Re: Representing the result of a node search in HTML

[stefp]

There is Graphviz written by Leon Brocard. This is a Perl front-end to ATT Graphviz. I used the later and it works well for graphs of moderate complexity; I mean by that graphs that can be transformed to forests by removing a few edges.

I have no personal experience of the frond though and don't know if you can generate image-maps. But, I once generated a representation of perl grammar using ATT graphviz.

-- stefp

[blakem]

For an example of what Graphviz and perl can do, check out this nifty DNS mapper.

-Blake

[perrin]

Or this example, by merlyn.

[merlyn]

Yes, of course I'm familiar with GraphViz. If the client will go for it, I'll use an image then. The thought of juggling all that HTML was making my eyes start to hurt. {grin}

In the last hour or so, after seeing the suggestions here for GraphViz, I started whacking out a test harness up that takes parent-child pairs and plots it in a client-side imagemap, allowing selection of the next node to visit. The image is cached using Cache::FileCache so that the browser can fetch it (similar to what I did for my barcode column, and the image map gets generated with stateful values and a specific tag to tell what node got hit. I think I'm on the right track.

The next step is to hook it up with the database. For a given search, I find the entire parental node tree and add those links, and one level down on the child side, for each found node. If there are overlaps, great, GraphViz will find them. If all the results are disjoint, who cares? {grin} Selecting one of the nodes will bring me back to the program, and I can show the GraphViz for that specific node.

I think to make this work I'll also have to cache the results for a given query and for a given single node. Generating the image is expensive, I've seen.

Thanks for your help, y'all!

-- Randal L. Schwartz, Perl hacker