): pfGroup 327780 group
> preCallback(): pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> preCallback(): pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> preCallback(): pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> preCallback(): pfGroup 327780 group
> preCallback(): pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> preCallback(): pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> preCallback(): pfGroup 327780 group
> preCallback(): pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> preCallback(): pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
> postCallback() pfGeode 196708 geode
>
> So the postCallback() gets a pfGeode instead of pfGroup or others.
>
> Is this a bug or am I just doing something wrong?
>
> Any help would be appreciated.
>
> Alf
>
>
>
>
>
> --
> =============================================================================
> Alf Ritter _/ _/ _/_/_/_/ alf++at++vr1.engin.umich.edu
> University of Michigan _/_/ _/ _/ "Life would be so much easier
> VR Laboratory _/ _/ _/ _/_/_/ if we could just look at the
> _/_/_/_/ _/ _/ source code."
> (313) 763-7798 _/ _/ _/_/_/ _/ http://www.engin.umich.edu/~aritter/
> =============================================================================
>
>

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:51:36 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.